Metabolomic and proteomic signatures of cardiorespiratory fitness for predicting all-cause mortality and non-communicable disease risk: a prospective study in the UK Biobank

Metabolomic and proteomic signatures of cardiorespiratory fitness for predicting all-cause mortality and non-communicable disease risk: a prospective study in the UK Biobank | medRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (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];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search Metabolomic and proteomic signatures of cardiorespiratory fitness for predicting all-cause mortality and non-communicable disease risk: a prospective study in the UK Biobank Sergio Miras-Moreno , Álvaro Torres-Martos , Jonatan R. Ruiz , Jennifer Carter , Carolina Abreu de Carvalho , View ORCID Profile Concepción M Aguilera , Carmen Piernas , View ORCID Profile Borja Martinez-Tellez doi: https://doi.org/10.1101/2025.08.07.25332939 Sergio Miras-Moreno 1 Department of Physical Education and Sports, Faculty of Sport Sciences, Sport and Health University Research Institute (iMUDS), University of Granada , Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site Álvaro Torres-Martos 2 Department of Biochemistry and Molecular Biology II, Faculty of Pharmacy; Center of Biomedical Research, Institute of Nutrition and Food Technology, University of Granada , Granada, Spain 3 Instituto de Investigación Biosanitaria ibs.GRANADA , Granada, Spain 4 CIBER Physiopathology of Obesity and Nutrition (CIBEROBN), Instituto de Salud Carlos III , Madrid, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jonatan R. Ruiz 1 Department of Physical Education and Sports, Faculty of Sport Sciences, Sport and Health University Research Institute (iMUDS), University of Granada , Spain 3 Instituto de Investigación Biosanitaria ibs.GRANADA , Granada, Spain 4 CIBER Physiopathology of Obesity and Nutrition (CIBEROBN), Instituto de Salud Carlos III , Madrid, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jennifer Carter 5 Nuffield Department of Population Health, University of Oxford , UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site Carolina Abreu de Carvalho 6 Postgraduate Program in Public Health; Public Health Department; Federal University of Maranhão , Brazil Find this author on Google Scholar Find this author on PubMed Search for this author on this site Concepción M Aguilera 2 Department of Biochemistry and Molecular Biology II, Faculty of Pharmacy; Center of Biomedical Research, Institute of Nutrition and Food Technology, University of Granada , Granada, Spain 3 Instituto de Investigación Biosanitaria ibs.GRANADA , Granada, Spain 4 CIBER Physiopathology of Obesity and Nutrition (CIBEROBN), Instituto de Salud Carlos III , Madrid, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Concepción M Aguilera Carmen Piernas 2 Department of Biochemistry and Molecular Biology II, Faculty of Pharmacy; Center of Biomedical Research, Institute of Nutrition and Food Technology, University of Granada , Granada, Spain 3 Instituto de Investigación Biosanitaria ibs.GRANADA , Granada, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: carmenpiernas{at}ugr.es borjammt{at}ual.es Borja Martinez-Tellez 4 CIBER Physiopathology of Obesity and Nutrition (CIBEROBN), Instituto de Salud Carlos III , Madrid, Spain 7 Department of Medicine, Division of Endocrinology, Einthoven Laboratory for Experimental Vascular Medicine, Leiden University Medical Center , Albinusdreef 2, 2333 ZA Leiden, the Netherlands 8 Department of Nursing, Physiotherapy and Medicine and SPORT Research Group (CTS-1024), CIBIS Research Center, University of Almería , Almería, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Borja Martinez-Tellez For correspondence: carmenpiernas{at}ugr.es borjammt{at}ual.es Abstract Full Text Info/History Metrics Data/Code Preview PDF Summary Cardiorespiratory fitness (CRF) is a strong predictor of mortality and non-communicable disease risk, but its underlying molecular mechanisms are poorly understood. In this study, we identified CRF-associated metabolomics (n=30,010) and proteomics (n=4,235) signatures in UK Biobank participants. These signatures were validated in an independent sample of UK participants with data on metabolomics (n=198,871) and proteomics (n=29,961) to investigate prospective associations with all-cause mortality and non-communicable diseases. Our