Semaglutide cardiovascular outcomes align more closely with attained dose than achieved weight loss

Semaglutide cardiovascular outcomes align more closely with attained dose than achieved weight loss | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Semaglutide cardiovascular outcomes align more closely with attained dose than achieved weight loss Karthik Murugadoss, A. J. Venkatakrishnan, Christopher Gregg, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9407045/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 8 You are reading this latest preprint version Abstract Semaglutide is often optimized for weight loss, but whether longer-term cardiovascular benefit tracks achieved weight loss or therapeutic exposure levels remains unclear. We conducted a retrospective observational study using a federated, deidentified U.S. electronic health record network and applied multimodal AI-based methods to analyze 47,199 patients with baseline cardiovascular disease. We quantified dose escalation and weight change during the 0–2-year period after semaglutide initiation (landmark period) and assessed cardiovascular outcomes during the 2–4-year period (post-landmark). To mitigate confounding, we performed propensity-matched comparisons during the landmark period, in which semaglutide was associated with lower rates of cardiovascular events than metformin, DPP-4 inhibitors, and SGLT2 inhibitors; however, these findings should be interpreted as associative and remain susceptible to treatment selection bias. Higher maximum semaglutide dose was associated with greater weight loss during the landmark period (3.15% additional weight loss per 1 mg increase; r = − 0.97, P < 0.001) and with lower post-landmark risk of all-cause mortality (RR 0.42, P < 0.001), composite cardiovascular events (death, myocardial infarction, or stroke; RR 0.51, P < 0.001), cerebrovascular disease (RR 0.50, P < 0.001), heart failure (RR 0.55, P < 0.001), and valvular or rheumatic heart disease (RR 0.71, P = 0.025), providing robust associative evidence that supports prospective evaluation of causal relationships. In contrast, greater achieved weight loss during the landmark period did not show a consistent monotonic association with lower post-landmark cardiovascular risk (All-cause mortality p-value = 0.14, composite cardiovascular endpoint p-value = 0.55). Integrating insights from a single cell GLP1R expression atlas was used to infer how semaglutide pharmacology may tie into heart-specific signaling, beyond what is reflected by body-weight reduction alone. The strongest prevalence-weighted GLP1R signal was observed in the pancreas, followed by the heart, where GLP1R engagement potential (GEP) was considerable across cardiomyocyte, cardiac endothelial, and rarer immune cell populations. Together, in this retrospective observational study, semaglutide-associated cardiovascular benefit appears more closely aligned with maximum dose attained than with achieved weight-loss magnitude, supporting prospective validation and motivating beyond-obesity trial designs that integrate whole-body spatial intelligence. Health sciences/Cardiology Health sciences/Diseases Health sciences/Medical research Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Introduction Cardiovascular disease is among the most consequential afterlives of excess adiposity, with myocardial injury, vascular dysfunction, and heart failure causing premature death across millions of patients worldwide 1 . In this landscape, semaglutide has emerged not merely as a drug that lowers body weight, but as a therapy that appears to reshape the clinical terrain of cardiometabolic disease itself. Randomized trials have shown that it can produce substantial and sustained weight reduction while also lowering major cardiovascular event risk in high-risk populations with type 2 diabetes and/orobesity 2–4 . Yet, despite its widening therapeutic reach, semaglutide is still most often understood by patients through the lens of a single visible metric: the magnitude of weight lost. It is increasingly appreciated by physicians that this lens may be too narrow to capture the full biology of the drug. An expanding body of human evidence now suggests that semaglutide influences domains that extend beyond weight loss alone ( Table S1 ) 5 – 7 , including major adverse cardiovascular events (SELECT), symptoms and physical limitations in obesity-related heart failure with preserved ejection fraction (STEP-HFpEF), kidney outcomes in type 2 diabetes with chronic kidney disease (FLOW), cardiovascular outcomes with oral semaglutide (SOUL), and walking capacity in symptomatic peripheral artery disease (STRIDE) 8 – 10 . The recent trial of high-dose semaglutide 7.2 mg provides an additional dose–response benchmark beyond the standard 0.25–2.4 mg range (STEP UP) 7 . Together, these findings raise the possibility that body-weight reduction, although important, is only one visible manifestation of a broader therapeutic biology. This raises a question that is at once practical and mechanistic. In routine care, are dose, weight loss and cardiovascular benefit aligned along a single axis, or does semaglutide exert partially separable cardiovascular effects that are not fully captured by weight-loss magnitude? This question is especially relevant now that both treatment guidelines and commercial development strategies are increasingly organized around organ outcomes and comorbidity reduction, not weight loss alone 4 , 6 . Indeed, weight loss is a composite physiological readout, shaped not only by dose, but by treatment persistence, tolerability, metabolic context, baseline illness and individual biological variability. It is therefore possible that achieved weight reduction, although clinically salient, may serve as an incomplete surrogate for the signaling pathways most relevant to cardiovascular outcomes. Resolving this distinction is important not only for interpreting semaglutide response, but also for how cardiometabolic therapies are compared, optimized, and advanced beyond obesity alone. Here, we addressed this question using deidentified longitudinal clinical data from the nSights Federated EHR Network, contextualized through a whole-body spatial intelligence framework 11 – 14 . As summarized in Fig. 1 , the study couples a landmark clinical design and cohort-definition strategy with a broader conceptual framework linking therapeutic exposure, systems pharmacology, and diagnostics. Our findings position semaglutide as a cardiometabolic therapy whose clinical value may be organized along axes not captured by body-weight reduction alone and may be overlooked by conventional endpoints. Results Higher attained semaglutide dose was associated with greater weight loss during the landmark period Baseline demographic and clinical characteristics of patients stratified by maximum semaglutide dose are summarized in Table S2 . Across the first two years after semaglutide initiation, attained dose and achieved weight loss were closely correlated (Fig. 2 ). Mean maximum weight loss increased from approximately 8% at 0.25 mg to 15% at 2.4 mg, corresponding to an estimated 3.15% greater maximum weight loss per 1 mg increase in attained dose (Pearson R = − 0.97, P < 0.001). The same gradient emerged when patients were stratified by the maximum body-weight reduction achieved during the landmark period. Demographically, when comparing the 25% weight-loss strata, mean age decreased from 58.2 to 56.2 years, female representation increased from 61.8% to 80.9%, baseline BMI rose modestly from 35.2 to 36.9 kg/m 2 , baseline weight rose from 103.1 kg to 105.6 kg, and baseline HbA1c fell from 6.6 to 6.2. Regarding the association between dose and weight loss, the proportion reaching the highest dose category increased from 20.9% to 49.6%, and median semaglutide prescription counts during the first 2 years increased from 1 [IQR 1 to 2] to 5 [IQR 2 to 11] (Table 1 ). Weight loss, therefore, seemed to reflect a combination of cumulative treatment exposure and baseline phenotype. Table 1 Baseline and treatment characteristics across weight-loss categories among semaglutide-treated patients. Weight-loss categories were defined according to the maximum percentage reduction in body weight achieved within each landmark window. Baseline BMI and baseline weight are presented as mean (SD); dose intensity is summarized as the proportion reaching the highest dose category and prescription counts are summarized as median [interquartile range]. Weight Loss Category Number of unique patients Age, Mean (SD) Female, % Baseline BMI, mean (SD) Baseline Weight, kg mean (SD) T2DM Prevalence at Baseline, % Baseline HBA1C, Mean (SD) Max Dose, 2-year n (%) Prescription, 2-year median (IQR) < 5% 4,050 58.2 (12.6) 61.8 35.2 (6.4) 103.1 (25.6) 28.3 6.6 (1.7) 847 (20.9%) 1 [1, 2] 5–10% 2,805 59.8 (12.1) 62.6 35.4 (6.0) 102.6 (23.1) 30.1 6.5 (1.6) 763 (27.2%) 2 [1, 5] 10–15% 1,943 60.0 (11.9) 66.9 35.4 (5.8) 101.5 (22.3) 30.6 6.5 (1.7) 655 (33.7%) 2 [1, 7] 15–20% 1,223 58.4 (12.8) 73.7 35.5 (5.5) 101.0 (21.8) 27.5 6.3 (1.4) 504 (41.2%) 4 [1, 9] 20–25% 707 57.9 (11.6) 76.9 35.5 (5.2) 100.8 (21.3) 25.7 6.2 (1.3) 325 (46.0%) 5 [1, 10] > 25% 884 56.2 (12.1) 80.9 36.9 (5.7) 105.6 (23.6) 25.8 6.2 (1.4) 438 (49.6%) 5 [2, 11] Higher attained dose during the landmark period was associated with lower post-landmark cardiovascular risk When the post-landmark period was examined through the lens of dose attained during the first two years, patients who reached high-dose semaglutide by the landmark (≥ 1.7 mg; n = 3,794) had lower subsequent risk in the 2–4 years (“post-landmark”) period than those who remained in the low-dose range (0.25 to 1.0 mg; n = 8,725) across several major outcomes (Fig. 3 ). High attained dose was associated with lower: all-cause mortality (RR = 0.42, P < 0.001), risk of the composite cardiovascular endpoint (RR = 0.51, P < 0.001), incident cerebrovascular disease (RR = 0.50, P < 0.001), incident heart failure (RR = 0.55, P < 0.001) and incident valvular or rheumatic heart disease (RR = 0.71, P = 0.025) (see Methods ). For all-cause mortality, the composite cardiovascular endpoint, cerebrovascular disease and heart failure, time-to-event analyses showed divergence that emerged within the first 5–10 months post-landmark and persisted through 24 months of follow-up (log-rank p < 0.05 for all analyses) (Fig. 3 ). Cumulative event curves did not significantly differ for incident valvular or rheumatic heart disease. Ischemic heart disease, hypertension, arrhythmias or conduction disorders, peripheral vascular disease or atherosclerosis, cardiomyopathy, aortic disease and venous thromboembolism did not show clear dose-stratified differences in the post-landmark 2–4-year period ( Fig. S1 ). Analysis anchored on the time of first attainment of maximum dose yielded directionally concordant results, again showing lower risk across the same cardiovascular outcomes in the higher-dose group (Fig. S2 ). Dose attainment also correlated with treatment intensity beyond the landmark. During the first two years, median semaglutide prescription counts were 5 [IQR 1 to 9] in the high-dose group versus 1 [IQR 1 to 2] in the low-dose group ( Table S3 ). In the post-landmark period (2–4 years upon initiating semaglutide), prescribing attenuated in both groups, but remained modestly higher in the high-dose group (median 1 [IQR 0 to 2] versus 0 [IQR 0 to 2]). However, larger negative pre-to-post-landmark changes in prescription counts were observed in the high-dose group (− 4 [IQR − 9 to − 1] relative to the low-dose group − 1 [IQR − 2 to − 1]). Higher weight loss during the landmark period was associated with post-landmark glycemic and systolic blood pressure control, but not later cardiovascular risk When the same shared landmark was viewed through the lens of weight change rather than attained dose, the metabolic signals were as expected, but the cardiovascular signal was not. Greater maximum weight loss during the first two years was accompanied by progressively more favorable metabolic-surrogate measures at and after the landmark. Post-landmark HbA1c fell from approximately 6.4% in the 25% group, whereas patient-specific HbA1c change from the pre-treatment baseline deepened from − 0.2% to − 0.8% (one-way ANOVA P < 0.001; Fig. 4 a-b). Post-landmark systolic blood pressure similarly declined from about 133 mmHg to 126 mmHg, with change from baseline shifting from roughly + 2.0 to − 3.5 mmHg (both P < 0.001; Fig. 4 c-d). Post-landmark diastolic blood pressure declined more modestly, from about 78.0 to 75.4 mmHg, with progressively more negative change from baseline across strata (P = 0.002 for change from baseline). By conventional metabolic readouts, larger weight loss behaved exactly as a stronger systemic response should. Using the same 0 to 2-year weight-loss strata and the same 2-year landmark, post-landmark all-cause mortality, the composite cardiovascular endpoint, incident cerebrovascular disease, incident heart failure and incident valvular or rheumatic heart disease did not show a consistent monotonic gradient (Fig. 5 ). Global comparisons were not significant for all-cause mortality (P = 0.144), the composite cardiovascular endpoint (P = 0.547), incident cerebrovascular disease (P = 0.249), incident heart failure (P = 0.887) or incident valvular or rheumatic heart disease (P = 0.431). Among the additional cardiovascular outcomes, only incident peripheral vascular disease or atherosclerosis reached nominal significance across weight-loss strata (global P = 0.033), whereas incident hypertension showed only a suggestive downward trend with greater weight loss (global P = 0.107); the remaining outcomes were not significant (ischemic heart disease, arrhythmias, cardiomyopathy, aortic disease, and venous thromboembolism; Fig. S3 ). Taken together, these findings suggest that greater weight reduction during the 0–2 years landmark period did not necessarily translate into proportionally better cardiovascular outcomes in the 2–4 years post-landmark. Part of the explanation may lie in the fact that weight loss was a moving summary of the first 2 years, whereas drug exposure fell sharply thereafter. When stratified by maximum dose achieved, semaglutide prescribing declined more sharply after the landmark among patients who had reached higher doses than among those who remained in the low-dose group, indicating that more intensive treatment during the first 2 years (landmark period) did not necessarily translate into sustained exposure in the following post-landmark period (Fig. 6 a). A similar pattern was observed across weight-loss strata, with semaglutide prescriptions dropping markedly after the landmark (Fig. 6 b). Median post-landmark prescription counts fell to 0 or 1 in every group, with the steepest declines in the higher-response strata: −4 [IQR − 8 to − 1] in the 15 to 20% group, − 5 [IQR − 9 to − 1] in the 20 to 25% group, and − 4 [IQR − 8 to − 1] in the > 25% group ( Table S4 ). Thus, maximum weight loss accrued during the landmark period did not correspond to sustained semaglutide prescriptions thereafter. Across all weight-loss strata, median prescription counts after maximum weight loss attainment were only 0 or 1, indicating that the subsequent outcome window was generally characterized by sparse semaglutide exposure ( Table S4 ). Consistent with this context, a concurrent analysis anchored at the time of maximum weight loss showed that greater achieved weight loss was not associated with lower cardiovascular risk over the next 24 months (Fig. 7 ). Rather, all-cause mortality and the composite cardiovascular endpoint showed modest upward separation across increasing weight-loss categories, with the highest event probabilities generally observed among patients with the largest weight reductions (P > 0.05, Fig. 7 a-d). Incident cerebrovascular disease followed a similar directional trend, whereas incident heart failure and incident valvular or rheumatic heart disease showed no consistent gradient across weight-loss strata (Fig. 7 e-j). Together, these findings indicate that achieved weight loss alone did not recapitulate the favorable cardiovascular pattern observed across attained semaglutide dose tiers, particularly in a setting where semaglutide prescribing after maximum weight loss was minimal. Matched comparator analyses placed the semaglutide cardiovascular signal in the broader clinical context of antidiabetic medicines Semaglutide carried a broader favorable cardiovascular signal in matched comparator analyses. In 1:1 propensity-score-matched comparisons against metformin, the pooled all cardiovascular conditions subgroup contained 47,199 patients per arm and was closely balanced for age at index (59.4 versus 59.9 years), female representation (62.8% versus 61.3%), baseline BMI (35.6 versus 35.3 kg/m 2 ) and baseline type 2 diabetes prevalence (25.8% versus 26.0%) (Table 2 ). Across individual cardiovascular subgroups, mean age ranged from approximately 59 to 67 years, baseline BMI from 33.8 to 36.3 kg/m 2 , and baseline type 2 diabetes prevalence from 25.4% in aortic disease to 50.8% in heart failure, indicating broadly balanced matching across a clinically heterogeneous population. Table 2 Baseline characteristics of propensity score–matched semaglutide and metformin cohorts across cardiovascular disease subgroups. For each baseline cardiovascular subgroup, patients initiating semaglutide were matched 1:1 to metformin-treated patients using propensity scores estimated from age, sex, baseline BMI and T2DM status. Matched cohort sizes, age at index date (mean ± SD), proportion female, baseline BMI (mean ± SD) and proportion with type 2 diabetes mellitus (T2DM) at baseline are shown. Baseline CV Group Matched Number of unique patients Semaglutide age at index, mean (SD) Metformin age at index, mean (SD) Semaglutide female, % Metformin female, % Semaglutide baseline BMI, mean (SD) Metformin baseline BMI, mean (SD) Semaglutide T2DM, % Metformin T2DM, % All CV conditions 47199 59.4 (12.9) 59.9 (13.3) 62.8 61.3 35.6 (6.0) 35.3 (6.0) 25.8 26 Aortic disease 899 67.0 (10.8) 67.5 (10.5) 35.4 31.4 34.5 (5.5) 34.3 (5.6) 25.4 27.7 Arrhythmia / Conduction disorders 9590 63.6 (13.6) 64.2 (14.7) 56.9 54.3 35.1 (5.7) 34.8 (6.0) 29.8 30.4 Cardiomyopathy 1494 62.0 (13.0) 62.4 (13.1) 48.7 44 35.3 (5.8) 34.7 (6.2) 35.3 35.4 Cerebrovascular disease 2821 65.8 (11.8) 66.1 (12.6) 61.6 60.9 34.2 (5.7) 34.2 (5.6) 39.5 40.3 Heart failure 3777 65.9 (12.0) 66.7 (12.6) 53.2 48.7 36.2 (6.0) 35.4 (6.6) 50.8 50.2 Hypertension 40773 59.8 (12.6) 60.4 (13.1) 62 60.5 35.8 (5.9) 35.5 (6.0) 27.8 28.3 Ischemic heart disease 9203 66.4 (10.3) 66.6 (10.5) 44.8 41.8 34.6 (5.7) 34.2 (5.7) 38.7 37.8 Peripheral / Vascular disease atherosclerosis 3425 66.8 (11.3) 67.2 (11.9) 59.4 58.5 34.5 (6.1) 33.8 (5.9) 43.2 42.5 Valvular / Rheumatic heart disease 3082 65.2 (13.0) 65.7 (13.8) 60.4 58.4 34.5 (5.8) 34.2 (6.0) 34.7 35.1 Venous thromboembolism 2750 58.6 (13.6) 59.4 (14.9) 64.7 61.2 36.3 (6.1) 35.9 (6.3) 29.7 30.5 Analyzed from the time of treatment initiation (Fig. 8 a), semaglutide was associated with lower approximate 2-year event risk than metformin for all-cause mortality (1.7% versus 3.4%; P < 0.001), the composite cardiovascular endpoint (9.7% versus 13.3%; P < 0.001), incident ischemic heart disease (8.0% versus 8.6%; P = 0.002), incident cerebrovascular disease (3.5% versus 4.8%; P < 0.001), incident hypertension (23.8% versus 40.5%; P < 0.001) and incident arrhythmias or conduction disorders (8.5% versus 10.4%; P < 0.001) (Fig. 8 b, Fig. S4a ). Broadly favorable patterns were also observed when comparing semaglutide to dipeptidyl peptidase-4 (DPP-4) inhibitors (Fig. 8 c, Fig. S4b ) or sodium-glucose cotransporter-2 (SGLT2) inhibitors (Fig. 8 d, Fig. S4c ). These comparator analyses support an overall favorable cardiovascular association for semaglutide. Whole-body GLP1R geography provides biologic plausibility for organ-directed effects The landmark analysis raised a biological question: if differences in subsequent cardiovascular outcomes among semaglutide-treated patients were not fully explained by the magnitude of weight loss achieved during the first 2 years, what aspects of GLP1R biology might account for this apparent organ-directed effect? We hypothesized that cell-type-specific receptor prevalence and downstream signaling readiness may be important determinants, and to test this idea, we generated a schematic model of GLP1R engagement potential (GEP) from a whole-body single-cell atlas (Fig. 9 a). This model showed that the pancreas had the highest prevalence-weighted GEP (1.944%), but the heart emerged as the next most prominent tissue by GEP (0.729%) and, notably, as the largest aggregate GLP1R+ reservoir by absolute target load (ATL ; 19,913), marginally exceeding the brain (19,217) and approximately doubling the pancreas (9,967) (Fig. 9 b, Table 3 ). Thus, while the pancreas remained the dominant prevalence-weighted site of GLP1R signal, the heart stood out as a potential extra-pancreatic target organ. Table 3 Single Cell RNA atlas of GLP1R Engagement Potential (GEP) and Absolute Target Load (ATL), with tissues sorted by decreasing GEP%. Tissue Cell Count Expression (CP10K) % Expressing Cells Positive Fraction Cells Expr. GLP1R GEP GEP % ATL pancreas 512856 1.8 1.08 0.0108 5537 0.01944 1.944 9966.6 heart 2730947 1.8 0.405 0.00405 11063 0.00729 0.729 19913.4 axilla 84584 1.73 0.346 0.00346 293 0.0059858 0.59858 506.89 human breast milk 47514 1.71 0.183 0.00183 87 0.0031293 0.31293 148.77 abdomen 429748 1.68 0.092 0.00092 397 0.0015456 0.15456 666.96 spinal cord 44639 1.79 0.081 0.00081 36 0.0014499 0.14499 64.44 mucosa 109366 2.05 0.069 0.00069 76 0.0014145 0.14145 155.8 esophagus 155431 2.2 0.059 0.00059 92 0.001298 0.1298 202.4 embryo 277234 2 0.063 0.00063 176 0.00126 0.126 352 chest wall 19459 1.75 0.057 0.00057 11 0.0009975 0.09975 19.25 prostate gland 207838 2.05 0.039 0.00039 81 0.0007995 0.07995 166.05 stomach 303597 1.93 0.041 0.00041 125 0.0007913 0.07913 241.25 brain 26689529 1.78 0.04 0.0004 10796 0.000712 0.0712 19216.88 adrenal gland 244608 1.73 0.04 0.0004 98 0.000692 0.0692 169.54 adipose tissue 448846 1.87 0.037 0.00037 164 0.0006919 0.06919 306.68 lymph node 1731461 1.73 0.033 0.00033 576 0.0005709 0.05709 996.48 musculature 508632 1.98 0.028 0.00028 144 0.0005544 0.05544 285.12 kidney 1245614 1.85 0.028 0.00028 352 0.000518 0.0518 651.2 exocrine gland 193695 2.02 0.022 0.00022 42 0.0004444 0.04444 84.84 liver 1444734 1.73 0.024 0.00024 346 0.0004152 0.04152 598.58 breast 3100226 1.84 0.022 0.00022 696 0.0004048 0.04048 1280.64 intestine 322530 1.85 0.02 0.0002 64 0.00037 0.037 118.4 lung 4526576 1.83 0.02 0.0002 