Circulating Levels of Angiotensinogen, Sex, and Hormone Therapy - The Multi-Ethnic Study of Atherosclerosis (MESA)

1

Angiotensinogen, the unique precursor of all angiotensin hormones of the Renin -2 Angiotensin-Aldosterone System (RAAS), is now a potential target in a novel pharmacological 3 approach to hypertension. Investigating the factors that influence angiotensinogen levels , 4 including sex hormones, may have important therapeutic implications. 5 6

Plasma angiotensinogen and sex hormones levels were measured in 5,171 Multi -7 Ethnic Study of Atherosclerosis (MESA) participants. Linear models were employed to 8 determine the associations of angiotensinogen with sex hormones, and mediation analysis was 9 performed to evaluate the effect of HT on blood pressure (BP) and hypertension through 10 angiotensinogen. 11 12

Angiotensinogen levels were significantly higher in postmenopausal women receiving 13 HT (n=760) compared to women not receiving HRT (n=1,675) and i n men (n=2,736). A positive 14 association was present between angiotensinogen and estrogen levels that differed in magnitude 15 between sexes and by HT status among postmenopausal women (women on HT: r=0.44, p< 16 0.0001; women not on HT: r=0.09, p=0.0002; and men : r= 0.07, p=0.0003). The type of HT 17 formulation (estrogen or estrogen/progesterone) and its duration of use did not significantly 18 affect angiotensinogen levels. HT indirectly increased systolic BP (β=1.24) and the odds of 19 hypertension (OR=1.065) through its effect of increasing angiotensinogen. 20 21

A positive association was present between angiotensinogen and estrogen levels 22 that differed by HT status. HT impacts systolic BP and hypertension in directly by increasing 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint angiotensinogen. This study underscores the role of angiotensinogen in hypertension, and the 1 complex relationship between HT and hypertension. 2 3

angiotensinogen, estradiol, hormone therapy, Renin -Angiotensin-Aldosterone 4 System 5 6

7 Angiotensinogen is the liver -derived protein precursor of all angiotensin (Ang) peptides of the 8 renin-angiotensin-aldosterone system (RAAS) including Ang I an d Ang II 1–3. The RAAS is 9 critical to the pathology of hypertension . Therapeutics that target the downs tream metabolism 10 and reception of metabolites of angiotensinogen are proven to improve clinical outcomes for 11 patients with hypertension, and to reduce the risk of heart failure and myocardial infarction 4–6. 12 However, t herapies targeting the metabolism or reception of Ang peptides derived from 13 angiotensinogen often lead to a compensatory pathway that limits therapeutic efficacy as 14 monotherapy 2. This necessitates the use of a combination with other classes of drugs to achieve 15 adequate blood pressure control in many patients. 16 17 Angiotensinogen is produced primarily by the liver, but the other components of the RAAS 18 pathway are mainly active in the kidney. Angiotensinogen is now a target of therapy in phase 1 19 and 2 trials. It is proposed that targeting angiotensinogen may optimally inhibit the RAAS 20 pathway and provide more effective and possibly safer alternative by inhibit ing angiotensinogen 21 in the liver instead of the kidney 7,8. 22 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 4 Although angiotensinogen plays an essential role in regulating RAAS, little is known about key 1 determinants of angiotensinogen plasma levels. Estrogen may be a significant contributor to 2 expression and activity of RAAS components through several pathways 9–11. Previous studies 3 have demonstrated that circulating levels of angiotensinogen are related to sex 12,13, and that the 4 administration of exogenous estrogen to premenopausal women in contraceptive formulat ions 14 5 and to postmenopausal women as HT 15–17 increases plasma concentrations of circulating 6 angiotensinogen. However, most of these studies are limited by small sample size and were 7 conducted in exclusively normotensive women. 8 9 The effect of natural and supplemental estrogen on angiotensinogen levels and the factors 10 underlying how sex and sex hormones impact angiotensinogen levels remains poorly understood. 11 The current study aimed to better delineate the complex interactions betwe en sex, estrogen, and 12 HT with circulating plasma levels of angiotensinogen and their effect on blood pressure/ 13 hypertension in a community-dwelling multi-ethnic cohort. 14 15