findings reveal that higher CRF is characterized by downregulation of pathways related to inflammation, triglyceride metabolism, glycolysis, and vascular dysfunction, and upregulation of pathways related to cholesterol transport, apolipoprotein particle size, and cytoskeletal remodeling. Leveraging these insights, we developed two novel metabolic CRF signatures—one metabolomic and one proteomic—that robustly reflect CRF levels (R 2 : 0.49-0.60). Over an average of 9_years of follow-up, we observed 27,659 cases of all-cause mortality. Across the discovery and validation cohorts, we found that the metabolomic CRF signature was strongly associated with a 34–39% lower risk of all-cause mortality and markedly reduced risk of type 2 diabetes (89–91%), cardiovascular disease (35–39%), and colorectal cancer (32–54%). Additionally, the proteomic CRF signature was associated with a 17% lower risk of all-cause mortality, and with a 22–39% lower risk of type 2 diabetes and cardiovascular disease. Together, these findings suggest that circulating metabolites and proteins can capture the physiological imprint of CRF and may serve as indirect biomarkers for predicting mortality and non-communicable disease risk. Introduction Cardiorespiratory fitness (CRF), defined as the ability of the circulatory and respiratory systems to supply oxygen to skeletal muscles during sustained physical activity, is associated with overall health and functional capacity, and a strong predictor of mortality and non-communicable disease risk 1 . In a landmark study involving over 13,000 participants, Blair et al. were among the first to demonstrate that low CRF was a stronger predictor of all-cause mortality than traditional risk factors such as smoking, hypertension, or high cholesterol 2 , 3 . These findings have since been replicated and extended, establishing CRF as one of the most powerful modifiable risk factors for premature death 4 . Yet, despite its critical importance, the underlying molecular mechanisms of CRF remain poorly understood. This knowledge gap, coupled with the fact that CRF is rarely assessed in clinical or public health settings, highlights the urgent need for biomarkers to track this vital health indicator. Given the complex and multifactorial nature of CRF, it is unlikely that a single molecule can serve as an accurate predictor of fitness or related health outcomes. Instead, integrative multi-omic approaches combining large-scale metabolomic and proteomic data are needed to uncover molecular signatures that reflect the physiological adaptations associated with higher CRF. Such signatures may provide novel insights into the biological underpinnings of CRF and offer clinically useful proxies, especially in settings where direct or indirect measurements (e.g., VO_max testing) are not feasible. Moreover, it is critical to determine whether these molecular signatures are prospectively associated with mortality risk and major health outcomes. In the present study, we used data from the UK Biobank study 5 to identify molecular signatures associated with CRF and subsequently validated these omics signatures for predicting all-cause mortality and non-communicable diseases, including type 2 diabetes, cardiovascular disease, and several types of cancer. Supported by the unprecedented scale and depth of the UK Biobank, our study provides one of the most detailed omic characterizations of CRF to date. It identifies molecular signatures with strong predictive value for mortality and non-communicable disease risk and offers new insights into the biological mechanisms and potential clinical applications of CRF. Results Participant characteristics Our initial analyses examined the association between metabolomic and proteomic profiles and CRF in individuals with valid data for either metabolomic or proteomic measures, and CRF ( Fig. S1 ). These participants are referred to as the discovery cohort ( Table 1 ). The metabolomic discovery cohort included 30,010 individuals (54.3% female) with an average age of 56.0_±_8.1 years and a mean estimated VO_max of 28.8_±_6.7_ml/min/kg. The proteomic discovery cohort comprised 4,235 individuals (54.6% female), with an average age of 55.9_±_8.2 years and a mean estimated VO_max of 28.8_±_6.9_ml/min/kg. After developing the metabolomic and proteomic CRF signatures in the discovery cohorts, we validated these signatures in an independent