909 0.000366 0.0366 1663.47 eye 7700797 1.75 0.02 0.0002 1548 0.00035 0.035 2709 tendon of semitendinosus 10533 1.77 0.019 0.00019 2 0.0003363 0.03363 3.54 neck 19097 1.65 0.016 0.00016 3 0.000264 0.0264 4.95 colon 812980 1.83 0.012 0.00012 99 0.0002196 0.02196 181.17 cortex 149400 1.65 0.013 0.00013 19 0.0002145 0.02145 31.35 chest 15413 1.55 0.013 0.00013 2 0.0002015 0.02015 3.1 tongue 38749 1.47 0.013 0.00013 5 0.0001911 0.01911 7.35 testis 18528 1.58 0.011 0.00011 2 0.0001738 0.01738 3.16 small intestine 1191162 1.71 0.01 0.0001 119 0.000171 0.0171 203.49 cerebrospinal fluid 72565 1.76 0.008 8.00E-05 6 0.0001408 0.01408 10.56 spleen 537601 1.69 0.008 8.00E-05 42 0.0001352 0.01352 70.98 skin of body 817861 1.91 0.007 7.00E-05 57 0.0001337 0.01337 108.87 esophagogastric junction 11883 1.54 0.008 8.00E-05 1 0.0001232 0.01232 1.54 endocrine gland 654345 1.7 0.006 6.00E-05 38 0.000102 0.0102 64.6 blood 25136830 1.67 0.005 5.00E-05 1330 8.35E-05 0.00835 2221.1 ureter 41757 1.56 0.005 5.00E-05 2 7.80E-05 0.0078 3.12 ovary 346569 1.73 0.004 4.00E-05 14 6.92E-05 0.00692 24.22 bone marrow 727397 1.65 0.004 4.00E-05 30 6.60E-05 0.0066 49.5 urinary bladder 36100 1.51 0.003 3.00E-05 1 4.53E-05 0.00453 1.51 fallopian tube 238661 1.98 0.002 2.00E-05 4 3.96E-05 0.00396 7.92 large intestine 422380 1.77 0.002 2.00E-05 8 3.54E-05 0.00354 14.16 pleural fluid 83186 1.46 0.002 2.00E-05 2 2.92E-05 0.00292 2.92 uterus 456208 2.05 0.001 1.00E-05 4 2.05E-05 0.00205 8.2 vasculature 349181 1.69 0.001 1.00E-05 4 1.69E-05 0.00169 6.76 placenta 561073 1.68 0.001 1.00E-05 7 1.68E-05 0.00168 11.76 omentum 213994 1.62 0.001 1.00E-05 2 1.62E-05 0.00162 3.24 bladder organ 68317 1.46 0.001 1.00E-05 1 1.46E-05 0.00146 1.46 hindlimb 85565 1.46 0.001 1.00E-05 1 1.46E-05 0.00146 1.46 Within the heart, the most prevalent GLP1R+ populations were cardiomyocytes, particularly atrial cardiomyocytes, followed by endocardial cells, pericytes, and cardiac blood vessel endothelial cells (Fig. 9 c; Table S5 ). Schwann cells and specific cardiac immune-cell populations showed some of the highest per-cell expressions but were rare and ultra-rare, respectively. The GLP1R+ rare cardio-immune cells included CD4-positive alpha-beta T cells and mature natural killer T cells. Among cardiac myocytes, regular atrial cardiac myocytes showed GLP1R expression of 1.83 CP10K, with 992 expressing cells and 1.364% of cells positive, yielding a salience score of 2.492, whereas regular ventricular cardiac myocytes showed comparable per-cell expression intensity (1.72 CP10K) but substantially lower prevalence, with 926 expressing cells and only 0.382% of cells positive, yielding a lower salience score of 0.659. A second ventricular population, ventricular cardiac muscle cells, showed slightly higher per-cell expression (2.03 CP10K) but extremely sparse positivity (0.116%; salience 0.236). The dominant contribution in the pancreas arose from the exocrine compartment including acinar cells, which had the top-ranked GEP status across all analyzed organs of the human body (Fig. 9 d). Consistent with these insights from single cell analysis, whole-body bulk RNA sequencing data shows that GLP1R expression was highest in the pancreas, followed next by the heart, and then by selected neuroendocrine tissues ( Fig. S5 ). Specifically, heart muscle ranked among the highest non-pancreatic tissues for GLP1R expression, significantly exceeding most peripheral organs, including lung, liver, kidney, and adipose tissue. Within cardiac compartments, GLP1R expression showed meaningful signal in atrial tissue and, to a lesser extent, ventricular tissue, with distributions indicating heterogeneous but reproducible expression across samples. Atrial samples (n = 432) and left ventricular samples (n = 432, 550) showed median TPM was 1.01 in atrial tissues (Q1–Q3, 0.56–1.47) versus 0.44–0.46 in left ventricular tissues (Q1–Q3, 0.17–0.87). Triangulating bulk RNA and single-cell RNA-seq, atrial tissue showed a stronger GLP1R signal than ventricular tissue, with the bulk distributions shifted upward in atrium across the full range of samples: median TPM was 1.01 in atrial tissues (Q1–Q3, 0.56–1.47) versus 0.44–0.46 in left ventricular tissues (Q1–Q3, 0.17–0.87). Together, these data indicate that the stronger atrial GLP1R signal in bulk tissue is better explained by higher prevalence of GLP1R-positive cardiomyocytes in the atrium than by markedly higher per-cell expression intensity. Collectively, these data establish that while the pancreas dominates GLP1R expression, the heart represents a prominent extra-pancreatic site of GLP1R transcription, supporting the plausibility of cardiac cells as biologically relevant targets for GLP1 receptor agonists which may contribute to their broader effects beyond body-weight reduction. Discussion Semaglutide appears to write on more than one physiologic axis The most important lesson of these data is that semaglutide seems to leave at least two signatures on human physiology. One signature is immediately visible: higher attained dose tracked closely with greater weight loss across the first 2 years after initiation. The second is quieter, but potentially more consequential: later cardiovascular benefit aligned more closely with dose attained than with the magnitude of weight loss itself. If weight loss were the dominant mediator of cardiovascular protection in this setting, one would expect a graded decline in post-landmark events across progressively deeper weight-loss strata. We did not observe that pattern. Instead, conventional metabolic markers such as HbA1c and blood pressure improved monotonically with weight loss, whereas post-landmark cardiovascular outcomes did not. In this cohort, body weight change therefore behaved as an important pharmacodynamic readout, but not as a complete surrogate for the biology most relevant to later cardiovascular risk. The prescribing trajectory after the landmark sharpens that distinction. Weight loss was a cumulative summary of the first 2 years, whereas semaglutide prescribing attenuated substantially in the subsequent interval in which cardiovascular outcomes were assessed. In that setting, a patient's maximum weight loss may memorialize earlier exposure without faithfully representing the therapeutic state that followed. That temporal mismatch provides one explanation for why weight-loss magnitude tracked systemic metabolic improvement yet failed to organize later cardiovascular risk as cleanly as dose attainment did. Cardiac GLP1R geography offers biologic plausibility for organ-directed effects Our atlas-based analyses offer a biologically plausible framework for this dissociation. The pancreas remained the dominant prevalence-weighted site of GLP1R signal, as expected for an incretin therapy. But the heart emerged immediately behind it by GLP1R engagement potential and, notably, as the largest transcript-weighted reservoir of GLP1R-positive cells (Fig. 9 , Table 3 ). Within that reservoir, signal was not confined to a single rare niche. It extended across cardiomyocyte and contractile populations, endothelial and endocardial compartments, stromal and perivascular states, and selected immune populations. These data do not demonstrate direct cardiac causality. They do, however, make it biologically plausible that the cardiovascular effects of semaglutide are written partly through organ-directed biology in the heart and vasculature, and not solely through the arithmetic of kilograms lost. That broader view is also consistent with the emerging clinical literature. Semaglutide has shown cardiovascular and heart-failure benefits across multiple contexts, including obesity with established cardiovascular disease, obesity-related HFpEF, chronic kidney disease with diabetes, oral semaglutide cardiovascular outcomes, and symptomatic peripheral artery disease 3 – 10 . Our data suggest that these benefits may be better understood as the net result of several partially overlapping programs, including systemic metabolic change, vascular-interface effects, and myocardial or immunometabolic signaling. An extended mechanistic interpretation of the whole-body atlas, cardiac cell-state programs, and functional-readiness modules is provided in the Supplementary Discussion ( Figs. S5 - S6 ). Finally, differential expression of GLP1R in the heart also motivates future analyses of cardiac remodeling in semaglutide-treated patients through radiological diagnostic imaging data collected before and after semaglutide treatment initiation. Implications for endpoint selection, dose optimization and trial design The matched comparator analyses place this within-drug observation into a broader therapeutic frame. From treatment initiation, semaglutide showed a favorable cardiovascular association relative to metformin, DPP-4 inhibitors, and SGLT2 inhibitors across several major endpoints. Taken together with the landmark analyses, the implication is not that weight loss is irrelevant. Rather, it is that cardiovascular medicine should resist reducing the value of semaglutide to a single visible phenotype. Weight loss may be the most conspicuous manifestation of drug activity, but it is not necessarily the most discriminating measure of organ benefit. This distinction matters for both clinical practice and drug development. In routine care, incretin therapies are often titrated and compared through the lens of body-weight reduction. Yet the outcomes that matter most to patients with cardiovascular disease are myocardial infarction, stroke, heart failure, disability, and death. If these data are confirmed prospectively, then dose attainment, exposure patterns and organ-specific endpoints should stand alongside weight loss, rather than behind it, in the optimization of semaglutide and in the design of next-generation trials. The broader opportunity is methodological as well: a whole-body spatial intelligence framework that brings together federated longitudinal EHR data, therapeutic exposure, physiologic biomarkers, and molecular atlases may help reveal therapeutic value axes that conventional obesity endpoints do not fully capture. Limitations and future directions This study has important limitations. As an observational analysis of de-identified EHR data, it remains vulnerable to residual confounding, treatment-selection bias, unequal measurement density, unmeasured adherence, and misclassification of outcomes defined from diagnosis codes rather than adjudicated events. Dose was inferred from prescriptions rather than verified administration. The marked attenuation of prescribing after the landmark further complicates interpretation of exposure during the later risk window. The transcriptomic analyses provide biologic plausibility, not mechanistic proof: single-cell and bulk RNA measurements do not establish receptor occupancy, downstream pathway activation, or causal mediation in human cardiac tissue. The metrics GEP and ATL should be interpreted as atlas-derived pharmacologic heuristics rather than literal measures of in vivo receptor occupancy or absolute organ-level target burden. In particular, ATL is influenced by uneven cell recovery and differential susceptibility of cell types to dissociation, capture and sequencing in single-cell RNA-sequencing datasets, and therefore should not be treated as a direct estimate of absolute cellular proportions across tissues. Accordingly, these metrics are most useful for prioritizing candidate tissues and cell compartments for semaglutide responsiveness, rather than for making exact quantitative claims about organ-level pharmacologic exposure. The absence of a monotonic cardiovascular gradient across weight-loss strata should therefore not be read as evidence that large semaglutide-mediated weight loss is harmful. It should be read more carefully, and more usefully, as evidence that weight-loss magnitude alone does not fully explain the later cardiovascular signal observed here. Even with those caveats, a coherent picture emerges. Semaglutide appears to act not only as an agent of body-mass reduction, but as a therapy whose clinical signature extends into cardiovascular biology. One part of that signature is visible on the scale. Another may be written more quietly in the heart, the vessel wall, and the clinical arc of risk over time. The next generation of cardiometabolic trials will need to read both. Methods Data Source and Study Design We conducted a retrospective cohort study using deidentified longitudinal electronic health record data from the nSights Federated EHR Network 11 . All analyses were conducted on deidentified data. Study Cohorts and Exposure Definitions From approximately 29 million patients, 505,874 had at least one semaglutide prescription record, of whom 269,390 initiated semaglutide between March 2018 and January 2024; 47,199 of these had baseline cardiovascular disease (see below). Medication exposure was ascertained from prescription or medication event records using the recorded event timestamp. The index date was defined as the date of the first qualifying prescription for the cohort-defining medication or medication class. Follow-up data were available through January 2026. To create incident-user cohorts with minimal treatment contamination, patients were excluded if they had exposure to metformin, DPP-4 inhibitors, or SGLT2 inhibitors during the 365 days before the index date or at any time after the index date. Constituent medications within DPP-4 and SGLT2 groups are listed in Table S6 . Other GLP-1RA medications (tirzepatide, liraglutide, dulaglutide, exenatide, albiglutide, lixisenatide) were used as exclusion exposures when defining the semaglutide cohort. Demographic and Clinical Covariates Demographic data included age, sex, and death date. Age at index was obtained from the index medication record. Baseline BMI was defined as the BMI measurement closest to the index date, recorded from 365 days before through 14 days after the index date. Baseline weight for the weight-loss analyses was defined separately as the weight measurement closest to the index date, recorded from 90 days before through 14 days after the index date. Baseline type 2 diabetes was defined by the presence of at least three encounters carrying a type 2 diabetes code on three distinct dates at any time before the index date. The same logic was used to define prespecified baseline cardiovascular comorbidity groups. These groups were ischemic heart disease, cerebrovascular disease, hypertension, heart failure, arrhythmias or conduction disorders, peripheral vascular disease, or atherosclerosis, valvular or rheumatic heart disease, cardiomyopathy, aortic disease, and venous thromboembolism. Diagnosis definitions were based on prespecified ICD-9 and ICD-10 code prefixes; the full diagnosis code sets are provided in Table S7 . An all-cardiovascular conditions subgroup, defined as the union of these groups, served as the primary analytic cohort for the dose- and weight-loss-based landmark analyses; individual cardiovascular subgroups were examined separately in the comparative outcome analyses. We also calculated the interval between the index date and the most recent qualifying preindex diagnosis code for each cardiovascular subgroup; these summaries are provided in Table S8 . Cardiovascular Outcomes The primary comparative outcomes were all-cause mortality, composite cardiovascular events (all-cause death, MI, or stroke), and incident cardiovascular diagnoses. All-cause mortality was defined by the recorded death date. The composite cardiovascular endpoint was defined as all-cause death or the first post-index diagnosis consistent with acute myocardial infarction, acute coronary syndrome, or cerebrovascular event; component ICD-9 and ICD-10 code definitions are provided in Table S9 . Incident cardiovascular outcomes included incident ischemic heart disease, cerebrovascular disease, hypertension, heart failure, arrhythmias or conduction disorders, peripheral vascular disease, or atherosclerosis, valvular or rheumatic heart disease, cardiomyopathy, aortic disease, and venous thromboembolism. For each incident outcome analysis, patients with evidence of that same condition before the start of follow-up were excluded. Dose–Weight-Loss Relationship To characterize the dose–response relationship between semaglutide and weight reduction, we examined the distribution of maximum percent body-weight change achieved within the 2-year landmark window according to the maximum semaglutide dose reached during that period. Dose categories corresponded to labeled dosing increments for injectable semaglutide formulations (Ozempic and Wegovy): 0.25 mg, 0.5 mg, 1.0 mg, 1.7 mg, 2.0 mg, and 2.4 mg or greater. Observed semaglutide dose values were used when directly available in the medication record, and missing dose values were extracted from medication descriptions using a large language model-assisted dose-mapping workflow based on GPT-OSS-20B 15 . The resulting normalized dose values were used to assign patients to labeled dosing categories. This workflow increased dose ascertainment coverage from 56.0% to 98.2% of semaglutide prescription records. Manual review of 148 unique medication descriptions confirmed correct dose assignment in every case; the only apparent ambiguities (13 descriptions, 8.8%) arose from multi-dose pen devices listing two possible doses (e.g., "0.25 mg or 0.5 mg"), which were assigned the higher of the two labeled doses as a convention. A linear regression was fitted to the mean maximum weight-loss percentage at each dose level to quantify the overall dose-response trend. Landmark Analyses of Cardiovascular Outcomes by Dose and Weight Loss at 2 Years We performed within-drug landmark analyses to evaluate the association of dose intensity and weight loss with subsequent cardiovascular outcomes among semaglutide-treated patients. The landmark was defined at 2 years after the index date. Outcomes were assessed only after the landmark date, and patients who experienced the outcome of interest on or before the landmark were excluded from the corresponding analysis. In the post-landmark period, patients were followed until the outcome of interest, death, or the last observed clinical record, whichever occurred first. For dose-based analyses, we used all prescriptions from the index date through the landmark to define the maximum dose reached by the landmark. Semaglutide dose was dichotomized as low (0.25–1.0 mg) or high (1.7 mg or greater). Patients were considered analyzable if follow-up extended through the landmark and at least one dose-bearing prescription was available by that time point. For weight-loss-based analyses, baseline weight was defined as the measurement closest to the index date, from 90 days before through 14 days after index. Post-index weight change was evaluated beginning 15 days after index through the landmark to avoid overlap between baseline ascertainment and follow-up weight assessment. Maximum percent body-weight change was calculated as the difference between baseline weight and the lowest observed weight before the landmark, divided by baseline weight, multiplied by 100. Patients were categorized into prespecified strata: less than 5%, 5–10%, 10–15%, 15–20%, 20–25%, and 25% or greater. Analyses were restricted to patients with an available baseline weight and at least one follow-up weight before the landmark. For both dose and weight-loss landmark analyses, cardiovascular outcomes were evaluated using cumulative event curves over the 24-month post-landmark period and overall post-landmark event proportions. Post-Landmark Metabolic Surrogates, Prescription Continuity, and Sensitivity Analyses Post-landmark metabolic surrogates included HbA1c and systolic and diastolic blood pressure. For each patient, the post-landmark value was defined as the closest measurement to the landmark date within the 24-month post-landmark period. Change from baseline was calculated as the difference between the post-landmark value and the most recent pre-treatment measurement. These surrogates were summarized across weight-loss strata. To assess semaglutide prescription continuity relative to the landmark, we computed per-patient prescription counts in two windows: the pre-landmark window spanning from the first semaglutide prescription date to the 2-year landmark date, and the post-landmark window spanning the 24 months following the landmark date (truncated at the last observation date). The per-patient difference in prescription count (post minus pre) was computed and summarized as median [IQR] stratified by dose group and weight-loss category. In addition to the shared 2-year landmark, two sensitivity analyses were performed: the first anchored follow-up to the date of first attainment of the patient's maximum semaglutide dose, and the second to the date of maximum body-weight reduction. In both cases, outcomes were evaluated over the subsequent 24 months to test whether the patterns observed with the fixed landmark were robust when aligned to the time points at which peak exposure or peak response was achieved. Propensity-Matched Comparative Cardiovascular Analyses For comparator analyses, three additional mutually exclusive incident-user cohorts were constructed for metformin, DPP-4 inhibitors, and SGLT2 inhibitors using the same index-date and exclusion logic described above. Within each baseline cardiovascular subgroup, we performed pairwise comparisons anchored on semaglutide as the primary exposure cohorts. Separate analyses were conducted for semaglutide versus metformin, DPP-4 inhibitors, and SGLT2 inhibitors. Propensity scores were estimated with logistic regression using age at index, sex, baseline BMI, and baseline type 2 diabetes status. Patients were matched 1:1 without replacement using nearest-neighbor matching and a caliper of 0.2 on the propensity score scale. Follow-up for comparative outcome analyses began on the index date. Patients were followed until the outcome of interest, death, or the last observed clinical record. For incident cardiovascular outcomes, patients with prevalent disease of the same type at baseline were excluded from that specific analysis. We summarized matched cohort size, event counts, person-time, and cumulative 2-year event estimates using Kaplan-Meier methods and log-rank P values. Analysis of GLP1R Expression Using Single-Cell and Bulk Tissue Atlases Single-cell expression of GLP1R was obtained from the CELLxGENE data portal 16 ( https://cellxgene.cziscience.com/ ), leveraging 1556 publicly available human single-cell and single-nucleus RNA sequencing datasets corresponding to 86 million cells from 11,633 human donors. Using the CELLxGENE Census, GLP1R expression was obtained across available