16 Study participants and design 17 The Multiethnic Study of Atherosclerosis (MESA) enrolled 6,814 participants (3,601 women and 18 3,213 men) with no -known clinical atherosclerotic cardiovascular disease of four ethnic 19 backgrounds (White, Chinese, Black, and Hispanic), with age between 45 -84 years from six 20 communities in the United States (Forsyth County, NC; Northern Manhattan and the Bronx, NY; 21 Baltimore, MD; St. Paul, MN; Chicago, IL; and Los Angeles County, CA). The design of the 22 study has been described in detail in an earlier publication 18. An ancillary study was conducted 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 5 by our group that measured angiotensinogen levels in MESA participants who were not included 1 in the MESA -1000. The MESA -1000 is a random sample of approximately 1,000 participants 2 that has undergone additional tests using baseline blood samples. To minimize depletion of these 3 specimens, ancillary studies exclude the MESA -1000 participants, unless otherwise justified. 4 Therefore, this study included 5,786 participants. Medical history, laboratory data, and 5 anthropometric me asurements were ascertained as described previously 18,19. The study was 6 approved by the institutional review boards of the participating institutions, and all participants 7 gave written informed consent. 8 9 Our study population ( Figure 1 ) consisted of all men and postmenopausal wome n who had 10 plasma angiotensinogen levels and sex hormones measured at baseline (n=5,292). Measured sex 11 hormones included total testosterone, dehydroepiandrosterone (DHEA), and estradiol (E2). 12 Additionally, sex hormone binding globulin (SHBG) was measured gi ven its role in transporting 13 hormones in circulation. We sequentially excluded 97 pre -menopausal women, and 24 women 14 with missing information on HT use. Our analytical sample was composed of 2,435 15 postmenopausal women and 2,736 men. 16 17 Menopausal status was determined through an algorithm developed using answers to a series of 18 self-reported questions that included age, age at menopause, history of surgical menopause and 19 age at surgical menopause as compared to age ( Supplemental Figure 1 ) 20. In determinate 20 menopausal women were not included in this analysis. Women were considered postm enopausal 21 if they were older than 55 years of age or had undergone a bilateral oophorectomy and/or self -22 reported being postmenopausal or having an absence of menstrual periods in the preceding year 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 6 20. HT consisted of estrogen alone (Premarin or Estratab) (n = 251, 33.0%) or estrogen with 1 progesterone (Premarin plus Provera, Estrata b plus Provera, Prempro, or Premphase) (n = 98, 2 12.9%), while 54.1% (n = 411) had no data on the type of HT. Time duration of HT use ranged 3 from 0 to 49 years (Median = 10.0 years). 4 5 Laboratory measurements 6 Circulating angiotensinogen levels and sex hormones were measured at the baseline (Exam 1) 7 visit of the included MESA participants. Angiotensinogen measurements were made using an 8 enzyme-linked immunoassay that has been described previously 7 and was executed by Medpace 9

Labs (Cincinnati, Ohio). Briefly, the coating antibody was rabbit anti -human 10 immunoglobulin monoclonal antibody (IBL -America, Minneapolis, Minnesota). Total 11 angiotensinogen (including intact angiotensinogen and des[Ang I] angiotensinogen) was then 12 detected from 1:10,000 diluted EDTA plasma with horseradish peroxidase mouse anti -human 13 angiotensinogen monoclonal Fab’ fragment (IBL -America). A standard curve was generated for 14 quantitation using purified human angiotensinogen. The coefficient of variation was observed to 15 be 9% with an analytical range of 13.9 µg/mL to 75.9 µg/mL. 16 17 Endogenous sex hormone levels were measured from fasting serum samples at the University of 18 Massachusetts Medical Center in Worcester, MA. The assays and kits used to measure the 19 hormone levels were an ultrasensitive radioimmunoassay kit for E2 (Diagnostic System 20 Laboratories, Webster, TX), radioimmunoassay kits for total testosterone and DHEA and a 21 chemiluminescence enzyme immunometric assay using Immulite kits for SHBG (Diagnostic 22 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 7 Products Corporation, Los Angeles, CA) 20–23 . The intra-assay coefficients of variation for total 1 testosterone, SHBG, DHEA, and E2 were 12.3%, 9.0%, 11.2%, and 10.5%, respectively. 2 3 Other covariates 4 Covariates were obtained from standardized questionnaires, physical exam, and laboratory 5 measures at study visit 1 as previously reported 18,19. Age, race/ethnicity, smoking status, alcohol 6 history, and age at menopause were self -reported. A medication inventor y determined 7 medication use including use of HT and antihypertensive medications at visit 1 (baseline) . 8 Diabetes was assessed by self -reported physician diagnosis, a fasting glucose level of ≥126 9 mg/dL, or hypoglycemic medication use. Height and weight, measured using standardized 10 procedures, were used to calculate body mass index (BMI). Prevalent hyperten sion was defined 11 as systolic blood pressure (BP) >130 mm Hg, diastolic BP >80 mm Hg, or hypertension 12 medication use at examination 1 (baseline). 13 14 Statistical analysis 15 The distribution of cohort characteristics was determined within each sex, including the use of 16 HT in postmenopausal women. Median and quartile values (25 th and 75 th percentiles) of 17 circulating angiotensinogen levels were determined by sex, race/ethnicity, and HT in 18 postmenopausal women. 