group of participants from the UK Biobank—specifically, individuals with valid metabolomic or proteomic data but without available CRF data (named as validation cohort, Fig. S1 ). In this independent cohort, which was followed for an average of 9 years, validation was performed by assessing the association between the predicted metabolomic or proteomic CRF signatures and all-cause mortality and non-communicable disease risks. During this follow-up period, we documented 27,659 cases of all-cause mortality, 15,267 incident cases of type 2 diabetes, and 40,455 cases of cardiovascular disease. Regarding cancer outcomes, we observed 4,942 cases of colorectal cancer, including 3,341 cases of colon cancer. Additional cancer cases included 9,338 cases of prostate cancer, 8,120 of breast cancer, 3,260 of lung cancer, 1,282 of endometrial cancer, and 906 of bladder cancer. The metabolomic validation cohort included 198,871 participants (55.9% female) with an average age of 55.8_±_8.1 years, while the proteomic validation cohort comprised 29,961 participants (56.1% female) with an average age of 56.0_±_8.2 years (see Table S1 ). Metabolomic and proteomic profiling identifies widespread molecular associations with cardiorespiratory fitness In the metabolomic discovery cohort (n_=_30,010), we performed multivariable analyses to examine associations between the concentrations of 251 metabolites and CRF levels, adjusting for sex, age, and ethnicity. After False Discovery Rate correction, only 15 metabolites (5.9%) remained non-significantly associated with CRF levels, including acetate, histidine, omega-3 fatty acids, and total cholesterol ( Table S2 ). However, we found that 237 metabolites (94%) were significantly associated with CRF, of which 152 (60%) showed negative associations and 85 (34%) showed positive associations ( Table S2 ). The effect sizes (β) ranged from –0.205 to 0.274. Among the negatively associated metabolites, the percentage of triglycerides within various lipoprotein particles (VLDL, HDL, IDL, and LDL) and particle sizes were consistently and inversely related to CRF ( Fig. 1A ). Notably, glucose, lactate, pyruvate, and alanine levels were also negatively associated with CRF, though the effect sizes were all β_≤_–0.03 ( Table S2 ). Conversely, the percentages of cholesterol, phospholipids, and cholesteryl esters within different lipoprotein particles and sizes were consistently and positively associated with CRF ( Fig. 1A and Table S2 ). We then applied elastic net regression to derive a metabolomic CRF signature ( Table S3 ), retaining 221 metabolites in the model. The 30 most influential metabolites contributing to the metabolomic CRF signature are shown in Fig. 1B , which showed a moderate positive correlation with measured CRF (R² = 0.49; RMSE = 0.70; MAE = 0.53, Fig. 1C ), supporting the predictive value of the metabolic profile. Using the beta coefficients obtained in Fig. 1B , we performed enrichment analyses that revealed that several metabolic pathways were positively associated with CRF, such as ApoLP size and lipoprotein concentration (NES = +1.27 and +1.10, respectively; both p < 0.01, Fig. 1D ), while others were negatively associated, including phospholipids and inflammation (NES = –1.23 and –1.05, respectively; both p < 0.05, see Fig. 1D and Table S4 ). Download figure Open in new tab Figure 1: Multivariable associations between metabolomic molecules and cardiorespiratory fitness (CRF) levels in the discovery cohort (A) Circular bar plot showing individual metabolites significantly associated with CRF (FDR < 0.05) in the discovery cohort. Bars indicate the direction (positive or negative) and magnitude of association. Metabolites are grouped by biochemical class. (B) Top 30 metabolites selected via elastic net regularization as part of the metabolomic CRF signature. Bars represent the standardized beta coefficients from the final penalized regression model. (C) Scatter plots showing predicted scaled CRF based on metabolomic signatures on the x-axis and scaled CRF on the y-axis. The solid black line represents the line of good agreement between CRF and CRF prediction using metabolomic data. Values indicate the mean predictive performance across 10-fold cross-validation (k = 10). CRF, cardiorespiratory fitness; RMSE, root mean square error; MAE, mean absolute error. (D) Pathway enrichment analysis of metabolites contributing to