cell types. Gene expression values were extracted as normalized counts (counts per 10,000 [CP10K], log-transformed where applicable) and analyzed at the cell-type level using curated annotations provided within each dataset. For each cell type, GLP1R expression was summarized as both mean normalized expression and the proportion of expressing cells (non-zero counts). Where multiple datasets contributed to the same tissue, results were aggregated to ensure robustness across cohorts, and analyses were restricted to cell types with adequate representation to minimize sparsity-driven artifacts. Bulk tissue-level expression of GLP1R was obtained from the Human Protein Atlas (proteinatlas.org). Normalized transcript expression values (e.g., nTPM) were retrieved across human tissues, with particular emphasis on cardiac subregions where available. Single Cell RNA-sequencing profiling of GLP1R across human body tissues ​​ Single-cell GLP1R atlas data were analyzed as follows. For each tissue-cell annotation, Expression (CP10K), number of GLP1R-positive cells, and percentage of expressing cells were analyzed. The primary metric was GLP1R Engagement Potential (GEP) , defined as Expression (CP10K) × (% expressing cells / 100). Conceptually, GEP is a prevalence-weighted expression metric: it becomes high when a cell class has both appreciable GLP1R transcript abundance and a meaningful proportion of cells that are GLP1R-positive. It is therefore useful for ranking which cellular compartments are most likely to be engaged under a broad systemic exposure model because it balances intensity and prevalence rather than privileging either one alone. A compartment with very high expression in only a tiny number of cells may not rank as highly by GEP as a compartment with slightly lower expression but a much broader GLP1R-positive fraction. The other salient metric was Absolute Target Load (ATL) , defined as the product of per-cell GLP1R expression and the absolute number of GLP1R-positive cells in that compartment. This is computed as Expression (CP10K) × NumberPositiveCells, where NumberPositiveCells is the number of cells expressing GLP1R in that tissue-cell class. Conceptually, ATL estimates the total transcript-weighted burden of GLP1R-positive cells and is therefore a whole-compartment burden metric rather than a prevalence-weighted one. It becomes especially informative at the tissue or organ level, where a compartment may have only modest prevalence but still contain a very large absolute reservoir of GLP1R-positive cells because the total sampled population is large. For this reason, ATL is particularly useful for understanding overall organ-scale target burden and for distinguishing tissues that may represent major aggregate pharmacologic reservoirs even if their percentage of positive cells is not the highest. Stable cell classes were filtered for visualization using Cell Count > = 100 and NumberPositiveCells > = 2. Statistical Analysis Continuous variables were summarized as means with standard deviations or medians with interquartile ranges, as appropriate; categorical variables were summarized as counts and percentages. In the comparative cardiovascular outcome analyses, between-cohort differences were assessed using propensity score–matched Kaplan-Meier estimates of 2-year event risk; statistical significance in KM-difference heatmaps was determined by log-rank test. In the dose landmark analyses, cumulative event distributions were compared between high-dose and low-dose groups using log-rank tests, and post-landmark event rates were compared using relative risks with associated Wald p-values. In the weight-loss landmark analyses, cumulative event distributions and post-landmark event rates across weight-loss strata were compared using a global log-rank test. Trends in post-landmark metabolic surrogates (HbA1c, systolic and diastolic blood pressure) across weight-loss strata were assessed using one-way ANOVA. Analyses were performed in Python 3.13.1 using pandas 3.0.0, NumPy 2.4.2, SciPy 1.17.0, and Matplotlib 3.10.8; propensity score modeling, survival analyses and single-cell expression profiling were implemented within the same analytic pipeline. Data Source This study analyzed de-identified EHR data from academic medical centers in the United States via the nference nSights Analytics Platform. Prior to analysis, all data underwent expert determination de-identification satisfying HIPAA Privacy Rule requirements (45 CFR § 164.514(b)(1)), employing a multi-layered transformation approach for both structured data (cryptographic hashing of identifiers, date-shifting, geographic truncation) and unstructured clinical text (ensemble deep learning and rule-based methods with > 99% recall for personally identifiable information detection) 17 , 18 . nference established secure data environments within each participating center, housing these de-identified patient data governed by expert determination. These de-identified data environments were specifically designed to enable data access and analysis without requiring Institutional Review Board oversight, approval, or exemption confirmation. Accordingly, informed consent and IRB review were not required for this study. Data Availability This study involves the analysis of de-identified Electronic Health Record (EHR) data via the nference nSights Federated Clinical Analytics Platform (nSights). Data shown and reported in this manuscript were extracted from this environment using an established protocol for data extraction, aimed at preserving patient privacy. The data has been de-identified pursuant to an expert determination in accordance with the HIPAA Privacy Rule. Any data beyond what is reported in the manuscript, including but not limited to the raw EHR data, cannot be shared or released due to the parameters of the expert determination to maintain the data de-identification. The corresponding author should be contacted for additional details regarding nSights. De-identification and HIPAA compliance certification Prior to analysis, all EHR data were de-identified under an expert determination consistent with the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule (45 CFR § 164.514(b)(1)). The de-identification methodology employed a multi-layered transformation approach to both structured and unstructured data fields 17 , 18 . In structured data, direct identifiers including patient names and precise geographic locations were excluded entirely, while indirect identifiers underwent specific transformations: patient identifiers, medical record numbers, and accession numbers were replaced with one-way cryptographic hashes using confidential salts to preserve linkage across patient encounters; all dates were shifted backward by patient-specific random offsets (1–31 days) to preserve temporal relationships while obscuring exact event timing; the ZIP codes were truncated to two-digit state-level resolution; and continuous variables including age, height, weight, and body mass index were thresholded to prevent identification of extreme values (for example, ages ≥ 89 years transformed to ‘89+’ and BMI > 40 transformed to ‘40+’). In unstructured clinical text, an ensemble de-identification system that combines attention-based deep learning models with rule-based methods achieved an estimated > 99% recall for personally identifiable information (PII) detection, with detected identifiers replaced by plausible fictional surrogates 17 . Data Harmonization To address heterogeneity in EHR data, we harmonized clinical variables including medications, anthropometric measurements, and diagnoses to standardized concepts. For medications, we first constructed a standardized drug concept database combining the nSights knowledge graph with RXNorm ( https://www.nlm.nih.gov/research/umls/rxnorm/index.html ) hierarchies to capture ingredient, brand, and dose-specific information 11 . EHR medication records were matched using a hierarchical approach prioritizing RXNorm codes when available, followed by ingredient-level matching, and finally natural language processing and pattern matching on free-text medication orders when structured codes were absent. For anthropometric measurements (height, weight, BMI), we created a unified vocabulary from SNOMED ( https://www.snomed.org/ , https://athena.ohdsi.org ) and LOINC ( https://loinc.org/ ) terminologies and matched EHR measurement descriptions using standardized text matching algorithms with abbreviation expansion and synonym resolution; ambiguous mappings were resolved using OpenAI GPT-4o ( https://platform.openai.com/docs/models/gpt-4o ) with summary statistics as context, followed by manual verification. For diagnoses, we developed a hierarchical disease concept database from the nSights knowledge graph and matched EHR diagnosis descriptions and codes by identifying the most specific common child concept in the hierarchy. This approach enabled consistent identification of clinical entities while preserving granularity where available. Declarations Conflict of Interest Statement The authors are employees of nference, inc., which conducts research collaborations with various biopharmaceutical companies whose therapeutic products are included in this study. None of these companies, nor any other nference collaborator, funded, supported, or had any role in the independent study design, data acquisition, analysis, interpretation, manuscript preparation, or the decision to submit this work for publication. All analyses were conducted by the authors using de-identified electronic health record data. The authors declare no additional competing interests. Competing Interests The authors are employees of nference, inc., which conducts research collaborations with various biopharmaceutical companies whose therapeutic products are included in this study. None of these companies, nor any other nference collaborator, funded, supported, or had any role in the independent study design, data acquisition, analysis, interpretation, manuscript preparation, or the decision to submit this work for publication. All analyses were conducted by the authors using de-identified electronic health record data. The authors declare no additional competing interests. Funding This research received no external funding. Author Contribution V.S. conceived the study and supervised the overall project. A.J.V. and K.M. designed the analytical framework. K.M. led the data curation and statistical analyses, with support from A.J.V and V.S. K.M. implemented the AI-enabled phenotyping pipelines and performed the longitudinal modeling. A.J.V., K.M. and V.S. contributed equally to data interpretation. V.S., A.J.V. and K.M. drafted the initial manuscript and critically revised the manuscript with inputs from C.G. All authors reviewed, edited, and approved the final version of the manuscript. Acknowledgement We thank the nference engineering team for the development of the nSights federated AI platform, and Patrick Lenehan for critical review and manuscript feedback. Data Availability This study involves the analysis of de-identified Electronic Health Record (EHR) data via the nference nSights Federated Clinical Analytics Platform (nSights). Data shown and reported in this manuscript were extracted from this environment using an established protocol for data extraction, aimed at preserving patient privacy. The data has been de-identified pursuant to an expert determination in accordance with the HIPAA Privacy Rule. Any data beyond what is reported in the manuscript, including but not limited to the raw EHR data, cannot be shared or released due to the parameters of the expert determination to maintain the data de-identification. The corresponding author should be contacted for additional details regarding nSights. References Zhou, X.-D. et al. Burden of disease attributable to high body mass index: an analysis of data from the Global Burden of Disease Study 2021. EClinicalMedicine 76, 102848 (2024). Wilding, J. P. H. et al. Once-Weekly Semaglutide in Adults with Overweight or Obesity. N Engl J Med 384, 989–1002 (2021). Marso, S. P. et al. Semaglutide and Cardiovascular Outcomes in Patients with Type 2 Diabetes. N Engl J Med 375, 1834–1844 (2016). Lincoff, A. M. et al. Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. N Engl J Med 389, 2221–2232 (2023). Kosiborod, M. N. et al. Semaglutide in Patients with Heart Failure with Preserved Ejection Fraction and Obesity. N Engl J Med 389, 1069–1084 (2023). Deanfield, J. et al. Semaglutide and cardiovascular outcomes in patients with obesity and prevalent heart failure: a prespecified analysis of the SELECT trial. Lancet 404, 773–786 (2024). Wharton, S. et al. Once-weekly semaglutide 7·2 mg in adults with obesity (STEP UP): a randomised, controlled, phase 3b trial. Lancet Diabetes Endocrinol 13, 949–963 (2025). Perkovic, V. et al. Effects of Semaglutide on Chronic Kidney Disease in Patients with Type 2 Diabetes. N Engl J Med 391, 109–121 (2024). McGuire, D. K. et al. Oral Semaglutide and Cardiovascular Outcomes in High-Risk Type 2 Diabetes. N Engl J Med 392, 2001–2012 (2025). Bonaca, M. P. et al. Semaglutide and walking capacity in people with symptomatic peripheral artery disease and type 2 diabetes (STRIDE): a phase 3b, double-blind, randomised, placebo-controlled trial. Lancet 405, 1580–1593 (2025). Venkatakrishnan, A. J. et al. Clinical nSights: A software platform to accelerate real world oncology analyses. Journal of Clinical Oncology (2024) doi: 10.1200/JCO.2024.42.16_suppl.e23316 . Venkatakrishnan, A. J., Murugadoss, K. & Soundararajan, V. Decoding the hallmarks of GLP-1RA weight-loss super responders. medRxiv 2025.11.15.25340314 (2025) doi: 10.1101/2025.11.15.25340314 . Venkatakrishnan, A. J., Murugadoss, K. & Soundararajan, V. Weight Loss Dynamics and Health Burden Changes with Tirzepatide versus Semaglutide. medRxiv 2025.11.30.25341294 (2025) doi: 10.64898/2025.11.30.25341294 . Murugadoss, K., Varma, G., Venkatakrishnan, A. J., Gibson, M. C. & Soundararajan, V. Weight trajectories after last Tirzepatide or Semaglutide prescription across a federated health network. medRxiv 2026.01.26.26344839 (2026) doi: 10.64898/2026.01.26.26344839 . OpenAI et al. gpt-oss-120b & gpt-oss-20b Model Card. (2025). CZI Cell Science Program et al. CZ CELLxGENE Discover: a single-cell data platform for scalable exploration, analysis and modeling of aggregated data. Nucleic Acids Res 53, D886–D900 (2025). Murugadoss, K. et al. Building a best-in-class automated de-identification tool for electronic health records through ensemble learning. Patterns (N Y) 2, 100255 (2021). Murugadoss, K. et al. Scaling text de-identification using locally augmented ensembles. medRxiv 2024.06.20.24308896 (2024) doi: 10.1101/2024.06.20.24308896 . Packer, M. et al. Tirzepatide for Heart Failure with Preserved Ejection Fraction and Obesity. N Engl J Med 392, 427–437 (2025). McLean, B. A., Wong, C. K., Kabir, M. G. & Drucker, D. J. Glucagon-like Peptide-1 receptor Tie2 + cells are essential for the cardioprotective actions of liraglutide in mice with experimental myocardial infarction. Mol Metab 66, 101641 (2022). Website. https://www.fda.gov/news-events/press-announcements/fda-approves-fourth-product-under-national-priority-voucher-program-higher-dose-semaglutide . Kosiborod, M. N. et al. Semaglutide in Patients with Obesity-Related Heart Failure and Type 2 Diabetes. N Engl J Med 390, 1394–1407 (2024). Helmstädter, J. et al. Endothelial GLP-1 (Glucagon-Like Peptide-1) Receptor Mediates Cardiovascular Protection by Liraglutide In Mice With Experimental Arterial Hypertension. Arterioscler Thromb Vasc Biol 40, 145–158 (2020). Park, B. et al. GLP-1 receptor agonists and atherosclerosis protection: the vascular endothelium takes center stage. Am J Physiol Heart Circ Physiol 326, H1159–H1176 (2024). Additional Declarations Competing interest reported. The authors are employees of nference, inc., which conducts research collaborations with various biopharmaceutical companies whose therapeutic products are included in this study. None of these companies, nor any other nference collaborator, funded, supported, or had any role in the independent study design, data acquisition, analysis, interpretation, manuscript preparation, or the decision to submit this work for publication. All analyses were conducted by the authors using de-identified electronic health record data. The authors declare no additional competing interests. Supplementary Files Soundararajansupplement.pdf Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 17 May, 2026 Reviewers agreed at journal 13 May, 2026 Reviewers agreed at journal 24 Apr, 2026 Reviewers agreed at journal 22 Apr, 2026 Reviewers invited by journal 22 Apr, 2026 Editor assigned by journal 19 Apr, 2026 Submission checks completed at journal 16 Apr, 2026 First submitted to journal 13 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9407045","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":630815263,"identity":"5ba9688f-fb02-4c68-9eaf-e22ea6346ca5","order_by":0,"name":"Karthik Murugadoss","email":"","orcid":"","institution":"Nference (United States)","correspondingAuthor":false,"prefix":"","firstName":"Karthik","middleName":"","lastName":"Murugadoss","suffix":""},{"id":630815265,"identity":"b86d476e-258b-471e-ac73-8dff49bb346b","order_by":1,"name":"A. J. Venkatakrishnan","email":"","orcid":"","institution":"Nference (United States)","correspondingAuthor":false,"prefix":"","firstName":"A.","middleName":"J.","lastName":"Venkatakrishnan","suffix":""},{"id":630815267,"identity":"fb7ca4fa-3e09-452c-be6a-98325497aa2c","order_by":2,"name":"Christopher Gregg","email":"","orcid":"","institution":"Nference (United States)","correspondingAuthor":false,"prefix":"","firstName":"Christopher","middleName":"","lastName":"Gregg","suffix":""},{"id":630815269,"identity":"7a8a04b3-24e2-4931-a7e5-9600d9c49cb2","order_by":3,"name":"Venky Soundararajan","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2klEQVRIiWNgGAWjYHACZgaGCgZ+UrWcYZBsgPPZiNHC2EaKFoPjvYeNeefdkeCXPnzswY8Khjx++QYCWs6cS07m3fZMQrIvLd2w5wxDsWQbAVskZ+QYH+bddrjO4AyPmTTQhYkbjhHSMv8NUMucwxL2Z/i/STP+Y0jcT0gLvwSPcTJvw2EJAx4eNmnGBqAthLzPz5NjbDjn2GEJiTNsZpI9xyQSZxxLwK+Fjf2MscSbmsMS/D3MzyR+1Ngk9jcfIGANEDDxINgShJWDAOMP4tSNglEwCkbBSAUA5jY713RXHt4AAAAASUVORK5CYII=","orcid":"","institution":"Nference (United States)","correspondingAuthor":true,"prefix":"","firstName":"Venky","middleName":"","lastName":"Soundararajan","suffix":""}],"badges":[],"createdAt":"2026-04-13 17:23:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9407045/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9407045/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108406317,"identity":"68bf6800-24c2-4cd6-805f-65c7e0bbea54","added_by":"auto","created_at":"2026-05-04 09:41:58","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":438205,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStudy design and timeline for the semaglutide cardioprotective landmark analysis. \u003c/strong\u003e(A) Patients initiating semaglutide between March 2018 and January 2024 were identified across the nSights Federated EHR Network. The index date was defined as the date of first semaglutide prescription. An incident user design was employed, requiring a one-year pre-index medication washout period to exclude patients with prior exposure to comparator medications. Baseline cardiovascular (CV) burden was ascertained from the entire available electronic health record (EHR) history prior to the index date, requiring ≥3 distinct ICD code dates per condition. Baseline body mass index (BMI) and body weight were captured as the nearest recorded value within windows of −365 to +14 days and −90 to +14 days relative to the index date, respectively. A two-year landmark date was defined for each patient, and only patients with documented follow-up through this date were included in the primary analysis. Dose escalation and weight loss were quantified over the landmark period (0–2 years). Post-landmark cardiovascular outcomes: all-cause mortality, composite cardiovascular events, and incident CV conditions, were assessed from the landmark date through the data cutoff of January 2026. (B) Cohort funnel showing the inclusion of patients based on semaglutide treatment and prior cardiovascular disease. (C) Conceptual framework for whole-body spatial intelligence connecting therapeutic exposure, systems pharmacology and 4D anatomical changes captured using radiological data.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/c77db839c6ffe782198e8de3.jpeg"},{"id":108406316,"identity":"f7aa708a-ddb3-4300-a406-c395a8efc479","added_by":"auto","created_at":"2026-05-04 09:41:58","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":99210,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRelationship between semaglutide maximum dose and maximum weight loss through the 2-year landmark date.\u003c/strong\u003e \u003cstrong\u003e(A) \u003c/strong\u003eStudy design highlighting the 0-2 year landmark period where the current analysis is focused. \u003cstrong\u003e(B)\u003c/strong\u003eBoxplots display the distribution of maximum percent body-weight change achieved before the 2-year landmark according to the maximum semaglutide dose reached within the landmark window. Negative values indicate weight loss. Center lines indicate medians, boxes indicate interquartile ranges, whiskers indicate the spread of the data, and black dots indicate group means. The dotted line shows the fitted linear trend between dose and weight change, with the corresponding regression equation shown on the plot. Sample sizes are indicated below each dose category. Baseline demographic and clinical characteristics of patients stratified by maximum semaglutide dose are summarized in \u003cstrong\u003eTable S2\u003c/strong\u003e.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/ed2ae402382534ffdd5ea95b.jpg"},{"id":108493147,"identity":"17cfde52-5946-4cf2-a388-d633a520095b","added_by":"auto","created_at":"2026-05-05 09:59:30","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":701443,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHigher maximum semaglutide dose was associated with lower post-landmark cardiovascular risk.\u003c/strong\u003e Patients were grouped by the maximum semaglutide dose reached by the 2-year landmark as low dose (0.25–1.0 mg; \u003cem\u003en\u003c/em\u003e = 8,725) or high dose (≥1.7 mg; \u003cem\u003en\u003c/em\u003e = 3,794). Left-column panels show cumulative event curves after the landmark and right-column panels show corresponding post-landmark event rates for \u003cstrong\u003e(A-B)\u003c/strong\u003e all-cause mortality, \u003cstrong\u003e(C-D) \u003c/strong\u003ecomposite cardiovascular events, \u003cstrong\u003e(E-F) \u003c/strong\u003eincident cerebrovascular disease, \u003cstrong\u003e(G-H) \u003c/strong\u003eheart failure, and \u003cstrong\u003e(I-J) \u003c/strong\u003eincident valvular/rheumatic heart disease. Relative risks and nominal \u003cem\u003eP\u003c/em\u003e values are shown on the bar plots. Higher dose semaglutide was associated with lower risk of all-cause mortality, composite cardiovascular events, incident cerebrovascular disease, incident heart failure, and valvular/rheumatic heart disease.