19 20 To determine the relationship between circulating angiotensinogen levels and sex by HT, an un-21 adjusted linear model was estimated by regressing levels of angiotensinogen on sex by HT and 22 estradiol levels. This model was then refitted to incorporate other c ovariates that had univariable 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 8 R2 values greater than 0.025 [race/ethnicity, BMI, total cholesterol, high-sensitivity C-reactive 1 protein (hs-CRP), total testosterone, DHEA, and SHBG). Un-adjusted and adjusted linear models 2 set men as the reference group, and each analyte was log -transformed and then scaled to have 3 mean zero and standard deviation 1. The model coefficients thus reflect a one standard deviation 4 change. 5 6 To investigate the potential mediating role of angiotensinogen in the relationship between HT 7 use and blood pressure/ hypertension, a mediation analysis was conducted to delineate: (1) the 8 direct effect of HT on blood pressure/ hypertension unrelated to angiotensinogen; and (2) the 9 indirect effect of HT on blood press ure/hypertension mediated by angiotensinogen. Regression 10 models were adjusted for age, BMI, race/ethnicity, and smoking history, while the use of any 11 hypertension medication was included as a covariate only for the analysis of blood pressure. 12 13 The analyses reported in the current work were conducted using the R statistical language 24 14 (version 4.0.2) and the following packages: ppcor 25, boot 26, emmeans 27, dplyr 28, ggplot2 29, 15 and mediation 30. 16 17

18 Baseline demographics of the 5,171 participants are reported in Table 1, stratified by sex and the 19 use of HT in postmenopausal women. Across all race/ethnicity groups, median angiotensinogen 20 levels were highest in postmenopausal women on HT, followed by women participants not on 21 therapy, followed by men (Medians: 36.0 µg/mL, 20.7 µg/mL, and 18.3 µg/mL, respectively) 22 (Figure 2). The most pronounced relative difference across race / ethnicity in angiotensinogen 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 9 levels was observed in the ratio between White and Chinese individuals. Postmenopausal White 1 women on HT have an angiotensinog en ratio of 1.26 (95% CI of 1.15 -1.38); followed by 2 postmenopausal white women not on HT (ratio = 1.10 and 95% CI of 1.05 -1.16); and men (ratio 3 = 1.06 and 95% CI of 1.02 -1.10) as compared to Chinese women and men (Figure 2, 4 Supplemental Table 1). 5 6 Positive associations between estradiol and angiotensinogen levels were observed in all 7 participants and were of greater magnitude in postmenopausal women on HT than those not on 8 HT and men (postmenopausal women on HT: r = 0.44 , p < 0.0001; postmenopausal women not 9 on HT: r = 0.09, p = 0.0002; and men: r = 0.07, p = 0.0003). In the regression model presented in 10 Figure 3, the association between estradiol and angiotensinogen differs by specific group ( men, 11 postmenopausal women not on HT, and postmenopausal women on HT). Using the model, we 12 compared angiotensinogen levels at the average estradiol levels for each group. At t he average 13 log-estradiol level for men, angiotensinogen was 18.2 g/mL with a 95% confidence interval of 14 (18.0, 18.4). At the average log -estradiol level for women not on HT it was 20.8 g/mL with a 15 CI of (20.6, 21.1). At the average log -estradiol levels f or postmenopausal women on HT it was 16 33.9 g/mL with a CI of (33.4, 34.5). For the same estradiol level, 0.2 nmol/L (mean for entire 17 cohort), men had a mean circulating angiotensinogen level of 18.7 g/mL, postmenopausal 18 women not on HT had a mean angiotensinogen level of 19.5 g/mL, and postmenopausal 19 women on HT had a mean angiotensinogen level of 33.9 g/mL (Figure 3). 20 21 An estimate one standard deviation (SD) higher log -estradiol level was associated with 0.195 22 SD, 0.585 SD, and 1. 716 SD higher log -angiotensinogen (95% CI: 0.169, 0.220) in m en, 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 10 postmenopausal women not on HT, and postmenopausal women on HT, respectively (Table 2). 1 Indicating smaller differences in angiotensinogen for the same amount of change in estradiol in 2 men and postmenopausal women not on HT as compared to those on HT (Table 2, Figure 3). 3 4 To further examine if the differences in the association between estrogen and angiotensinogen 5 between men, postmenopausal women not on HT and postmenopausal women on HT, we 6 adjusted the model described in Table 2 for variables with univariable model R 2 values greater 7 than 0.025 ( Supplemental Table 2 ). Following adjustment for race / ethnicity, BMI, total 8 cholesterol, hs-CRP, total testosterone, DHEA, and SHBG, the model coefficients for sex, HT, 9 and estradiol changed little in terms of direction or magnitude ( Supplemental Table 3 ). The 10 proportion of the variance in angiotensinogen explained by estrogen, sex and HT status alone is 11 43.5%; increases to 48.0% with the addition of the variables noted above (Table 2 and 12 Supplemental Table 3 ) and the partial R2 values for estradiol changed little following model 13 adjustment (Supplemental Table 4). 14 15 Different formulations of HT, only with estrogen or combined estrogen/proge sterone, showed 16 similar levels of angiotensinogen with median value of 36.0 µg/mL (95% CI: 32.7, 40.5) and 17 36.7 µg/mL (95% CI: 33.7, 40.5), respectively. In addition, the duration of HT use was not 18 associated with angiotensinogen levels. 19 20 In the mediation analysis, we examined the role of angiotensinogen in explaining the relationship 21 between HT and blood pressure/ prevalent hypertension. The indirect effect of HT mediated 22 through angiotensinogen levels was associated with increased systolic BP (β = 1.24, p = 0.040) 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 11 and odds of hypertension ( OR = 1.065, p < 0.001). The total effect of HT slightly increased the 1 odds of hypertension (OR = 1.082, p < 0.001) (Table 3). 2 3