the metabolomic CRF signature, using beta coefficients derived from the multivariable model (B). The most significantly enriched metabolic pathways are highlighted. Circle size indicates pathway size, and color denotes FDR-corrected significance. All analyses were performed in the discovery cohort (n = 30,010). Metabolite associations were adjusted for age, sex, and ethnicity. The CRF signature shown in panel B was derived using elastic net models and 10-fold cross-validation to select the hyperparameters that optimized predictive performance. Multiple testing correction was performed using FDR via the Benjamini–Hochberg procedure. Following a similar approach in the proteomic discovery cohort (n_=_4,235), we found that 885 out of 2,922 quantified proteins were not significantly associated with CRF levels. However, 2,037 proteins (69.7%) showed significant associations with CRF, of which 1,715 (58.7%) were negatively associated and 322 (11%) were positively associated ( Fig. 2A and Table S5 ). Effect sizes were modest, ranging from –0.11 to 0.08. Leptin (LEP) was the protein most strongly and negatively associated with CRF. Proteins negatively associated with CRF showed strong functional enrichment in immune, inflammatory, and vascular pathways ( Table S6 ). Over-representation analysis revealed significant enrichment in pathways that were downregulated such as cytokine–cytokine receptor interaction, PI3K-Akt signaling, MAPK signaling, chemokine signaling, and lipid and atherosclerosis ( Fig. 2B ). These were supported by network-based analyses highlighting downregulated proteins involved in platelet activation (e.g., VWF, FGA, VASP), cellular senescence (CDKN1A, FOXO3), and inflammatory signaling (IL1B, MMP9, CCL3) ( Fig. 2C ). In contrast, proteins positively associated with CRF showed weaker enrichment patterns. A smaller set of upregulated pathways emerged, including Ras signaling, regulation of actin cytoskeleton, focal adhesion, and ECM-receptor interaction ( Fig. 2D ). The 30 most influential proteins contributing to the protein CRF signature are shown in Fig. 2E and Table S7 , which showed a strong positive correlation with measured CRF (R² = 0.6; RMSE = 0.62; MAE = 0.46, Fig. 2F ), supporting the predictive value of the proteomic profile. Download figure Open in new tab Figure 2: Multivariable associations between proteomics molecules and cardiorespiratory fitness (CRF) levels (A) Volcano plot displaying the associations between individual plasma proteins and CRF in the discovery cohort. The x-axis shows the β-coefficients from linear models, and the y-axis shows the –log10 FDR-adjusted p-values. Proteins significantly positively associated with CRF are highlighted in red, and those negatively associated are shown in blue (FDR < 0.05). (B and D) Over-representation analysis (ORA) was conducted on proteins significantly associated with CRF. Panel (B) shows enriched GO pathways among proteins negatively associated with CRF (downregulated), while panel (D) displays enriched pathways among proteins positively associated with CRF (upregulated). Dot size reflects the number of proteins involved in each pathway (Count), color indicates the adjusted p-value, and the x-axis represents the GeneRatio—defined here as the number of significant proteins in the pathway divided by the total number of proteins analyzed that belong to that pathway. (C) Gene set enrichment analysis (GSEA) based on proteins associated with CRF using standardized coefficients was conducted. Enriched KEGG pathways are represented as colored clusters, and proteins are shown as nodes. Node size indicates the number of proteins of these clusters, while node color intensity reflects the beta coefficient. (E) Top 30 proteins included in the proteomic CRF signature, selected using elastic net regression with 10-fold cross-validation. Bars represent standardized beta coefficients from the final penalized regression model. (F) Scatter plots showing predicted scaled CRF based on proteomic signatures on the x-axis and scaled CRF on the y-axis. The solid black line represents the line of good agreement between CRF and CRF prediction using proteomic data. Values indicate the mean predictive performance across 10-fold cross-validation (k = 10). CRF, cardiorespiratory fitness; RMSE, root mean square error; MAE, mean absolute error. All analyses were performed in the discovery cohort (n = 