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/c235ce236619b1008f6134de.jpeg"},{"id":108492567,"identity":"693e8753-8b92-4384-bab8-bea60a78554e","added_by":"auto","created_at":"2026-05-05 09:58:04","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":996095,"visible":true,"origin":"","legend":"\u003cp\u003eSemaglutide 2-year max weight-loss landmark analysis of HbA1c and blood pressure in the all cardiovascular conditions subgroup. Patients were categorized according to the maximum percentage reduction in body weight achieved before the 2-year landmark. Panels A and B show post-landmark HbA1c and change in HbA1c from patient-specific baseline, respectively, across weight-loss categories. Panels C and D show post-landmark systolic blood pressure and change from patient-specific baseline, and panels E and F show post-landmark diastolic blood pressure and change from patient-specific baseline. Bars indicate group means with standard error bars, dotted gray lines indicate fitted trends, solid black lines connect observed group means, and one-way ANOVA P values are shown.\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/2183922e8c33c185dffadf63.jpeg"},{"id":108406319,"identity":"cd7c849f-eff7-4785-b01a-d53a0b3b145e","added_by":"auto","created_at":"2026-05-04 09:41:58","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":913665,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSemaglutide 2-year max weight-loss landmark analysis in cardiovascular conditions subgroup. \u003c/strong\u003ePatients were categorized according to the maximum percentage reduction in body weight achieved before the 2-year landmark as \u0026lt;5% (\u003cem\u003en\u003c/em\u003e = 4,050), 5–10% (\u003cem\u003en\u003c/em\u003e = 2,805), 10–15% (\u003cem\u003en\u003c/em\u003e = 1,943), 15–20% (\u003cem\u003en\u003c/em\u003e = 1,223), 20–25% (\u003cem\u003en\u003c/em\u003e = 707), or \u0026gt;25% (\u003cem\u003en\u003c/em\u003e = 884), and cardiovascular outcomes were assessed only after the landmark among patients event-free through that timepoint. Left-column panels show cumulative post-landmark event probability over follow-up for all-cause mortality (A), composite cardiovascular events (C), incident cerebrovascular disease (E), incident heart failure (G), and incident valvular/rheumatic heart disease (I). Right-column panels show the corresponding post-landmark event rates across weight-loss categories for all-cause mortality (B), composite cardiovascular events (D), incident cerebrovascular disease (F), incident heart failure (H), and incident valvular/rheumatic heart disease (J), with global \u003cem\u003eP\u003c/em\u003e values indicated. In contrast to the dose-defined analysis, achieved weight-loss strata did not show a consistent monotonic association with subsequent cardiovascular risk, and the representative global comparisons shown were not significant.\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/6a7a391188095893c3a019e3.jpeg"},{"id":108406324,"identity":"88e8a5dc-edce-4e26-8795-473ddca8de2c","added_by":"auto","created_at":"2026-05-04 09:41:59","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":189238,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eReduction in semaglutide prescriptions in the two years following the initial treatment period, stratified by (A) maximum dose achieved \u003c/strong\u003eand \u003cstrong\u003e(B) maximum weight loss attained \u003c/strong\u003eBars represent the median per-patient difference in prescription count between the 0–2 year and 2–4-year windows; error bars denote the interquartile range. Patients achieving higher doses or greater weight loss show larger declines in prescription fills, consistent with treatment tapering or discontinuation following peak therapeutic response.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/1bdb4c49a2945af923f46a35.png"},{"id":108493132,"identity":"55a18c8d-42dc-4f11-98ac-6d979d9e76e3","added_by":"auto","created_at":"2026-05-05 09:59:28","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":893125,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eConcurrent relationship between achieved weight loss and cardiovascular outcomes during semaglutide treatment. \u003c/strong\u003ePatients with at least one cardiovascular condition were categorized by the maximum percentage reduction in body weight achieved within a 2-year observation window as \u0026lt;5% (n=15,201), 5–10% (n=8,115), 10–15% (n=4,829), 15–20% (n=2,689), 20–25% (n=1,304), or \u0026gt;25% (n=1,461). For each patient, outcomes were tracked from the date they achieved their maximum weight loss, with follow-up extending up to 24 months thereafter. Left-column panels show cumulative event probability and right-column panels show the corresponding 24-month event rates for all-cause mortality (A, B), composite cardiovascular events (C, D), incident cerebrovascular disease (E, F), incident heart failure (G, H), and incident valvular/rheumatic heart disease (I, J). Error bars represent 95% confidence intervals derived from Greenwood's formula; global log-rank P values are shown. In contrast to the dose-defined analysis, greater weight loss was associated with incrementally higher cardiovascular event rates across most outcomes — a pattern consistent with reverse causation, whereby patients with greater underlying cardiometabolic burden tend to achieve more pronounced weight loss on semaglutide during the concurrent observation window.\u003c/p\u003e","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/753e545dd57a326b8d9e9d3a.jpeg"},{"id":108493202,"identity":"c48cd300-9060-4adc-a3c5-98f0615e52c1","added_by":"auto","created_at":"2026-05-05 09:59:37","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1547859,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSemaglutide vs comparator anti-diabetic drugs' absolute 2-year event risk differences across baseline cardiovascular burden subgroups. \u003c/strong\u003e(A) Study design schematic highlighting the 0–2-year landmark period. (B) Semaglutide vs Metformin; (C) Semaglutide vs DPP-4 inhibitors; (D) Semaglutide vs SGLT-2 inhibitors. In B–D, heatmaps show absolute 2-year event risk differences across baseline cardiovascular burden subgroups. Rows indicate baseline cardiovascular burden subgroups, whereas columns indicate all-cause mortality, composite cardiovascular endpoint, and incident cardiovascular outcomes. Cell values represent the difference in approximate 2-year event risk between semaglutide and the respective comparator, expressed in percentage points (pp). Asterisks denote statistical significance based on log-rank tests using Fisher's method (* p\u0026lt;0.05, ** p\u0026lt;0.01, *** p\u0026lt;0.001). Results for additional cardiovascular outcomes are shown in \u003cstrong\u003eFig. S4\u003c/strong\u003e. \u003cstrong\u003eTable S6\u003c/strong\u003e lists the constituent medications within the DPP-4 inhibitors and SGLT2 inhibitors comparator groups. \u003cstrong\u003eTable S7\u003c/strong\u003eprovides the full definitions of baseline cardiovascular burden groups. \u003cstrong\u003eTable S9\u003c/strong\u003e provides the component code lists used for the composite cardiovascular endpoint.\u003c/p\u003e","description":"","filename":"floatimage9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/70761a1c4bcfafd68b352b31.jpeg"},{"id":108406322,"identity":"ea24573d-3da2-44c7-be87-9adca8ac7bfb","added_by":"auto","created_at":"2026-05-04 09:41:58","extension":"jpg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":111912,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWhole-body GLP1R receptor geography distinguishes prevalence-weighted engagement from absolute target burden. A. \u003c/strong\u003eSchematic overview of relationship between semaglutide dosage, weight loss and cardiovascular benefits\u003cstrong\u003e B. \u003c/strong\u003e\u003cem\u003eSemaglutide engagement across cardiac cell types stratified by GLP1R Engagement Potential (GEP; expression × positive-cell fraction). \u003c/em\u003eSchematic illustrating a hypothesis-generating model in which increasing semaglutide exposure may broaden engagement from high-GEP cardiac cell types at lower doses to additional medium- and low-to-mid-GEP cell types at higher doses, contingent on the presence of downstream signaling and contextual gene programs. GEP is used as a prevalence-weighted proxy for receptor availability across heart cell types. This model is intended to motivate interrogation of dose-dependent cardiac target engagement and does not represent a validated pharmacokinetic–pharmacodynamic relationship. \u003cstrong\u003e\u0026nbsp;\u003c/strong\u003eIn addition to GEP which is the primary metric, the other metric computed is the Absolute Target Load (ATL; expression × number of GLP1R-positive cells). Tissues are ordered by descending GEP. Pancreas shows the highest prevalence-weighted GLP1R signal, whereas heart shows the largest transcript-weighted GLP1R-positive reservoir. \u003cstrong\u003eC, \u003c/strong\u003eHeart-focused single-cell engagement landscape. Each point represents a cell type, positioned by the percentage of cells expressing GLP1R on the x axis and GLP1R expression on the y axis; bubble area is proportional to ATL. Cardiomyocyte-related populations, including atrial cardiomyocytes, cardiac muscle cells, muscle cells and contractile cells, combine modest expression with appreciable prevalence, whereas Schwann cells show higher expression intensity but lower prevalence. Also shown alongside is the Pancreas-focused single-cell engagement landscape. Exocrine epithelial and acinar populations show both high prevalence and substantial expression, consistent with the dominance of pancreas in aggregate GEP. In b and c, heart or pancreas are highlighted, and all other tissues are shown in grayscale. Dashed lines indicate cohort-wide median values for percentage of cells expressing GLP1R and GLP1R expression.\u003c/p\u003e","description":"","filename":"9.jpg","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/1e83de44fa1b090f6611fa91.jpg"},{"id":108804522,"identity":"bc3e127a-b7a1-4d86-9caf-95dd202f23ba","added_by":"auto","created_at":"2026-05-08 15:21:15","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6567398,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/0cc0ad77-a29b-4e88-9c52-6fc5108cea7e.pdf"},{"id":108406315,"identity":"6b3b55e5-af75-4cd5-9774-1e66834f14c4","added_by":"auto","created_at":"2026-05-04 09:41:58","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":2399637,"visible":true,"origin":"","legend":"","description":"","filename":"Soundararajansupplement.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9407045/v1/9d4e0d077471864aefe5d085.pdf"}],"financialInterests":"Competing interest reported. The authors are employees of nference, inc., which conducts research collaborations with various biopharmaceutical companies whose therapeutic products are included in this study. None of these companies, nor any other nference collaborator, funded, supported, or had any role in the independent study design, data acquisition, analysis, interpretation, manuscript preparation, or the decision to submit this work for publication. All analyses were conducted by the authors using de-identified electronic health record data. The authors declare no additional competing interests.","formattedTitle":"Semaglutide cardiovascular outcomes align more closely with attained dose than achieved weight loss","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCardiovascular disease is among the most consequential afterlives of excess adiposity, with myocardial injury, vascular dysfunction, and heart failure causing premature death across millions of patients worldwide\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. In this landscape, semaglutide has emerged not merely as a drug that lowers body weight, but as a therapy that appears to reshape the clinical terrain of cardiometabolic disease itself. Randomized trials have shown that it can produce substantial and sustained weight reduction while also lowering major cardiovascular event risk in high-risk populations with type 2 diabetes and/orobesity\u003csup\u003e2\u0026ndash;4\u003c/sup\u003e. Yet, despite its widening therapeutic reach, semaglutide is still most often understood by patients through the lens of a single visible metric: the magnitude of weight lost.\u003c/p\u003e \u003cp\u003eIt is increasingly appreciated by physicians that this lens may be too narrow to capture the full biology of the drug. An expanding body of human evidence now suggests that semaglutide influences domains that extend beyond weight loss alone (\u003cb\u003eTable \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e)\u003csup\u003e\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e, including major adverse cardiovascular events (SELECT), symptoms and physical limitations in obesity-related heart failure with preserved ejection fraction (STEP-HFpEF), kidney outcomes in type 2 diabetes with chronic kidney disease (FLOW), cardiovascular outcomes with oral semaglutide (SOUL), and walking capacity in symptomatic peripheral artery disease (STRIDE)\u003csup\u003e\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. The recent trial of high-dose semaglutide 7.2 mg provides an additional dose\u0026ndash;response benchmark beyond the standard 0.25\u0026ndash;2.4 mg range (STEP UP)\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Together, these findings raise the possibility that body-weight reduction, although important, is only one visible manifestation of a broader therapeutic biology.\u003c/p\u003e \u003cp\u003eThis raises a question that is at once practical and mechanistic. In routine care, are dose, weight loss and cardiovascular benefit aligned along a single axis, or does semaglutide exert partially separable cardiovascular effects that are not fully captured by weight-loss magnitude? This question is especially relevant now that both treatment guidelines and commercial development strategies are increasingly organized around organ outcomes and comorbidity reduction, not weight loss alone\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e,\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Indeed, weight loss is a composite physiological readout, shaped not only by dose, but by treatment persistence, tolerability, metabolic context, baseline illness and individual biological variability. It is therefore possible that achieved weight reduction, although clinically salient, may serve as an incomplete surrogate for the signaling pathways most relevant to cardiovascular outcomes. Resolving this distinction is important not only for interpreting semaglutide response, but also for how cardiometabolic therapies are compared, optimized, and advanced beyond obesity alone.\u003c/p\u003e \u003cp\u003eHere, we addressed this question using deidentified longitudinal clinical data from the nSights Federated EHR Network, contextualized through a whole-body spatial intelligence framework\u003csup\u003e\u003cspan additionalcitationids=\"CR12 CR13\" citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e. As summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, the study couples a landmark clinical design and cohort-definition strategy with a broader conceptual framework linking therapeutic exposure, systems pharmacology, and diagnostics. Our findings position semaglutide as a cardiometabolic therapy whose clinical value may be organized along axes not captured by body-weight reduction alone and may be overlooked by conventional endpoints.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eHigher attained semaglutide dose was associated with greater weight loss during the landmark period\u003c/h2\u003e \u003cp\u003eBaseline demographic and clinical characteristics of patients stratified by maximum semaglutide dose are summarized in \u003cb\u003eTable S2\u003c/b\u003e. Across the first two years after semaglutide initiation, attained dose and achieved weight loss were closely correlated (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Mean maximum weight loss increased from approximately 8% at 0.25 mg to 15% at 2.4 mg, corresponding to an estimated 3.15% greater maximum weight loss per 1 mg increase in attained dose (Pearson R\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.97, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe same gradient emerged when patients were stratified by the maximum body-weight reduction achieved during the landmark period. Demographically, when comparing the \u0026lt;\u0026thinsp;5% weight-loss and \u0026gt;\u0026thinsp;25% weight-loss strata, mean age decreased from 58.2 to 56.2 years, female representation increased from 61.8% to 80.9%, baseline BMI rose modestly from 35.2 to 36.9 kg/m\u003csup\u003e2\u003c/sup\u003e, baseline weight rose from 103.1 kg to 105.6 kg, and baseline HbA1c fell from 6.6 to 6.2. Regarding the association between dose and weight loss, the proportion reaching the highest dose category increased from 20.9% to 49.6%, and median semaglutide prescription counts during the first 2 years increased from 1 [IQR 1 to 2] to 5 [IQR 2 to 11] (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Weight loss, therefore, seemed to reflect a combination of cumulative treatment exposure and baseline phenotype.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003eBaseline and treatment characteristics across weight-loss categories among semaglutide-treated patients.\u003c/b\u003e Weight-loss categories were defined according to the maximum percentage reduction in body weight achieved within each landmark window. Baseline BMI and baseline weight are presented as mean (SD); dose intensity is summarized as the proportion reaching the highest dose category and prescription counts are summarized as median [interquartile range].\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWeight Loss Category\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber of unique patients\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAge,\u003c/p\u003e \u003cp\u003eMean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFemale, %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eBaseline BMI, mean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eBaseline Weight, kg mean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eT2DM Prevalence at Baseline, %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eBaseline HBA1C, Mean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eMax Dose,\u003c/p\u003e \u003cp\u003e2-year n (%)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003ePrescription, 2-year median (IQR)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;5%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4,050\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e58.2 (12.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e61.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35.2 (6.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e103.1 (25.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e28.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6.6 (1.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e847 (20.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e1 [1, 2]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u0026ndash;10%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2,805\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e59.8 (12.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e62.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35.4 (6.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e102.6 (23.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e30.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6.5 (1.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e763 (27.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e2 [1, 5]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10\u0026ndash;15%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,943\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e60.0 (11.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e66.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35.4 (5.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e101.5 (22.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e30.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6.5 (1.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e655 (33.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e2 [1, 7]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e15\u0026ndash;20%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1,223\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e58.4 (12.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e73.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35.5 (5.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e101.0 (21.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e27.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6.3 (1.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e504 (41.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e4 [1, 9]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e20\u0026ndash;25%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e707\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e57.9 (11.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e76.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35.5 (5.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e100.8 (21.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e25.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6.2 (1.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e325 (46.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e5 [1, 10]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;25%\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e884\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e56.2 (12.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e80.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e36.9 (5.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e105.6 (23.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e25.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6.2 (1.4)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e438 (49.