4 The key findings of this study are as follows: (1) across all race/ethnicity groups, median 5 angiotensinogen levels were highest in postmenopausal women on HT, as compared to 6 postmenopausal women not on HT and men; (2) estrogen, sex and HT status alone accounts for 7 43.5% of variance in circulating angiotensinogen levels; (3) the indirect effect of HT mediated 8 through angiotensinogen levels increases systolic blood pressure and the odds of hypertension, 9 but did not have an impact on diastolic BP. These findings highlight the impact of 10 postmenopausal HT on circulating angiotensinogen levels, and the potential effect of HT on 11 prevalent hypertension through angiotensinogen levels. 12 13 Despite the availability of effective antihypertensive treatments, nearly half of patients with 14 hypertension fail to achieve BP goals recommended by guidelines. The pursuit of novel 15 antihypertensive therapies includes the d evelopment of antisense oligonucleotides (ASOs) and 16 small interfering RNA (siRNA), designed to reduce angiotensinogen secretion 2,3,8. Clinical trials 17 with ASO in Phase 1 (NCT03101878) on healthy volunteers and Phase 2 on patients with mild 18 (NCT03714776) and uncontrolled (NCT04083222) hypertension demonstrated favorable safety 19 and tolerability. These trials also showed targeted reductions in angiotensinogen leve ls, resulting 20 in significant decreases in both systolic and diastolic BP without compromising renal function or 21 increasing cardiovascular risk 7. Similarly, Phase 1 trials with siRNA (NCT03934307) in 22 uncontrolled hy pertension have shown promise in reducing blood pressure without worsening 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 12 renal function8,31,32. Phase II trials assessing efficacy and optimal dosing are currently underway, 1 with a predicted completion by 2025. Understanding the determinants of angiotensinogen levels, 2 including sex hormones, race, sex, and menopausal status may influence the efficacy of this new 3 class of therapeutics. 4 5 Consistent with our results, previous studies demonstrated that postmenopausal estrogen therapy 6 was associated with increased angiotensinogen plasma concentrations 16,33. The role of 7 exogenous estrogen in stimulating AGT gene expression or secretion have been demonstrated in 8 animal models and hepatic cell line 34–36. In addition Schunkert et al. 16 found increased 9 angiotensinogen levels to a similar extent in women taking formulations that contained only 10 estrogen or estrogen/progestin combination, both being significantly higher compared with 11 women without HT. Although there are recognized sex differences in blood pressure regulation 12 as well as cardiovascular disease susceptibility, onset, prevalence, clinical presentation, 13 pathophysiology, treatment responses and outcomes, the effects of exogenous estrogen on 14 angiotensinogen and its downstream metabolites, and consequently on blood pressure are limited 15 16,37. Trainor et al. (2023) 13, using the same cohort of the present study, reported positive 16 associations between angiotensinogen and blood pressure/prevalent hypertension. 17 18 The effect of HT on prevalent hypertension, evidenced by mediation analysis, indicates that HT 19 indirectly increases the odds of prevalent hypertension via increased angiotensinogen, which 20 contributes to the total effect of HT on hypertension ( Central Illustration). In systolic BP, this 21 indirect effect was counterbalanced by HT direct effects (independent of angiotensinogen) that 22 decrease systolic BP. This data suggests that the blood pressure effects of HT may be clinically 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 13 exploited by silencing the angiotensinogen mediated indirect blood pressure raising properties of 1 HT with newly developed therapeutics that directly target angiotensinogen production 2. 2 3 Previous studies have evaluated the relationship between HT use and the development of 4 hypertension. The Women’s Health Initiative I (WHI I, NCT00000611], a randomized controlled 5 trial (RCT) assessing primary prevention in 16,608 participants, observed an increase in systolic 6 blood pressure among postmenopausal women using combined estrogen/progestin therapy 38. 7 This is consistent with other prospective studies in normotensive postmenopausal women 39,40. 