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e5 [2, 11]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eHigher attained dose during the landmark period was associated with lower post-landmark cardiovascular risk\u003c/h3\u003e\n\u003cp\u003eWhen the post-landmark period was examined through the lens of dose attained during the first two years, patients who reached high-dose semaglutide by the landmark (\u0026ge;\u0026thinsp;1.7 mg; n\u0026thinsp;=\u0026thinsp;3,794) had lower subsequent risk in the 2\u0026ndash;4 years (\u0026ldquo;post-landmark\u0026rdquo;) period than those who remained in the low-dose range (0.25 to 1.0 mg; n\u0026thinsp;=\u0026thinsp;8,725) across several major outcomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). High attained dose was associated with lower: all-cause mortality (RR\u0026thinsp;=\u0026thinsp;0.42, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), risk of the composite cardiovascular endpoint (RR\u0026thinsp;=\u0026thinsp;0.51, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), incident cerebrovascular disease (RR\u0026thinsp;=\u0026thinsp;0.50, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), incident heart failure (RR\u0026thinsp;=\u0026thinsp;0.55, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and incident valvular or rheumatic heart disease (RR\u0026thinsp;=\u0026thinsp;0.71, P\u0026thinsp;=\u0026thinsp;0.025) (see \u003cb\u003eMethods\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFor all-cause mortality, the composite cardiovascular endpoint, cerebrovascular disease and heart failure, time-to-event analyses showed divergence that emerged within the first 5\u0026ndash;10 months post-landmark and persisted through 24 months of follow-up (log-rank p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 for all analyses) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Cumulative event curves did not significantly differ for incident valvular or rheumatic heart disease. Ischemic heart disease, hypertension, arrhythmias or conduction disorders, peripheral vascular disease or atherosclerosis, cardiomyopathy, aortic disease and venous thromboembolism did not show clear dose-stratified differences in the post-landmark 2\u0026ndash;4-year period (\u003cb\u003eFig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e). Analysis anchored on the time of first attainment of maximum dose yielded directionally concordant results, again showing lower risk across the same cardiovascular outcomes in the higher-dose group \u003cb\u003e(Fig. S2\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eDose attainment also correlated with treatment intensity beyond the landmark. During the first two years, median semaglutide prescription counts were 5 [IQR 1 to 9] in the high-dose group versus 1 [IQR 1 to 2] in the low-dose group (\u003cb\u003eTable S3\u003c/b\u003e). In the post-landmark period (2\u0026ndash;4 years upon initiating semaglutide), prescribing attenuated in both groups, but remained modestly higher in the high-dose group (median 1 [IQR 0 to 2] versus 0 [IQR 0 to 2]). However, larger negative pre-to-post-landmark changes in prescription counts were observed in the high-dose group (\u0026minus;\u0026thinsp;4 [IQR\u0026thinsp;\u0026minus;\u0026thinsp;9 to \u0026minus;\u0026thinsp;1] relative to the low-dose group\u0026thinsp;\u0026minus;\u0026thinsp;1 [IQR\u0026thinsp;\u0026minus;\u0026thinsp;2 to \u0026minus;\u0026thinsp;1]).\u003c/p\u003e \u003cp\u003e \u003cb\u003eHigher weight loss during the landmark period was associated with post-landmark glycemic and systolic blood pressure control, but not later cardiovascular risk\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWhen the same shared landmark was viewed through the lens of weight change rather than attained dose, the metabolic signals were as expected, but the cardiovascular signal was not.\u003c/p\u003e \u003cp\u003eGreater maximum weight loss during the first two years was accompanied by progressively more favorable metabolic-surrogate measures at and after the landmark. Post-landmark HbA1c fell from approximately 6.4% in the \u0026lt;\u0026thinsp;5% weight-loss group to 5.6% in the \u0026gt;\u0026thinsp;25% group, whereas patient-specific HbA1c change from the pre-treatment baseline deepened from \u0026minus;\u0026thinsp;0.2% to \u0026minus;\u0026thinsp;0.8% (one-way ANOVA P\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ea-b). Post-landmark systolic blood pressure similarly declined from about 133 mmHg to 126 mmHg, with change from baseline shifting from roughly\u0026thinsp;+\u0026thinsp;2.0 to \u0026minus;\u0026thinsp;3.5 mmHg (both P\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ec-d). Post-landmark diastolic blood pressure declined more modestly, from about 78.0 to 75.4 mmHg, with progressively more negative change from baseline across strata (P\u0026thinsp;=\u0026thinsp;0.002 for change from baseline). By conventional metabolic readouts, larger weight loss behaved exactly as a stronger systemic response should.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eUsing the same 0 to 2-year weight-loss strata and the same 2-year landmark, post-landmark all-cause mortality, the composite cardiovascular endpoint, incident cerebrovascular disease, incident heart failure and incident valvular or rheumatic heart disease did not show a consistent monotonic gradient (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). Global comparisons were not significant for all-cause mortality (P\u0026thinsp;=\u0026thinsp;0.144), the composite cardiovascular endpoint (P\u0026thinsp;=\u0026thinsp;0.547), incident cerebrovascular disease (P\u0026thinsp;=\u0026thinsp;0.249), incident heart failure (P\u0026thinsp;=\u0026thinsp;0.887) or incident valvular or rheumatic heart disease (P\u0026thinsp;=\u0026thinsp;0.431). Among the additional cardiovascular outcomes, only incident peripheral vascular disease or atherosclerosis reached nominal significance across weight-loss strata (global P\u0026thinsp;=\u0026thinsp;0.033), whereas incident hypertension showed only a suggestive downward trend with greater weight loss (global P\u0026thinsp;=\u0026thinsp;0.107); the remaining outcomes were not significant (ischemic heart disease, arrhythmias, cardiomyopathy, aortic disease, and venous thromboembolism; \u003cb\u003eFig. S3\u003c/b\u003e). Taken together, these findings suggest that greater weight reduction during the 0\u0026ndash;2 years landmark period did not necessarily translate into proportionally better cardiovascular outcomes in the 2\u0026ndash;4 years post-landmark.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003ePart of the explanation may lie in the fact that weight loss was a moving summary of the first 2 years, whereas drug exposure fell sharply thereafter. When stratified by maximum dose achieved, semaglutide prescribing declined more sharply after the landmark among patients who had reached higher doses than among those who remained in the low-dose group, indicating that more intensive treatment during the first 2 years (landmark period) did not necessarily translate into sustained exposure in the following post-landmark period (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). A similar pattern was observed across weight-loss strata, with semaglutide prescriptions dropping markedly after the landmark (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb). Median post-landmark prescription counts fell to 0 or 1 in every group, with the steepest declines in the higher-response strata: \u0026minus;4 [IQR\u0026thinsp;\u0026minus;\u0026thinsp;8 to \u0026minus;\u0026thinsp;1] in the 15 to 20% group, \u0026minus;\u0026thinsp;5 [IQR\u0026thinsp;\u0026minus;\u0026thinsp;9 to \u0026minus;\u0026thinsp;1] in the 20 to 25% group, and \u0026minus;\u0026thinsp;4 [IQR\u0026thinsp;\u0026minus;\u0026thinsp;8 to \u0026minus;\u0026thinsp;1] in the \u0026gt;\u0026thinsp;25% group (\u003cb\u003eTable S4\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThus, maximum weight loss accrued during the landmark period did not correspond to sustained semaglutide prescriptions thereafter. Across all weight-loss strata, median prescription counts after maximum weight loss attainment were only 0 or 1, indicating that the subsequent outcome window was generally characterized by sparse semaglutide exposure (\u003cb\u003eTable S4\u003c/b\u003e). Consistent with this context, a concurrent analysis anchored at the time of maximum weight loss showed that greater achieved weight loss was not associated with lower cardiovascular risk over the next 24 months (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). Rather, all-cause mortality and the composite cardiovascular endpoint showed modest upward separation across increasing weight-loss categories, with the highest event probabilities generally observed among patients with the largest weight reductions (P\u0026thinsp;\u0026gt;\u0026thinsp;0.05, Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ea-d). Incident cerebrovascular disease followed a similar directional trend, whereas incident heart failure and incident valvular or rheumatic heart disease showed no consistent gradient across weight-loss strata (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ee-j). Together, these findings indicate that achieved weight loss alone did not recapitulate the favorable cardiovascular pattern observed across attained semaglutide dose tiers, particularly in a setting where semaglutide prescribing after maximum weight loss was minimal.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eMatched comparator analyses placed the semaglutide cardiovascular signal in the broader clinical context of antidiabetic medicines\u003c/b\u003e \u003c/p\u003e \u003cp\u003eSemaglutide carried a broader favorable cardiovascular signal in matched comparator analyses. In 1:1 propensity-score-matched comparisons against metformin, the pooled \u003cem\u003eall cardiovascular conditions\u003c/em\u003e subgroup contained 47,199 patients per arm and was closely balanced for age at index (59.4 versus 59.9 years), female representation (62.8% versus 61.3%), baseline BMI (35.6 versus 35.3 kg/m\u003csup\u003e2\u003c/sup\u003e) and baseline type 2 diabetes prevalence (25.8% versus 26.0%) (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Across individual cardiovascular subgroups, mean age ranged from approximately 59 to 67 years, baseline BMI from 33.8 to 36.3 kg/m\u003csup\u003e2\u003c/sup\u003e, and baseline type 2 diabetes prevalence from 25.4% in aortic disease to 50.8% in heart failure, indicating broadly balanced matching across a clinically heterogeneous population.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline characteristics of propensity score\u0026ndash;matched semaglutide and metformin cohorts across cardiovascular disease subgroups. For each baseline cardiovascular subgroup, patients initiating semaglutide were matched 1:1 to metformin-treated patients using propensity scores estimated from age, sex, baseline BMI and T2DM status. Matched cohort sizes, age at index date (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD), proportion female, baseline BMI (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD) and proportion with type 2 diabetes mellitus (T2DM) at baseline are shown.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"10\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBaseline CV Group\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMatched Number of unique patients\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSemaglutide age at index, mean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMetformin age at index, mean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSemaglutide female, %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMetformin female, %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eSemaglutide baseline BMI, mean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eMetformin baseline BMI, mean (SD)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eSemaglutide T2DM, %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c10\"\u003e \u003cp\u003eMetformin T2DM, %\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAll CV conditions\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e47199\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e59.4 (12.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e59.9 (13.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e62.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e61.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e35.6 (6.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e35.3 (6.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e25.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e26\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAortic disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e899\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e67.0 (10.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e67.5 (10.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e35.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e31.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e34.5 (5.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e34.3 (5.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e25.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e27.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eArrhythmia / Conduction disorders\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9590\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e63.6 (13.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e64.2 (14.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e56.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e54.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e35.1 (5.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e34.8 (6.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e29.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e30.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCardiomyopathy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1494\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e62.0 (13.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e62.4 (13.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e48.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e35.3 (5.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e34.7 (6.2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e35.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e35.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCerebrovascular disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2821\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e65.8 (11.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e66.1 (12.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e61.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e60.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e34.2 (5.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e34.2 (5.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e39.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e40.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeart failure\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3777\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e65.9 (12.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e66.7 (12.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e53.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e48.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e36.2 (6.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e35.4 (6.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e50.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e50.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHypertension\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e40773\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e59.8 (12.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e60.4 (13.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e60.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e35.8 (5.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e35.5 (6.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e27.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e28.3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIschemic heart disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e9203\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e66.4 (10.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e66.6 (10.5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e44.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e41.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e34.6 (5.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e34.2 (5.7)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e38.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e37.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePeripheral / Vascular disease atherosclerosis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3425\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e66.8 (11.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e67.2 (11.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e59.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e58.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e34.5 (6.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e33.8 (5.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e43.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e42.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eValvular / Rheumatic heart disease\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3082\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e65.2 (13.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e65.7 (13.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e60.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e58.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e34.5 (5.8)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e34.2 (6.0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e34.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e35.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVenous thromboembolism\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2750\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e58.6 (13.6)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e59.4 (14.9)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e64.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e61.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e36.3 (6.1)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e35.9 (6.3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e29.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c10\"\u003e \u003cp\u003e30.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eAnalyzed from the time of treatment initiation (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ea), semaglutide was associated with lower approximate 2-year event risk than metformin for all-cause mortality (1.7% versus 3.4%; P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), the composite cardiovascular endpoint (9.7% versus 13.3%; P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), incident ischemic heart disease (8.0% versus 8.6%; P\u0026thinsp;=\u0026thinsp;0.002), incident cerebrovascular disease (3.5% versus 4.8%; P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), incident hypertension (23.8% versus 40.5%; P\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and incident arrhythmias or conduction disorders (8.5% versus 10.4%; P\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eb, \u003cb\u003eFig. S4a\u003c/b\u003e). Broadly favorable patterns were also observed when comparing semaglutide to dipeptidyl peptidase-4 (DPP-4) inhibitors (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ec, \u003cb\u003eFig. S4b\u003c/b\u003e) or sodium-glucose cotransporter-2 (SGLT2) inhibitors (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ed, \u003cb\u003eFig. S4c\u003c/b\u003e). These comparator analyses support an overall favorable cardiovascular association for semaglutide.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eWhole-body GLP1R geography provides biologic plausibility for organ-directed effects\u003c/h3\u003e\n\u003cp\u003eThe landmark analysis raised a biological question: if differences in subsequent cardiovascular outcomes among semaglutide-treated patients were not fully explained by the magnitude of weight loss achieved during the first 2 years, what aspects of GLP1R biology might account for this apparent organ-directed effect? We hypothesized that cell-type-specific receptor prevalence and downstream signaling readiness may be important determinants, and to test this idea, we generated a schematic model of \u003cem\u003eGLP1R engagement potential (GEP)\u003c/em\u003e from a whole-body single-cell atlas (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003ea). This model showed that the pancreas had the highest prevalence-weighted GEP (1.944%), but the heart emerged as the next most prominent tissue by GEP (0.729%) and, notably, as the largest aggregate GLP1R+ reservoir by \u003cem\u003eabsolute target load (ATL\u003c/em\u003e; 19,913), marginally exceeding the brain (19,217) and approximately doubling the pancreas (9,967) (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003eb, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Thus, while the pancreas remained the dominant prevalence-weighted site of GLP1R signal, the heart stood out as a potential extra-pancreatic target organ.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSingle Cell RNA atlas of GLP1R Engagement Potential (GEP) and Absolute Target Load (ATL), with tissues sorted by decreasing GEP%.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTissue\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCell Count\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eExpression (CP10K)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e% Expressing Cells\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003ePositive Fraction\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCells Expr. GLP1R\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eGEP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eGEP %\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eATL\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003epancreas\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e512856\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5537\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.01944\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e1.944\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e9966.