8 Conversely, the Kronos Early Estrogen Prevention Study (KEEPS, NCT00154180), a RCT with 9 727 participants reported a neutral effect of HT on systolic blood pressure 41. In postmenopausal 10 women with preexisting hypertension, randomized prospective studies with sample size less than 11 100 demonstrated that the HT use was associated with both a decrease 42,43 and neutral 44 effects 12 on blood pressure. Notably, the Heart and Estrogen/progestin Replacement Study I (HERS I, 13 NCT00319566) (RCT, n=2763) found no significant association between HT and the incidence 14 of myocardial infarction (MI) or stroke/transient ischemic attack in postmenopausal women with 15 established coronary disease 45. However, the WHI I study observed higher stroke rates in the 16 HT group compared to the placebo 38. A meta -analysis of 19 randomized controlled trials, 17 encompassing data from 40,410 postmenopausal women, furnished evidence that HT does not 18 affect the incidence of MI for either primary or secondary prevention but lead to a greater risk of 19 strokes 46. Given that HT status was one of the major determinants of angiotensinogen levels in 20 our study, further investigations are warranted to elucidate the specific interactions between HT, 21 angiotensinogen, and the development of hypertension and hypertension -related cardiovascular 22 diseases (e.g., stroke). 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 14 1 Study limitations : There are limitations to this study. First, we restricted our sample to 2 postmenopausal women as both sex hormone levels and cardiovascular risk differ between pre - 3 and postmenopausal women, and there were relatively few pre -menopausal women in MESA 4 (17%), which would limit analyses in this subset. Second, route and dosage of HT therapy was 5 not available for this analysis, both of which may impact on angiotensinogen levels. Third, sex 6 hormones in MESA were not measured using the current gold standard of mass spectrometry but 7 rather using radioimmunoassay and thus might be prone to measurement error. 8 9

10 Angiotensinogen and estrogen levels showed a positive correlation that differ in magnitude 11 between sexes, with the strongest association evidenced i n postmenopausal women on HT. We 12 observed evidence the HT indirectly increase systolic BP and odds of hypertension by increasing 13 angiotensinogen while lowering systolic BP via angiotensinogen independent direct effects. This 14 furthers our understanding of the implications of HT on angiotensinogen and BP in 15 postmenopausal women which may influence the utility of a new class of therapeutics designed 16 to directly reduce the production of angiotensinogen. 17 18 SOURCES OF FUNDING 19 This research was supported by an Institutional Development Award (IDeA) from the National 20 Institute of General Medical Sciences (NIGMS) of the National Institutes of Health under grant 21 number P20GM103451 and a Research Enhancement Award to PT from NIGMS under grant 22 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 15 number SC1GM139730. Add itional support for this research, including sample processing and 1 angiotensinogen measurements was provided by Ionis Pharmaceuticals, Inc. 2 3 This research was supported by contracts 75N92020D00001, HHSN268201500003I, N01 -HC-4 95159, 75N92020D00005, N01 -HC-95160, 75N92020D00002, N01 -HC-95161, 5 75N92020D00003, N01 -HC-95162, 75N92020D00006, N01 -HC-95163, 75N92020D00004, 6 N01-HC-95164, 75N92020D00007, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-7 95168 and N01 -HC-95169 from the National Heart, Lung, and Blood Institute, and by grants 8 UL1-TR-000040, UL1 -TR-001079, and UL1 -TR-001420 from the National Center for 9 Advancing Translational Sciences (NCATS). The authors thank the other investigators, the staff, 10 and the participants of the MESA study for their valuable contributions. A full list of 11 participating MESA investigators and institutions can be found at http://www.mesa -nhlbi.org. 12 This paper has been reviewed and approved by the MESA Publications and Presentations 13 Committee. 14 15 DISCLOSURES 16 ST is a co -inventor and receives royalties from patents owned by University of California San 17 Diego (UCSD) and is a co -founder and has an equity interest in Oxitope, LLC and Kleanthi 18 Diagnostics, LLC and has a dual appointment at UCSD and Ionis Pharmaceuticals. Although 19 these relationships have been identified for conflict -of-interest management based on the overall 20 scope of the project, the research findings included in this publication may not necessarily relate 21 to the interests of the above companies. The terms of this arrange ment have been reviewed and 22 approved by the University of California, San Diego in accordance with its conflict -of-interest 23 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 16 policies. ESM is employee of Ionis Pharmaceuticals. The other authors have reported that they 1 have no relationships relevant to this work to disclose. 2 3 DATA AVAILABILITY 4 Interested investigators may request access to the data by contacting the MESA data 5 coordinating center ([email protected]). 6

7 1. Martyniak A, Tomasik PJ. A New Perspective on the Renin -Angiotensin System. 8 Diagnostics. 2022;13(1):16. 9 2. Kahlon T, Carlisle S, Otero Mostacero D, Williams N, Trainor P, DeFi lippis AP. 10 Angiotensinogen. JACC: Heart Failure. 2022;10(10):699–713. 11 3. Cruz-López EO, Ye D, Wu C, Lu HS, Uijl E, Mirabito Colafella KM, Danser AHJ. 12 Angiotensinogen Suppression: A New Tool to Treat Cardiovascular and Renal Disease. 13 Hypertension. 2022;79(10):2115–2126. 14 4. Lonn EM, Yusuf S, Jha P, Montague TJ, Teo KK, Benedict CR, Pitt B. Emerging role of 15 angiotensin-converting enzyme inhibitors in cardiac and vascular protection. Circulation. 16 1994;90(4):2056–2069. 17 5. Yusuf S, Pepine CJ, Garces C, Pouleur H, Rousseau M, Salem D, Kostis J, Benedict C, 18 Bourassa M, Pitt B. Effect of enalapril on myocardial infarction and unstable angina in patients 19 with low ejection fractions. The Lancet. 1992;340(8829):1173–1178. 20 6. Pfeffer MA, Braunwald E, Moyé LA, Basta L , Brown EJ, Cuddy TE, Davis BR, Geltman 21 EM, Goldman S, Flaker GC, Klein M, Lamas GA, Packer M, Rouleau J, Rouleau JL, et al. Effect 22 of Captopril on Mortality and Morbidity in Patients with Left Ventricular Dysfunction after 23 Myocardial Infarction: Results o f the Survival and Ventricular Enlargement Trial. New England 24 Journal of Medicine. 1992;327(10):669–677. 25 7. Morgan ES, Tami Y, Hu K, Brambatti M, Mullick AE, Geary RS, Bakris GL, Tsimikas S. 26 Antisense Inhibition of Angiotensinogen With IONIS -AGT-LRx. JACC: Basic to Translational 27 Science. 2021;6(6):485–496. 28 8. Desai AS, Webb DJ, Taubel J, Casey S, Cheng Y, Robbie GJ, Foster D, Huang SA, Rhyee 29 S, Sweetser MT, Bakris GL. Zilebesiran, an RNA Interference Therapeutic Agent for 30 Hypertension. New England Journal of Medicine. 2023;389(3):228–238. 31 9. Dickson ME, Sigmund CD. Genetic Basis of Hypertension: Revisiting Angiotensinogen. 32 Hypertension. 2006;48(1):14–20. 33 10. Dickson ME, Zimmerman MB, Rahmouni K, Sigmund CD. The −20 and −217 Promoter 34 Variants Dominate Dif ferential Angiotensinogen Haplotype Regulation in Angiotensinogen -35 Expressing Cells. Hypertension. 2007;49(3):631–639. 36 11. Lu H, Cassis LA, Kooi CWV, Daugherty A. Structure and functions of angiotensinogen. 37 Hypertension Research. 2016;39(7):492–500. 38 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 17 12. White MC, Fleeman R, Arnold AC. Sex differences in the metabolic effects of the renin -1 angiotensin system. Biology of Sex Differences. 2019;10(1):31. 2 13. Trainor PJ, Brambatti M, Carlisle SM, Mullick AE, Shah SJ, Kahlon T, Mostacero DO, 3 Mousavi H, Morgan ES , Tami Y, Michos ED, Ouyang P, Tsimikas S, DeFilippis AP. Blood 4 Levels of Angiotensinogen and Hypertension in the Multi -Ethnic Study of Atherosclerosis 5 (MESA). Journal of the American College of Cardiology. 2023;81(13):1248–1259. 6 14. Kang AK, Duncan JA, C attran DC, Floras JS, Lai V, Scholey JW, Miller JA. Effect of oral 7 contraceptives on the renin angiotensin system and renal function. American Journal of 8 Physiology-Regulatory, Integrative and Comparative Physiology. 2001;280(3):R807–R813. 9 15. Harvey PJ, Wing LM, Savage J, Molloy D. The effects of different types and doses of 10 oestrogen replacement therapy on clinic and ambulatory blood pressure and the renin –11 angiotensin system in normotensive postmenopausal women: Journal of Hypertension . 12 1999;17(3):405–411. 13 16. Schunkert H, Danser AHJ, Hense H-W, Derkx FHM, Ku¨rzinger S, Riegger GAJ. Effects of 14 Estrogen Replacement Therapy on the Renin -Angiotensin System in Postmenopausal Women. 15 Circulation. 1997;95(1):39–45. 16 17. Harvey PJ, Morris BL, Miller JA, Floras J S. Estradiol Induces Discordant Angiotensin and 17 Blood Pressure Responses to Orthostasis in Healthy Postmenopausal Women. Hypertension. 18 2005;45(3):399–405. 19 18. Bild DE. Multi -Ethnic Study of Atherosclerosis: Objectives and Design. American Journal 20 of Epidemiology. 2002;156(9):871–881. 21 19. Blaha MJ, DeFilippis AP. Multi -Ethnic Study of Atherosclerosis (MESA). Journal of the 22 American College of Cardiology. 2021;77(25):3195–3216. 23 20. Ying W, Zhao D, Ouyang P, Subramanya V, Vaidya D, Ndumele CE, Sharma K, Shah SJ, 24 Heckbert SR, Lima JA, deFilippi CR, Budoff MJ, Post WS, Michos ED. Sex Hormones and 25 Change in N -Terminal Pro –B-Type Natriuretic Peptide Levels: The Multi -Ethnic Study of 26 Atherosclerosis. The Journal of Clinical Endocrinology & Metabolism . 2018;103(11):4304 –27 4314. 28 21. Zhao D, Guallar E, Ouyang P, Subramanya V, Vaidya D, Ndumele CE, Lima JA, Allison 29 MA, Shah SJ, Bertoni AG, Budoff MJ, Post WS, Michos ED. Endogenous Sex Hormones a nd 30 Incident Cardiovascular Disease in Post -Menopausal Women. Journal of the American College 31 of Cardiology. 2018;71(22):2555–2566. 32 22. Subramanya V, Zhao D, Ouyang P, Lima JA, Vaidya D, Ndumele CE, Bluemke DA, Shah 33 SJ, Guallar E, Nwabuo CC, Allison MA, Heckbert SR, Post WS, Michos ED. Sex hormone 34 levels and change in left ventricular structure among men and post -menopausal women: The 35 Multi-Ethnic Study of Atherosclerosis (MESA). Maturitas. 2018;108:37–44. 36 23. Subramanya V, Ambale-Venkatesh B, Ohyama Y, Zhao D, Nwabuo CC, Post WS, Guallar 37 E, Ouyang P, Shah SJ, Allison MA, Ndumele CE, Vaidya D, Bluemke DA, Lima JA, Michos 38 ED. Relation of Se x Hormone Levels With Prevalent and 10 -Year Change in Aortic 39 Distensibility Assessed by MRI: The Multi -Ethnic Study of Atherosclerosis. American Journal 40 of Hypertension. 2018;31(7):774–783. 41 24. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for 42 Statistical Computing. 2020. 43 25. Kim S. ppcor: An R Package for a Fast Calculation to Semi -partial Correlation Coefficients. 44 Communications for Statistical Applications and Methods. 2015;22(6):665–674. 