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eheart\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2730947\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.405\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00405\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e11063\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.00729\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.729\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e19913.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eaxilla\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e84584\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.346\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00346\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e293\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0059858\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.59858\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e506.89\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ehuman breast milk\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e47514\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.183\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00183\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0031293\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.31293\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e148.77\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eabdomen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e429748\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.092\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00092\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e397\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0015456\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.15456\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e666.96\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003espinal cord\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e44639\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.081\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00081\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0014499\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.14499\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e64.44\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003emucosa\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e109366\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.069\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00069\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0014145\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.14145\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e155.8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eesophagus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e155431\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.059\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00059\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e92\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.001298\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.1298\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e202.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eembryo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e277234\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.063\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00063\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e176\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.00126\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.126\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e352\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003echest wall\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e19459\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.057\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00057\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0009975\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.09975\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e19.25\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eprostate gland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e207838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0007995\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.07995\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e166.05\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003estomach\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e303597\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.041\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00041\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e125\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0007913\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.07913\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e241.25\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ebrain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e26689529\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e10796\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0712\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e19216.88\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eadrenal gland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e244608\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000692\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0692\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e169.54\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eadipose tissue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e448846\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.037\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00037\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e164\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0006919\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.06919\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e306.68\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elymph node\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1731461\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00033\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e576\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0005709\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.05709\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e996.48\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003emusculature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e508632\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.028\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00028\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e144\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0005544\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.05544\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e285.12\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ekidney\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1245614\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.028\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00028\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e352\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000518\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0518\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e651.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eexocrine gland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e193695\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0004444\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.04444\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e84.84\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eliver\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1444734\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e346\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0004152\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.04152\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e598.58\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ebreast\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3100226\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00022\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e696\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0004048\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.04048\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1280.64\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eintestine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e322530\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.00037\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.037\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e118.4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elung\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4526576\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e909\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000366\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0366\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1663.47\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eeye\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e7700797\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1548\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.00035\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.035\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2709\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003etendon of semitendinosus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10533\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00019\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0003363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.03363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e3.54\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eneck\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e19097\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.016\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00016\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000264\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0264\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e4.95\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ecolon\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e812980\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00012\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0002196\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.02196\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e181.17\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ecortex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e149400\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0002145\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.02145\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e31.35\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003echest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e15413\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0002015\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.02015\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e3.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003etongue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e38749\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00013\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0001911\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.01911\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e7.35\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003etestis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e18528\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.00011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0001738\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.01738\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e3.16\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003esmall intestine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1191162\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.71\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e119\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000171\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0171\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e203.49\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ecerebrospinal fluid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e72565\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.008\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0001408\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.01408\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e10.56\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003espleen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e537601\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.008\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0001352\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.01352\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e70.98\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eskin of body\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e817861\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.91\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e7.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0001337\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.01337\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e108.87\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eesophagogastric junction\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e11883\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.008\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.0001232\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.01232\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1.54\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eendocrine gland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e654345\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.006\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0.000102\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0102\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e64.6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eblood\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e25136830\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.67\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1330\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e8.35E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00835\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2221.1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eureter\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e41757\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e7.80E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0078\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e3.12\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eovary\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e346569\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.73\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e6.92E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00692\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e24.22\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ebone marrow\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e727397\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.65\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e6.60E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.0066\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e49.5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eurinary bladder\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e36100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.51\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e4.53E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00453\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1.51\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003efallopian tube\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e238661\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3.96E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00396\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e7.92\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003elarge intestine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e422380\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e3.54E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00354\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e14.16\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epleural fluid\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e83186\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.92E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00292\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e2.92\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003euterus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e456208\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e2.05E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00205\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e8.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003evasculature\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e349181\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.69E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00169\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e6.76\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eplacenta\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e561073\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.68\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.68E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00168\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e11.76\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eomentum\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e213994\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.62\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.62E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00162\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e3.24\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ebladder organ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e68317\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.46E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00146\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1.46\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ehindlimb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e85565\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.00E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e1.46E-05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.00146\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c9\"\u003e \u003cp\u003e1.46\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eWithin the heart, the most prevalent GLP1R+ populations were cardiomyocytes, particularly atrial cardiomyocytes, followed by endocardial cells, pericytes, and cardiac blood vessel endothelial cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003ec; \u003cb\u003eTable S5\u003c/b\u003e). Schwann cells and specific cardiac immune-cell populations showed some of the highest per-cell expressions but were rare and ultra-rare, respectively. The GLP1R+ rare cardio-immune cells included CD4-positive alpha-beta T cells and mature natural killer T cells. Among cardiac myocytes, regular atrial cardiac myocytes showed GLP1R expression of 1.83 CP10K, with 992 expressing cells and 1.364% of cells positive, yielding a salience score of 2.492, whereas regular ventricular cardiac myocytes showed comparable per-cell expression intensity (1.72 CP10K) but substantially lower prevalence, with 926 expressing cells and only 0.382% of cells positive, yielding a lower salience score of 0.659. A second ventricular population, ventricular cardiac muscle cells, showed slightly higher per-cell expression (2.03 CP10K) but extremely sparse positivity (0.116%; salience 0.236). The dominant contribution in the pancreas arose from the exocrine compartment including acinar cells, which had the top-ranked GEP status across all analyzed organs of the human body (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003ed).\u003c/p\u003e \u003cp\u003eConsistent with these insights from single cell analysis, whole-body bulk RNA sequencing data shows that GLP1R expression was highest in the pancreas, followed next by the heart, and then by selected neuroendocrine tissues (\u003cb\u003eFig. S5\u003c/b\u003e). Specifically, heart muscle ranked among the highest non-pancreatic tissues for GLP1R expression, significantly exceeding most peripheral organs, including lung, liver, kidney, and adipose tissue. Within cardiac compartments, GLP1R expression showed meaningful signal in atrial tissue and, to a lesser extent, ventricular tissue, with distributions indicating heterogeneous but reproducible expression across samples. Atrial samples (n\u0026thinsp;=\u0026thinsp;432) and left ventricular samples (n\u0026thinsp;=\u0026thinsp;432, 550) showed median TPM was 1.01 in atrial tissues (Q1\u0026ndash;Q3, 0.56\u0026ndash;1.47) versus 0.44\u0026ndash;0.46 in left ventricular tissues (Q1\u0026ndash;Q3, 0.17\u0026ndash;0.87). Triangulating bulk RNA and single-cell RNA-seq, atrial tissue showed a stronger GLP1R signal than ventricular tissue, with the bulk distributions shifted upward in atrium across the full range of samples: median TPM was 1.01 in atrial tissues (Q1\u0026ndash;Q3, 0.56\u0026ndash;1.47) versus 0.44\u0026ndash;0.46 in left ventricular tissues (Q1\u0026ndash;Q3, 0.17\u0026ndash;0.87). Together, these data indicate that the stronger atrial GLP1R signal in bulk tissue is better explained by higher prevalence of GLP1R-positive cardiomyocytes in the atrium than by markedly higher per-cell expression intensity.\u003c/p\u003e \u003cp\u003eCollectively, these data establish that while the pancreas dominates GLP1R expression, the heart represents a prominent extra-pancreatic site of GLP1R transcription, supporting the plausibility of cardiac cells as biologically relevant targets for GLP1 receptor agonists which may contribute to their broader effects beyond body-weight reduction.