45 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 18 26. Efron B, Hastie T. Computer age statistical inference: algorithms, evidence, and data 1 science. Student edition. Cambridge, United Kingdom New York, NY Port Melbourne, Australia 2 New Delhi, India Singapore: Cambridge University Press; 2021. 3 27. Lenth R. Emmeans: Estimated Marginal Means, Aka Least Square Means. 2020. 4 28. Wickham H, Francois R, Henry L, Muller K. dplyr: A Grammar of Data Manipulation (R 5 package). 2020. 6 29. Wickham. Ggplot2. New York, NY: Springer Science+Business Media, LLC; 2016. 7 30. Tingley, D., Yamamoto, T., Hirose, K., Imai, K., Keele, L. Mediation: R package for Causal 8 Mediation Analysis. 59(5):1–38. 9 31. Khan RS, Frishman WH. Zilebesiran: A Promising Antihypertensive Therapy Inhibiting 10 Angiotensinogen Synthesis. Cardiology in Review. 2024. 11 32. Liuzzo G, V olpe M. Silencing liver angiotensinogen synthesis as a novel approach to 12 hypertension management: promises and challenges. European Heart Journal . 13 2023;44(40):4217–4219. 14 33. Seely EW, Brosnihan KB, Jeunemaitre X, Okamura K, Williams GH, Hollenberg NK, 15 Herrington DM. Effects of conjugated oestrogen and droloxifene on the renin -angiotensin 16 system, blood pressure and renal blood flow in postmenopausal women: CEE, droloxifene and 17 the RAAS. Clinical Endocrinology. 2004;60(3):315–321. 18 34. Chang E, Perlman AJ. M ultiple Hormones Regulate Angiotensinogen Messenger 19 Ribonucleic Acid Levels in a Rat Hepatoma Cell Line. Endocrinology. 1987;121(2):513–519. 20 35. Voigt J, Koster H. Induction of Plasma Proangiotensin by Steroid Hormones in 21 Nephrectomized Rats. European Journal of Biochemistry. 1980;110(1):57–65. 22 36. AlSiraj Y, Woolley C, Thatcher SE, Cassis LA. Sex Differences and the Role of the Renin -23 Angiotensin System in Atherosclerosis and Abdominal Aortic Aneurysms. In: Sex Differences in 24 Cardiovascular Physiology and Pathophysiology. Elsevier; 2019:167–184. 25 37. Medina D, Mehay D, Arnold AC. Sex differences in cardiovascular actions of the renin –26 angiotensin system. Clinical Autonomic Research. 2020;30(5):393–408. 27 38. Writing Group For The Women’s Health Initiative In vestigators. Risks and Benefits of 28 Estrogen Plus Progestin in Healthy Postmenopausal Women: Principal Results From the 29 Women’s Health Initiative Randomized Controlled Trial. JAMA: The Journal of the American 30 Medical Association. 2002;288(3):321–333. 31 39. Chiu CL, Lujic S, Thornton C, O’Loughlin A, Makris A, Hennessy A, Lind JM. 32 Menopausal Hormone Therapy Is Associated with Having High Blood Pressure in 33 Postmenopausal Women: Observational Cohort Study. Jose PA, ed. PLoS ONE . 34 2012;7(7):e40260. 35 40. Madika A-L, MacDonald CJ, Fournier A, Mounier-Vehier C, Béraud G, Boutron-Ruault M-36 C. Menopausal hormone therapy and risk of incident hypertension: role of the route of estrogen 37 administration and progestogens in the E3N cohort. Menopause. 2021;28(11):1204–1208. 38 41. Miller VM, Naftolin F, Asthana S, Black DM, Brinton EA, Budoff MJ, Cedars MI, Dowling 39 NM, Gleason CE, Hodis HN, Jayachandran M, Kantarci K, Lobo RA, Manson JE, Pal L, et al. 40 The Kronos Early Estrogen Prevention Study (KEEPS): what have we learned? Menopause. 41 2019;26(9):1071–1084. 42 42. Affinito P, Palomba S, Bonifacio M, Fontana D, Izzo R, Trimarco B, Nappi C. Effects of 43 hormonal replacement therapy in postmenopausal hypertensive patients. Maturitas. 44 2001;40(1):75–83. 45 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 19 43. Kaya C, Dinçer Cengiz S, Cengiz B , Akgün G. The long -term effects of low -dose 17beta-1 estradiol and dydrogesterone hormone replacement therapy on 24 -h ambulatory blood pressure 2 in hypertensive postmenopausal women: a 1 -year randomized, prospective study. Climacteric: 3 The Journal of the International Menopause Society. 2006;9(6):437–445. 4 44. Kornhauser C, Malacara JM, Garay ME, Pérez -Luque EL. The effect of hormone 5 replacement therapy on blood pressure and cardiovascular risk factors in menopausal women 6 with moderate hypertension. Journal of Human Hypertension. 1997;11(7):405–411. 7 45. Hulley S, Grady D, Bush T, Furberg C, Herrington D, Riggs B, Vittinghoff E. Randomized 8 trial of estrogen plus progestin for secondary prevention of coronary heart disease in 9 postmenopausal women. Heart and Est rogen/progestin Replacement Study (HERS) Research 10 Group. JAMA. 1998;280(7):605–613. 11 46. Boardman HM, Hartley L, Eisinga A, Main C, Roqué I Figuls M, Bonfill Cosp X, Gabriel 12 Sanchez R, Knight B. Hormone therapy for preventing cardiovascular disease in post -13 menopausal women. Cochrane Heart Group, ed. Cochrane Database of Systematic Reviews . 14 2015;2015(8). 15 16 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 20 FIGURE LEGENDS 1 Graphical Abstract: Association between angiotensinogen and sex, and its clinical implications 2 in postmenopausal women on hormone therapy (HT). 3 4 Figure 1: Participant inclusion/exclusion in study. 5 HT, hormone therapy. 6 7 Figure 2: Distribution of circulating angiotensinogen levels by sex, race/ethnicity, and hormone 8 therapy (HT) status in postmenopausal (PM) women. Boxplot shows 25 th percentile, median, and 9 75th percentile. Table shows median and quartile values (25th and 75th percentiles). 10 11 Figure 3: Linear relationship between log -transformed estradiol and log -transformed 12 angiotensinogen levels stratified by sex and hormone therapy (HT) status with density curves in 13 the margin showing the distribution of each by gro up. Regression lines are shown based on an 14 interaction effect between sex and HT. Each dot represents a participant. 15 PM, postmenopausal women. 16 17 18 19 20 21 22 23 24 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 21 Table 1: Cohort characteristics of the 5,171 MESA participants stratified by sex and Hormone Therapy 1 (HT). Percentages presented are the number of participants in each row variable over the total number in 2 each column. 3 4 Characteristic Men (N = 2736) PM not on HT (N = 1675) PM on HT (N = 760) Mean age ± SD, years 62.6 ± 10.2 66.2 ± 9.2 62.6 ± 8.7 Race/Ethnicity, n (%) White 1037 (37.9) 487 (29.1) 410 (53.9) Black 737 (26.9) 524 (31.3) 177 (23.3) Chinese 351 (12.8) 248 (14.8) 56 (7.4) Hispanic 611 (22.3) 416 (24.8) 117 (15.4) Diabetes, n (%) 383 (14.0) 234 (14.0) 65 (8.6) Mean BMI ± SD, kg/m2 27.8 ± 4.4 28.9 ± 6.0 28.0 ± 5.9 Mean total cholesterol ± SD, mg/dL 188.0 ± 35.1 202.7 ± 36.5 199.4 ± 33.5 Mean HDL cholesterol ± SD, mg/dL 45.1 ± 11.8 54.7 ± 14.5 61.2 ± 16.8 Mean LDL cholesterol ± SD, mg/dL 116.5 ± 31.1 122.6 ± 32.7 110.3 ± 29.6 Mean systolic BP ± SD, mmHg 126.1 ± 19.4 131.3 ± 24.0 126.9 ± 22.0 Mean diastolic BP ± SD, mmHg 75.1 ± 9.4 69.6 ± 10.4 68.3 ± 9.9 Hypertension, n (%) 1185 (43.3) 873 (52.1) 374 (49.2) Any hypertension medication, n (%) 997 (36.4) 694 (41.4) 327 (43.1) Any lipid lowering medication, n (%) 454 (16.6) 325 (19.4) 143 (18.8) Current aspirin use, n (%) 787 (28.8) 391 (23.4) 199 (26.2) Smoking history, n (%) Never 1115 (40.9) 1066 (63.8) 365 (48.3) Former 1229 (45.1) 432 (25.9) 305 (40.3) Current 383 (14.0) 172 (10.3) 86 (11.4) Alcohol history, n (%) Never 279 (10.3) 601 (36.1) 168 (22.4) Former 748 (27.5) 360 (21.6) 156 (20.8) Current 1690 (62.2) 704 (42.3) 426 (56.8) Median hs-CRP [Q1, Q3], mg/L 1.390 [0.690, 3.060] 2.210 [0.978, 4.695] 3.515 [1.490, 6.945] Median Creatinine [Q1, Q3], mg/dL 1.02 [0.92, 1.22] 0.82 [0.72, 0.92] 0.82 [0.72, 0.92] Mean eGFR ± SD 75.7 ± 16.0 71.8 ± 15.8 71.6 ± 15.0 Total testosterone [Q1, Q3], nmol/L 14.12 [11.31, 17.67] 0.94 [0.62, 1.39] 0.76 [0.45, 1.18] Dehydroepiandrosterone [Q1, Q3], nmol/L 12.4 [9.1, 17.0] 10.9 [7.5, 15.2] 8.7 [6.0, 12.7] Sex Hormone Binding Globulin [Q1, Q3], nmol/L 41.2 [31.5, 53.0] 50.6 [37.2, 70.3] 103.0 [63.1, 163.2] Estradiol [Q1, Q3], nmol/L 0.114 [0.088, 0.139] 0.059 [0.040, 0.081] 0.231 [0.128, 0.356] 5 Note: data are shown as mean and standard deviation (SD), n (%), or median [Q1, Q3]. 6 BMI, body mass index; BP, blood pressure; HDL, high -density-lipoprotein; hs -CRP, high -sensitivity C -reactive 7 protein; LDL, low-density-lipoprotein; eGFR; Q1, first quartile (25th percentile); Q3, third quartile (75th percentile). 8 9 10 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 22 Table 2 : Un-adjusted model for predicting angiotensinogen from estradiol, in men, post -menopausal 1 women not on HT , and post-menopausal women on HT. 2 3 Note: Each analyte was log-transformed and then scaled to have mean zero and standard deviation 1. The 4 model coefficients are thus in terms of standard deviation units. Men were considered the reference level 5 for sex. 6 7 Variable Estimate (95% CI) p-value Intercept -0.443 (-0.471, -0.414) <0.0001 PM not on HT 0.585 (0.534, 0.636) <0.0001 PM on HT 1.716 (1.652, 1.781) <0.0001 Estradiol 0.195 (0.169, 0.220) <0.0001 R2 = 0.435 †Estradiol concentrations were log-transformed, centered, and standard deviation scaled. PM: postmenopausal women. HT: hormone therapy 8 9 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint 23 Table 3 . Direct, indirect, and total effects of HT on systolic blood pressure (BP), diastolic BP, and 1 hypertension with angiotensinogen as a mediator. 2 3 Note: Angiotensinogen levels have been log -transformed and standardized. Models to regress 4 angiotensinogen on HT were adjusted for age, race/ethnicity, BMI, smoking history, and use of any 5 hypertension medication. Models to regress blood pressure/ hypertension on HT were adjusted for 6 angiotensinogen levels, age, race/ethnicity, BMI, smoking history, and use of any hypertension 7 medication (only for blood pressure). 8 9 Dependent Variable Effects Estimate (95% CI) p-value Systolic BP Direct Effect HT  Systolic BP -1.19 (-3.36, 1.14) 0.306 Indirect Effect HT  Angiotensinogen  Systolic BP 1.24 (0.357, 2.48) 0.040 Total Effect 0.048 (-1.88, 1.87) 0.953 Diastolic BP Direct Effect HT  Diastolic BP -1.710 (-2.78, -0.55) < 0.0001 Indirect Effect HT  Angiotensinogen  Diastolic BP 0.472 (-0.16, 1.08) 0.126 Total Effect -1.234 (-2.13, -0.27) 0.014 Hypertension Direct Effect HT  Hypertension 1.016 (0.97, 1.06) 0.530 Indirect Effect HT  Angiotensinogen  Hypertension 1.065 (1.04, 1.09) < 0.001 Total Effect 1.082 (1.04, 1.13) < 0.001 BMI, body mass index; HT, hormone therapy; BP, blood pressure. 10 11 12 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 24, 2024. ; https://doi.org/10.1101/2024.03.22.24304764doi: medRxiv preprint