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eSemaglutide appears to write on more than one physiologic axis\u003c/h2\u003e \u003cp\u003eThe most important lesson of these data is that semaglutide seems to leave at least two signatures on human physiology. One signature is immediately visible: higher attained dose tracked closely with greater weight loss across the first 2 years after initiation. The second is quieter, but potentially more consequential: later cardiovascular benefit aligned more closely with dose attained than with the magnitude of weight loss itself. If weight loss were the dominant mediator of cardiovascular protection in this setting, one would expect a graded decline in post-landmark events across progressively deeper weight-loss strata. We did not observe that pattern. Instead, conventional metabolic markers such as HbA1c and blood pressure improved monotonically with weight loss, whereas post-landmark cardiovascular outcomes did not. In this cohort, body weight change therefore behaved as an important pharmacodynamic readout, but not as a complete surrogate for the biology most relevant to later cardiovascular risk.\u003c/p\u003e \u003cp\u003eThe prescribing trajectory after the landmark sharpens that distinction. Weight loss was a cumulative summary of the first 2 years, whereas semaglutide prescribing attenuated substantially in the subsequent interval in which cardiovascular outcomes were assessed. In that setting, a patient's maximum weight loss may memorialize earlier exposure without faithfully representing the therapeutic state that followed. That temporal mismatch provides one explanation for why weight-loss magnitude tracked systemic metabolic improvement yet failed to organize later cardiovascular risk as cleanly as dose attainment did.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eCardiac GLP1R geography offers biologic plausibility for organ-directed effects\u003c/h2\u003e \u003cp\u003eOur atlas-based analyses offer a biologically plausible framework for this dissociation. The pancreas remained the dominant prevalence-weighted site of GLP1R signal, as expected for an incretin therapy. But the heart emerged immediately behind it by GLP1R engagement potential and, notably, as the largest transcript-weighted reservoir of GLP1R-positive cells (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003e, Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). Within that reservoir, signal was not confined to a single rare niche. It extended across cardiomyocyte and contractile populations, endothelial and endocardial compartments, stromal and perivascular states, and selected immune populations. These data do not demonstrate direct cardiac causality. They do, however, make it biologically plausible that the cardiovascular effects of semaglutide are written partly through organ-directed biology in the heart and vasculature, and not solely through the arithmetic of kilograms lost.\u003c/p\u003e \u003cp\u003eThat broader view is also consistent with the emerging clinical literature. Semaglutide has shown cardiovascular and heart-failure benefits across multiple contexts, including obesity with established cardiovascular disease, obesity-related HFpEF, chronic kidney disease with diabetes, oral semaglutide cardiovascular outcomes, and symptomatic peripheral artery disease\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e3\u003c/span\u003e–\u003cspan class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Our data suggest that these benefits may be better understood as the net result of several partially overlapping programs, including systemic metabolic change, vascular-interface effects, and myocardial or immunometabolic signaling. An extended mechanistic interpretation of the whole-body atlas, cardiac cell-state programs, and functional-readiness modules is provided in the Supplementary Discussion (\u003cb\u003eFigs. S5\u003c/b\u003e-\u003cb\u003eS6\u003c/b\u003e). Finally, differential expression of GLP1R in the heart also motivates future analyses of cardiac remodeling in semaglutide-treated patients through radiological diagnostic imaging data collected before and after semaglutide treatment initiation.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eImplications for endpoint selection, dose optimization and trial design\u003c/h3\u003e\n\u003cp\u003eThe matched comparator analyses place this within-drug observation into a broader therapeutic frame. From treatment initiation, semaglutide showed a favorable cardiovascular association relative to metformin, DPP-4 inhibitors, and SGLT2 inhibitors across several major endpoints. Taken together with the landmark analyses, the implication is not that weight loss is irrelevant. Rather, it is that cardiovascular medicine should resist reducing the value of semaglutide to a single visible phenotype. Weight loss may be the most conspicuous manifestation of drug activity, but it is not necessarily the most discriminating measure of organ benefit.\u003c/p\u003e \u003cp\u003eThis distinction matters for both clinical practice and drug development. In routine care, incretin therapies are often titrated and compared through the lens of body-weight reduction. Yet the outcomes that matter most to patients with cardiovascular disease are myocardial infarction, stroke, heart failure, disability, and death. If these data are confirmed prospectively, then dose attainment, exposure patterns and organ-specific endpoints should stand alongside weight loss, rather than behind it, in the optimization of semaglutide and in the design of next-generation trials. The broader opportunity is methodological as well: a whole-body spatial intelligence framework that brings together federated longitudinal EHR data, therapeutic exposure, physiologic biomarkers, and molecular atlases may help reveal therapeutic value axes that conventional obesity endpoints do not fully capture.\u003c/p\u003e\n\u003ch3\u003eLimitations and future directions\u003c/h3\u003e\n\u003cp\u003eThis study has important limitations. As an observational analysis of de-identified EHR data, it remains vulnerable to residual confounding, treatment-selection bias, unequal measurement density, unmeasured adherence, and misclassification of outcomes defined from diagnosis codes rather than adjudicated events. Dose was inferred from prescriptions rather than verified administration. The marked attenuation of prescribing after the landmark further complicates interpretation of exposure during the later risk window. The transcriptomic analyses provide biologic plausibility, not mechanistic proof: single-cell and bulk RNA measurements do not establish receptor occupancy, downstream pathway activation, or causal mediation in human cardiac tissue. The metrics GEP and ATL should be interpreted as atlas-derived pharmacologic heuristics rather than literal measures of in vivo receptor occupancy or absolute organ-level target burden. In particular, ATL is influenced by uneven cell recovery and differential susceptibility of cell types to dissociation, capture and sequencing in single-cell RNA-sequencing datasets, and therefore should not be treated as a direct estimate of absolute cellular proportions across tissues. Accordingly, these metrics are most useful for prioritizing candidate tissues and cell compartments for semaglutide responsiveness, rather than for making exact quantitative claims about organ-level pharmacologic exposure.\u003c/p\u003e \u003cp\u003eThe absence of a monotonic cardiovascular gradient across weight-loss strata should therefore not be read as evidence that large semaglutide-mediated weight loss is harmful. It should be read more carefully, and more usefully, as evidence that weight-loss magnitude alone does not fully explain the later cardiovascular signal observed here. Even with those caveats, a coherent picture emerges. Semaglutide appears to act not only as an agent of body-mass reduction, but as a therapy whose clinical signature extends into cardiovascular biology. One part of that signature is visible on the scale. Another may be written more quietly in the heart, the vessel wall, and the clinical arc of risk over time. The next generation of cardiometabolic trials will need to read both.\u003c/p\u003e"},{"header":"Methods","content":"\u003ch2\u003eData Source and Study Design\u003c/h2\u003e\u003cp\u003eWe conducted a retrospective cohort study using deidentified longitudinal electronic health record data from the nSights Federated EHR Network\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. All analyses were conducted on deidentified data.\u003c/p\u003e\u003ch2\u003eStudy Cohorts and Exposure Definitions\u003c/h2\u003e\u003cp\u003eFrom approximately 29\u0026nbsp;million patients, 505,874 had at least one semaglutide prescription record, of whom 269,390 initiated semaglutide between March 2018 and January 2024; 47,199 of these had baseline cardiovascular disease (see below). Medication exposure was ascertained from prescription or medication event records using the recorded event timestamp. The index date was defined as the date of the first qualifying prescription for the cohort-defining medication or medication class. Follow-up data were available through January 2026. To create incident-user cohorts with minimal treatment contamination, patients were excluded if they had exposure to metformin, DPP-4 inhibitors, or SGLT2 inhibitors during the 365 days before the index date or at any time after the index date. Constituent medications within DPP-4 and SGLT2 groups are listed in \u003cb\u003eTable S6\u003c/b\u003e. Other GLP-1RA medications (tirzepatide, liraglutide, dulaglutide, exenatide, albiglutide, lixisenatide) were used as exclusion exposures when defining the semaglutide cohort.\u003c/p\u003e\u003ch2\u003eDemographic and Clinical Covariates\u003c/h2\u003e\u003cp\u003eDemographic data included age, sex, and death date. Age at index was obtained from the index medication record. Baseline BMI was defined as the BMI measurement closest to the index date, recorded from 365 days before through 14 days after the index date. Baseline weight for the weight-loss analyses was defined separately as the weight measurement closest to the index date, recorded from 90 days before through 14 days after the index date.\u003c/p\u003e\u003cp\u003eBaseline type 2 diabetes was defined by the presence of at least three encounters carrying a type 2 diabetes code on three distinct dates at any time before the index date. The same logic was used to define prespecified baseline cardiovascular comorbidity groups. These groups were ischemic heart disease, cerebrovascular disease, hypertension, heart failure, arrhythmias or conduction disorders, peripheral vascular disease, or atherosclerosis, valvular or rheumatic heart disease, cardiomyopathy, aortic disease, and venous thromboembolism. Diagnosis definitions were based on prespecified ICD-9 and ICD-10 code prefixes; the full diagnosis code sets are provided in \u003cb\u003eTable S7\u003c/b\u003e. An all-cardiovascular conditions subgroup, defined as the union of these groups, served as the primary analytic cohort for the dose- and weight-loss-based landmark analyses; individual cardiovascular subgroups were examined separately in the comparative outcome analyses. We also calculated the interval between the index date and the most recent qualifying preindex diagnosis code for each cardiovascular subgroup; these summaries are provided in \u003cb\u003eTable S8\u003c/b\u003e.\u003c/p\u003e\u003ch2\u003eCardiovascular Outcomes\u003c/h2\u003e\u003cp\u003eThe primary comparative outcomes were all-cause mortality, composite cardiovascular events (all-cause death, MI, or stroke), and incident cardiovascular diagnoses. All-cause mortality was defined by the recorded death date. The composite cardiovascular endpoint was defined as all-cause death or the first post-index diagnosis consistent with acute myocardial infarction, acute coronary syndrome, or cerebrovascular event; component ICD-9 and ICD-10 code definitions are provided in \u003cb\u003eTable S9\u003c/b\u003e. Incident cardiovascular outcomes included incident ischemic heart disease, cerebrovascular disease, hypertension, heart failure, arrhythmias or conduction disorders, peripheral vascular disease, or atherosclerosis, valvular or rheumatic heart disease, cardiomyopathy, aortic disease, and venous thromboembolism. For each incident outcome analysis, patients with evidence of that same condition before the start of follow-up were excluded.\u003c/p\u003e\u003ch2\u003eDose–Weight-Loss Relationship\u003c/h2\u003e\u003cp\u003eTo characterize the dose–response relationship between semaglutide and weight reduction, we examined the distribution of maximum percent body-weight change achieved within the 2-year landmark window according to the maximum semaglutide dose reached during that period. Dose categories corresponded to labeled dosing increments for injectable semaglutide formulations (Ozempic and Wegovy): 0.25 mg, 0.5 mg, 1.0 mg, 1.7 mg, 2.0 mg, and 2.4 mg or greater. Observed semaglutide dose values were used when directly available in the medication record, and missing dose values were extracted from medication descriptions using a large language model-assisted dose-mapping workflow based on GPT-OSS-20B\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. The resulting normalized dose values were used to assign patients to labeled dosing categories. This workflow increased dose ascertainment coverage from 56.0% to 98.2% of semaglutide prescription records. Manual review of 148 unique medication descriptions confirmed correct dose assignment in every case; the only apparent ambiguities (13 descriptions, 8.8%) arose from multi-dose pen devices listing two possible doses (e.g., \"0.25 mg or 0.5 mg\"), which were assigned the higher of the two labeled doses as a convention. A linear regression was fitted to the mean maximum weight-loss percentage at each dose level to quantify the overall dose-response trend.\u003c/p\u003e\u003ch2\u003eLandmark Analyses of Cardiovascular Outcomes by Dose and Weight Loss at 2 Years\u003c/h2\u003e\u003cp\u003eWe performed within-drug landmark analyses to evaluate the association of dose intensity and weight loss with subsequent cardiovascular outcomes among semaglutide-treated patients. The landmark was defined at 2 years after the index date. Outcomes were assessed only after the landmark date, and patients who experienced the outcome of interest on or before the landmark were excluded from the corresponding analysis. In the post-landmark period, patients were followed until the outcome of interest, death, or the last observed clinical record, whichever occurred first.\u003c/p\u003e\u003cp\u003eFor dose-based analyses, we used all prescriptions from the index date through the landmark to define the maximum dose reached by the landmark. Semaglutide dose was dichotomized as low (0.25–1.0 mg) or high (1.7 mg or greater). Patients were considered analyzable if follow-up extended through the landmark and at least one dose-bearing prescription was available by that time point.\u003c/p\u003e\u003cp\u003eFor weight-loss-based analyses, baseline weight was defined as the measurement closest to the index date, from 90 days before through 14 days after index. Post-index weight change was evaluated beginning 15 days after index through the landmark to avoid overlap between baseline ascertainment and follow-up weight assessment. Maximum percent body-weight change was calculated as the difference between baseline weight and the lowest observed weight before the landmark, divided by baseline weight, multiplied by 100. Patients were categorized into prespecified strata: less than 5%, 5–10%, 10–15%, 15–20%, 20–25%, and 25% or greater. Analyses were restricted to patients with an available baseline weight and at least one follow-up weight before the landmark.\u003c/p\u003e\u003cp\u003eFor both dose and weight-loss landmark analyses, cardiovascular outcomes were evaluated using cumulative event curves over the 24-month post-landmark period and overall post-landmark event proportions.\u003c/p\u003e\u003ch2\u003ePost-Landmark Metabolic Surrogates, Prescription Continuity, and Sensitivity Analyses\u003c/h2\u003e\u003cp\u003ePost-landmark metabolic surrogates included HbA1c and systolic and diastolic blood pressure. For each patient, the post-landmark value was defined as the closest measurement to the landmark date within the 24-month post-landmark period. Change from baseline was calculated as the difference between the post-landmark value and the most recent pre-treatment measurement. These surrogates were summarized across weight-loss strata.\u003c/p\u003e\u003cp\u003eTo assess semaglutide prescription continuity relative to the landmark, we computed per-patient prescription counts in two windows: the pre-landmark window spanning from the first semaglutide prescription date to the 2-year landmark date, and the post-landmark window spanning the 24 months following the landmark date (truncated at the last observation date). The per-patient difference in prescription count (post minus pre) was computed and summarized as median [IQR] stratified by dose group and weight-loss category.\u003c/p\u003e\u003cp\u003eIn addition to the shared 2-year landmark, two sensitivity analyses were performed: the first anchored follow-up to the date of first attainment of the patient's maximum semaglutide dose, and the second to the date of maximum body-weight reduction. In both cases, outcomes were evaluated over the subsequent 24 months to test whether the patterns observed with the fixed landmark were robust when aligned to the time points at which peak exposure or peak response was achieved.\u003c/p\u003e\u003ch2\u003ePropensity-Matched Comparative Cardiovascular Analyses\u003c/h2\u003e\u003cp\u003eFor comparator analyses, three additional mutually exclusive incident-user cohorts were constructed for metformin, DPP-4 inhibitors, and SGLT2 inhibitors using the same index-date and exclusion logic described above. Within each baseline cardiovascular subgroup, we performed pairwise comparisons anchored on semaglutide as the primary exposure cohorts. Separate analyses were conducted for semaglutide versus metformin, DPP-4 inhibitors, and SGLT2 inhibitors. Propensity scores were estimated with logistic regression using age at index, sex, baseline BMI, and baseline type 2 diabetes status. Patients were matched 1:1 without replacement using nearest-neighbor matching and a caliper of 0.2 on the propensity score scale.\u003c/p\u003e\u003cp\u003eFollow-up for comparative outcome analyses began on the index date. Patients were followed until the outcome of interest, death, or the last observed clinical record. For incident cardiovascular outcomes, patients with prevalent disease of the same type at baseline were excluded from that specific analysis. We summarized matched cohort size, event counts, person-time, and cumulative 2-year event estimates using Kaplan-Meier methods and log-rank P values.\u003c/p\u003e\u003ch2\u003eAnalysis of GLP1R Expression Using Single-Cell and Bulk Tissue Atlases\u003c/h2\u003e\u003cp\u003eSingle-cell expression of \u003cem\u003eGLP1R\u003c/em\u003e was obtained from the CELLxGENE data portal\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cellxgene.cziscience.com/\u003c/span\u003e\u003cspan class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), leveraging 1556 publicly available human single-cell and single-nucleus RNA sequencing datasets corresponding to 86\u0026nbsp;million cells from 11,633 human donors. Using the CELLxGENE Census, GLP1R expression was obtained across available cell types. Gene expression values were extracted as normalized counts (counts per 10,000 [CP10K], log-transformed where applicable) and analyzed at the cell-type level using curated annotations provided within each dataset. For each cell type, \u003cem\u003eGLP1R\u003c/em\u003e expression was summarized as both mean normalized expression and the proportion of expressing cells (non-zero counts). Where multiple datasets contributed to the same tissue, results were aggregated to ensure robustness across cohorts, and analyses were restricted to cell types with adequate representation to minimize sparsity-driven artifacts.\u003c/p\u003e\u003cp\u003eBulk tissue-level expression of GLP1R was obtained from the Human Protein Atlas (proteinatlas.org). Normalized transcript expression values (e.g., nTPM) were retrieved across human tissues, with particular emphasis on cardiac subregions where available.\u003c/p\u003e\u003ch2\u003eSingle Cell RNA-sequencing profiling of GLP1R across human body tissues\u003c/h2\u003e\u003cp\u003e \u003cb\u003e​​\u003c/b\u003eSingle-cell GLP1R atlas data were analyzed as follows. For each tissue-cell annotation, Expression (CP10K), number of GLP1R-positive cells, and percentage of expressing cells were analyzed.\u003c/p\u003e\u003cp\u003eThe primary metric was \u003cem\u003eGLP1R Engagement Potential (GEP)\u003c/em\u003e, defined as Expression (CP10K) × (% expressing cells / 100). Conceptually, GEP is a prevalence-weighted expression metric: it becomes high when a cell class has both appreciable GLP1R transcript abundance and a meaningful proportion of cells that are GLP1R-positive. It is therefore useful for ranking which cellular compartments are most likely to be engaged under a broad systemic exposure model because it balances intensity and prevalence rather than privileging either one alone. A compartment with very high expression in only a tiny number of cells may not rank as highly by GEP as a compartment with slightly lower expression but a much broader GLP1R-positive fraction.\u003c/p\u003e\u003cp\u003eThe other salient metric was \u003cem\u003eAbsolute Target Load (ATL)\u003c/em\u003e, defined as the product of per-cell GLP1R expression and the absolute number of GLP1R-positive cells in that compartment. This is computed as Expression (CP10K) × NumberPositiveCells, where NumberPositiveCells is the number of cells expressing GLP1R in that tissue-cell class. Conceptually, ATL estimates the total transcript-weighted burden of GLP1R-positive cells and is therefore a whole-compartment burden metric rather than a prevalence-weighted one. It becomes especially informative at the tissue or organ level, where a compartment may have only modest prevalence but still contain a very large absolute reservoir of GLP1R-positive cells because the total sampled population is large. For this reason, ATL is particularly useful for understanding overall organ-scale target burden and for distinguishing tissues that may represent major aggregate pharmacologic reservoirs even if their percentage of positive cells is not the highest.\u003c/p\u003e\u003cp\u003eStable cell classes were filtered for visualization using Cell Count \u0026gt; = 100 and NumberPositiveCells \u0026gt; = 2.\u003c/p\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eContinuous variables were summarized as means with standard deviations or medians with interquartile ranges, as appropriate; categorical variables were summarized as counts and percentages. In the comparative cardiovascular outcome analyses, between-cohort differences were assessed using propensity score–matched Kaplan-Meier estimates of 2-year event risk; statistical significance in KM-difference heatmaps was determined by log-rank test. In the dose landmark analyses, cumulative event distributions were compared between high-dose and low-dose groups using log-rank tests, and post-landmark event rates were compared using relative risks with associated Wald p-values. In the weight-loss landmark analyses, cumulative event distributions and post-landmark event rates across weight-loss strata were compared using a global log-rank test. Trends in post-landmark metabolic surrogates (HbA1c, systolic and diastolic blood pressure) across weight-loss strata were assessed using one-way ANOVA. Analyses were performed in Python 3.13.1 using pandas 3.0.0, NumPy 2.4.2, SciPy 1.17.0, and Matplotlib 3.10.8; propensity score modeling, survival analyses and single-cell expression profiling were implemented within the same analytic pipeline.\u003c/p\u003e\u003ch2\u003eData Source\u003c/h2\u003e\u003cp\u003eThis study analyzed de-identified EHR data from academic medical centers in the United States via the nference nSights Analytics Platform. Prior to analysis, all data underwent expert determination de-identification satisfying HIPAA Privacy Rule requirements (45 CFR §\u0026nbsp;164.514(b)(1)), employing a multi-layered transformation approach for both structured data (cryptographic hashing of identifiers, date-shifting, geographic truncation) and unstructured clinical text (ensemble deep learning and rule-based methods with \u0026gt; 99% recall for personally identifiable information detection)\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. nference established secure data environments within each participating center, housing these de-identified patient data governed by expert determination. These de-identified data environments were specifically designed to enable data access and analysis without requiring Institutional Review Board oversight, approval, or exemption confirmation. Accordingly, informed consent and IRB review were not required for this study.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThis study involves the analysis of de-identified Electronic Health Record (EHR) data via the nference nSights Federated Clinical Analytics Platform (nSights). Data shown and reported in this manuscript were extracted from this environment using an established protocol for data extraction, aimed at preserving patient privacy. The data has been de-identified pursuant to an expert determination in accordance with the HIPAA Privacy Rule. Any data beyond what is reported in the manuscript, including but not limited to the raw EHR data, cannot be shared or released due to the parameters of the expert determination to maintain the data de-identification. The corresponding author should be contacted for additional details regarding nSights.\u003c/p\u003e\u003ch2\u003eDe-identification and HIPAA compliance certification\u003c/h2\u003e\u003cp\u003ePrior to analysis, all EHR data were de-identified under an expert determination consistent with the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule (45 CFR §\u0026nbsp;164.514(b)(1)). The de-identification methodology employed a multi-layered transformation approach to both structured and unstructured data fields\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e,\u003cspan class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. In structured data, direct identifiers including patient names and precise geographic locations were excluded entirely, while indirect identifiers underwent specific transformations: patient identifiers, medical record numbers, and accession numbers were replaced with one-way cryptographic hashes using confidential salts to preserve linkage across patient encounters; all dates were shifted backward by patient-specific random offsets (1–31 days) to preserve temporal relationships while obscuring exact event timing; the ZIP codes were truncated to two-digit state-level resolution; and continuous variables including age, height, weight, and body mass index were thresholded to prevent identification of extreme values (for example, ages ≥ 89 years transformed to ‘89+’ and BMI \u0026gt; 40 transformed to ‘40+’). In unstructured clinical text, an ensemble de-identification system that combines attention-based deep learning models with rule-based methods achieved an estimated \u0026gt; 99% recall for personally identifiable information (PII) detection, with detected identifiers replaced by plausible fictional surrogates\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003ch2\u003eData Harmonization\u003c/h2\u003e\u003cp\u003eTo address heterogeneity in EHR data, we harmonized clinical variables including medications, anthropometric measurements, and diagnoses to standardized concepts. For medications, we first constructed a standardized drug concept database combining the nSights knowledge graph with RXNorm (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.nlm.nih.gov/research/umls/rxnorm/index.html\u003c/span\u003e\u003cspan class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) hierarchies to capture ingredient, brand, and dose-specific information\u003csup\u003e\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. EHR medication records were matched using a hierarchical approach prioritizing RXNorm codes when available, followed by ingredient-level matching, and finally natural language processing and pattern matching on free-text medication orders when structured codes were absent. For anthropometric measurements (height, weight, BMI), we created a unified vocabulary from SNOMED (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.snomed.org/\u003c/span\u003e\u003cspan class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://athena.ohdsi.org\u003c/span\u003e\u003cspan class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and LOINC (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://loinc.org/\u003c/span\u003e\u003cspan class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) terminologies and matched EHR measurement descriptions using standardized text matching algorithms with abbreviation expansion and synonym resolution; ambiguous mappings were resolved using OpenAI GPT-4o (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://platform.openai.com/docs/models/gpt-4o\u003c/span\u003e\u003cspan class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) with summary statistics as context, followed by manual verification. For diagnoses, we developed a hierarchical disease concept database from the nSights knowledge graph and matched EHR diagnosis descriptions and codes by identifying the most specific common child concept in the hierarchy. This approach enabled consistent identification of clinical entities while preserving granularity where available.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eConflict of Interest Statement\u003c/h2\u003e \u003cp\u003eThe authors are employees of nference, inc., which conducts research collaborations with various biopharmaceutical companies whose therapeutic products are included in this study. None of these companies, nor any other nference collaborator, funded, supported, or had any role in the independent study design, data acquisition, analysis, interpretation, manuscript preparation, or the decision to submit this work for publication. All analyses were conducted by the authors using de-identified electronic health record data. The authors declare no additional competing interests.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003cp\u003eThe authors are employees of nference, inc., which conducts research collaborations with various biopharmaceutical companies whose therapeutic products are included in this study. None of these companies, nor any other nference collaborator, funded, supported, or had any role in the independent study design, data acquisition, analysis, interpretation, manuscript preparation, or the decision to submit this work for publication. All analyses were conducted by the authors using de-identified electronic health record data. The authors declare no additional competing interests.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis research received no external funding.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eV.S. conceived the study and supervised the overall project. A.J.V. and K.M. designed the analytical framework. K.M. led the data curation and statistical analyses, with support from A.J.V and V.S. K.M. implemented the AI-enabled phenotyping pipelines and performed the longitudinal modeling. A.J.V., K.M. and V.S. contributed equally to data interpretation. V.S., A.J.V. and K.M. drafted the initial manuscript and critically revised the manuscript with inputs from C.G. All authors reviewed, edited, and approved the final version of the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003e We thank the nference engineering team for the development of the nSights federated AI platform, and Patrick Lenehan for critical review and manuscript feedback.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThis study involves the analysis of de-identified Electronic Health Record (EHR) data via the nference nSights Federated Clinical Analytics Platform (nSights). Data shown and reported in this manuscript were extracted from this environment using an established protocol for data extraction, aimed at preserving patient privacy. The data has been de-identified pursuant to an expert determination in accordance with the HIPAA Privacy Rule. Any data beyond what is reported in the manuscript, including but not limited to the raw EHR data, cannot be shared or released due to the parameters of the expert determination to maintain the data de-identification. The corresponding author should be contacted for additional details regarding nSights.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eZhou, X.-D. \u003cem\u003eet al.\u003c/em\u003e Burden of disease attributable to high body mass index: an analysis of data from the Global Burden of Disease Study 2021. \u003cem\u003eEClinicalMedicine\u003c/em\u003e 76, 102848 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilding, J. P. H. \u003cem\u003eet al.\u003c/em\u003e Once-Weekly Semaglutide in Adults with Overweight or Obesity. \u003cem\u003eN Engl J Med\u003c/em\u003e 384, 989\u0026ndash;1002 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarso, S. P. \u003cem\u003eet al.\u003c/em\u003e Semaglutide and Cardiovascular Outcomes in Patients with Type 2 Diabetes. \u003cem\u003eN Engl J Med\u003c/em\u003e 375, 1834\u0026ndash;1844 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLincoff, A. M. \u003cem\u003eet al.\u003c/em\u003e Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes. \u003cem\u003eN Engl J Med\u003c/em\u003e 389, 2221\u0026ndash;2232 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKosiborod, M. N. \u003cem\u003eet al.\u003c/em\u003e Semaglutide in Patients with Heart Failure with Preserved Ejection Fraction and Obesity. \u003cem\u003eN Engl J Med\u003c/em\u003e 389, 1069\u0026ndash;1084 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeanfield, J. \u003cem\u003eet al.\u003c/em\u003e Semaglutide and cardiovascular outcomes in patients with obesity and prevalent heart failure: a prespecified analysis of the SELECT trial. \u003cem\u003eLancet\u003c/em\u003e 404, 773\u0026ndash;786 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWharton, S. \u003cem\u003eet al.\u003c/em\u003e Once-weekly semaglutide 7\u0026middot;2 mg in adults with obesity (STEP UP): a randomised, controlled, phase 3b trial. \u003cem\u003eLancet Diabetes Endocrinol\u003c/em\u003e 13, 949\u0026ndash;963 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerkovic, V. \u003cem\u003eet al.\u003c/em\u003e Effects of Semaglutide on Chronic Kidney Disease in Patients with Type 2 Diabetes. \u003cem\u003eN Engl J Med\u003c/em\u003e 391, 109\u0026ndash;121 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcGuire, D. K. \u003cem\u003eet al.\u003c/em\u003e Oral Semaglutide and Cardiovascular Outcomes in High-Risk Type 2 Diabetes. \u003cem\u003eN Engl J Med\u003c/em\u003e 392, 2001\u0026ndash;2012 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBonaca, M. P. \u003cem\u003eet al.\u003c/em\u003e Semaglutide and walking capacity in people with symptomatic peripheral artery disease and type 2 diabetes (STRIDE): a phase 3b, double-blind, randomised, placebo-controlled trial. \u003cem\u003eLancet\u003c/em\u003e 405, 1580\u0026ndash;1593 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVenkatakrishnan, A. J. \u003cem\u003eet al.\u003c/em\u003e Clinical nSights: A software platform to accelerate real world oncology analyses. \u003cem\u003eJournal of Clinical Oncology\u003c/em\u003e (2024) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1200/JCO.2024.42.16_suppl.e23316\u003c/span\u003e\u003cspan address=\"10.1200/JCO.2024.42.16_suppl.e23316\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVenkatakrishnan, A. J., Murugadoss, K. \u0026amp; Soundararajan, V. Decoding the hallmarks of GLP-1RA weight-loss super responders. \u003cem\u003emedRxiv\u003c/em\u003e 2025.11.15.25340314 (2025) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/2025.11.15.25340314\u003c/span\u003e\u003cspan address=\"10.1101/2025.11.15.25340314\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVenkatakrishnan, A. J., Murugadoss, K. \u0026amp; Soundararajan, V. Weight Loss Dynamics and Health Burden Changes with Tirzepatide versus Semaglutide. \u003cem\u003emedRxiv\u003c/em\u003e 2025.11.30.25341294 (2025) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.64898/2025.11.30.25341294\u003c/span\u003e\u003cspan address=\"10.64898/2025.11.30.25341294\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurugadoss, K., Varma, G., Venkatakrishnan, A. J., Gibson, M. C. \u0026amp; Soundararajan, V. Weight trajectories after last Tirzepatide or Semaglutide prescription across a federated health network. \u003cem\u003emedRxiv\u003c/em\u003e 2026.01.26.26344839 (2026) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.64898/2026.01.26.26344839\u003c/span\u003e\u003cspan address=\"10.64898/2026.01.26.26344839\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOpenAI \u003cem\u003eet al.\u003c/em\u003e gpt-oss-120b \u0026amp; gpt-oss-20b Model Card. (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCZI Cell Science Program \u003cem\u003eet al.\u003c/em\u003e CZ CELLxGENE Discover: a single-cell data platform for scalable exploration, analysis and modeling of aggregated data. \u003cem\u003eNucleic Acids Res\u003c/em\u003e 53, D886\u0026ndash;D900 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurugadoss, K. \u003cem\u003eet al.\u003c/em\u003e Building a best-in-class automated de-identification tool for electronic health records through ensemble learning. \u003cem\u003ePatterns (N Y)\u003c/em\u003e 2, 100255 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurugadoss, K. \u003cem\u003eet al.\u003c/em\u003e Scaling text de-identification using locally augmented ensembles. \u003cem\u003emedRxiv\u003c/em\u003e 2024.06.20.24308896 (2024) doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/2024.06.20.24308896\u003c/span\u003e\u003cspan address=\"10.1101/2024.06.20.24308896\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePacker, M. \u003cem\u003eet al.\u003c/em\u003e Tirzepatide for Heart Failure with Preserved Ejection Fraction and Obesity. \u003cem\u003eN Engl J Med\u003c/em\u003e 392, 427\u0026ndash;437 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcLean, B. A., Wong, C. K., Kabir, M. G. \u0026amp; Drucker, D. J. Glucagon-like Peptide-1 receptor Tie2\u0026thinsp;+\u0026thinsp;cells are essential for the cardioprotective actions of liraglutide in mice with experimental myocardial infarction. \u003cem\u003eMol Metab\u003c/em\u003e 66, 101641 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWebsite. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.fda.gov/news-events/press-announcements/fda-approves-fourth-product-under-national-priority-voucher-program-higher-dose-semaglutide\u003c/span\u003e\u003cspan address=\"https://www.fda.gov/news-events/press-announcements/fda-approves-fourth-product-under-national-priority-voucher-program-higher-dose-semaglutide\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKosiborod, M. N. \u003cem\u003eet al.\u003c/em\u003e Semaglutide in Patients with Obesity-Related Heart Failure and Type 2 Diabetes. \u003cem\u003eN Engl J Med\u003c/em\u003e 390, 1394\u0026ndash;1407 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHelmst\u0026auml;dter, J. \u003cem\u003eet al.\u003c/em\u003e Endothelial GLP-1 (Glucagon-Like Peptide-1) Receptor Mediates Cardiovascular Protection by Liraglutide In Mice With Experimental Arterial Hypertension. \u003cem\u003eArterioscler Thromb Vasc Biol\u003c/em\u003e 40, 145\u0026ndash;158 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark, B. \u003cem\u003eet al.\u003c/em\u003e GLP-1 receptor agonists and atherosclerosis protection: the vascular endothelium takes center stage. \u003cem\u003eAm J Physiol Heart Circ Physiol\u003c/em\u003e 326, H1159\u0026ndash;H1176 (2024).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-cardiovascular-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [npj Cardiovascular Health](https://www.nature.com/npjcardiohealth)","snPcode":"44325","submissionUrl":"https://submission.springernature.com/new-submission/44325/3","title":"npj Cardiovascular Health","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-9407045/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9407045/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSemaglutide is often optimized for weight loss, but whether longer-term cardiovascular benefit tracks achieved weight loss or therapeutic exposure levels remains unclear. We conducted a retrospective observational study using a federated, deidentified U.S. electronic health record network and applied multimodal AI-based methods to analyze 47,199 patients with baseline cardiovascular disease. We quantified dose escalation and weight change during the 0\u0026ndash;2-year period after semaglutide initiation (landmark period) and assessed cardiovascular outcomes during the 2\u0026ndash;4-year period (post-landmark). To mitigate confounding, we performed propensity-matched comparisons during the landmark period, in which semaglutide was associated with lower rates of cardiovascular events than metformin, DPP-4 inhibitors, and SGLT2 inhibitors; however, these findings should be interpreted as associative and remain susceptible to treatment selection bias. Higher maximum semaglutide dose was associated with greater weight loss during the landmark period (3.15% additional weight loss per 1 mg increase; r\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.97, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and with lower post-landmark risk of all-cause mortality (RR 0.42, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), composite cardiovascular events (death, myocardial infarction, or stroke; RR 0.51, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), cerebrovascular disease (RR 0.50, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), heart failure (RR 0.55, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and valvular or rheumatic heart disease (RR 0.71, P\u0026thinsp;=\u0026thinsp;0.025), providing robust associative evidence that supports prospective evaluation of causal relationships. In contrast, greater achieved weight loss during the landmark period did not show a consistent monotonic association with lower post-landmark cardiovascular risk (All-cause mortality p-value\u0026thinsp;=\u0026thinsp;0.14, composite cardiovascular endpoint p-value\u0026thinsp;=\u0026thinsp;0.55). Integrating insights from a single cell GLP1R expression atlas was used to infer how semaglutide pharmacology may tie into heart-specific signaling, beyond what is reflected by body-weight reduction alone. The strongest prevalence-weighted GLP1R signal was observed in the pancreas, followed by the heart, where GLP1R engagement potential (GEP) was considerable across cardiomyocyte, cardiac endothelial, and rarer immune cell populations. Together, in this retrospective observational study, semaglutide-associated cardiovascular benefit appears more closely aligned with maximum dose attained than with achieved weight-loss magnitude, supporting prospective validation and motivating beyond-obesity trial designs that integrate whole-body spatial intelligence.\u003c/p\u003e","manuscriptTitle":"Semaglutide cardiovascular outcomes align more closely with attained dose than achieved weight loss","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-04 09:41:52","doi":"10.21203/rs.3.rs-9407045/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-18T01:32:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"216337269410938221042610054600360347750","date":"2026-05-13T05:56:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"336699271479188899347313400870030107244","date":"2026-04-24T15:42:49+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"152505262885852687411265293139828703582","date":"2026-04-22T11:09:02+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-22T10:38:43+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-19T16:15:39+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-16T14:55:28+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Cardiovascular Health","date":"2026-04-13T17:07:05+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-cardiovascular-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [npj Cardiovascular Health](https://www.nature.com/npjcardiohealth)","snPcode":"44325","submissionUrl":"https://submission.springernature.com/new-submission/44325/3","title":"npj Cardiovascular Health","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0d170bc4-16a0-4f9e-bc77-61776d038925","owner":[],"postedDate":"May 4th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-18T01:32:59+00:00","index":20,"fulltext":""},{"type":"reviewerAgreed","content":"216337269410938221042610054600360347750","date":"2026-05-13T05:56:00+00:00","index":19,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":67141207,"name":"Health sciences/Cardiology"},{"id":67141208,"name":"Health sciences/Diseases"},{"id":67141209,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2026-05-04T09:41:53+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-04 09:41:52","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9407045","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9407045","identity":"rs-9407045","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}