4,235). Protein levels were standardized and adjusted for age, sex, and ethnicity. Multiple testing correction was performed using FDR via the Benjamini– Hochberg procedure. Metabolomic CRF signature predicts lower risk of mortality and chronic disease, while the proteomic CRF signature predicts all-cause mortality, type 2 diabetes and cardiovascular disease To investigate the association between CRF and the risk of all-cause mortality and non-communicable diseases, we performed the analyses in all participants with available CRF data, regardless of whether metabolomic or proteomic data were available (n = 78,236, see flow chart Fig. S1 ). As expected, CRF was strongly and progressively associated with a lower risk of all-cause mortality. Compared with individuals in the lowest CRF quintile (quintile 1: reference group), those in the highest quintile (quintile 5) had a 22% lower risk of motality (HR_=_0.78, 95% CI: 0.66–0.93; P_=_0.01; Fig. 3 ). Higher CRF was also associated with lower risk of type 2 diabetes (51%, HR_=_0.49, 95% CI: 0.38–0.63; P<0.001 Fig. 3 ), colorectal cancer (54%, HR_=_0.46, 95% CI: 0.31–0.67; P<0.001; Fig. 3 ), lung cancer (43%, HR_=_0.57, 95% CI: 0.35–0.94; P=0.03 Table S8 ), and colon cancer (59%, HR_=_0.41, 95% CI: 0.26–0.65; P<0.001; Table S8 ). No significant associations were observed for cardiovascular disease ( Fig. 3 ), prostate cancer, bladder cancer, breast cancer, or endometrial cancer ( Table S8 ). In the discovery cohort, the metabolomic CRF signature showed even stronger associations. Those in the highest quintile had a 44% lower risk of all-cause mortality (HR_=_0.56, 95% CI: 0.42–0.75; P<0.001; Fig. 3 ), and markedly lower risk of type 2 diabetes (89%, HR_=_0.11, 95% CI: 0.07–0.17; P<0.001; Fig. 3 ), cardiovascular disease (40%, HR_=_0.60, 95% CI: 0.48–0.75; P<0.001; Fig. 3 ), and lung cancer (72%, HR_=_0.28, 95% CI: 0.12–0.66; P<0.001; Table S9 ). No significant associations were found with colorectal, prostate, bladder, breast, endometrial, or colon cancer ( Table S9 ). Download figure Open in new tab Figure 3: Associations between cardiorespiratory fitness (CRF) and the metabolomic CRF signature with all-cause mortality, type 2 diabetes (T2D), cardiovascular disease (CVD), and colorectal cancer in the UK biobank population. Hazard ratios (HR) and 95% confidence intervals (CI) are presented across quintiles of estimated CRF and the CRF-related metabolomic signature. Results are shown separately for the CRF cohort (left panels), the discovery cohort (middle panels), and the validation cohort (right panel). CRF was categorized into quintiles as follows (n = 78,236): Q1 (n = 11,177), Q2 (n = 11,177), Q3 (n = 11,176), Q4 (n = 11,176), and Q5 (n = 11,176). The number of events (%) for each disease is shown in Table S8. These analyses were performed in all participants with available CRF data, regardless of whether metabolomic or proteomic data were available. For the metabolomic CRF signature in the discovery cohort (n = 30,010), quintiles were: Q1 (n = 6,002), Q2 (n = 6,002), Q3 (n = 6,002), Q4 (n = 6,002), and Q5 (n = 6,002). The number of events (%) for each disease is shown in Table S9. In the validation cohort (n = 198,735), the metabolomic CRF signature was similarly divided: Q1 (n = 39,747), Q2 (n = 39,747), Q3 (n = 39,747), Q4 (n = 39,747), and Q5 (n = 39,747). The number of events (%) for each disease is shown in Table S10. All Cox proportional hazards models were adjusted for sex, ethnicity, socio-economic status (Townsend index and education), country, lifestyle behaviours (smoking, alcohol intake, physical activity, fruit and vegetable and saturated fat consumption), BMI category, family history of chronic diseases (diabetes, cardiovascular diseases or cancer), medication use (cholesterol, blood pressure and others) and menopausal status where applicable. Age was used as the underlying time scale. P-values indicate the significance of associations compared to the reference group. These findings were replicated in the validation cohort (n_=_198,871), where the metabolomic CRF signature was strongly associated with a 34% lower risk of all-cause mortality (HR_=_0.66, 95% CI: 0.60– 0.73; P<0.001; Fig. 3 ). Participants in the highest quintile had a 91% lower risk of type 2 diabetes (HR_=_0.09, 95% CI: 0.08–0.11; P<0.001; Fig. 3 ), as well as significantly lower risk of cardiovascular disease (35%, HR_=_0.65, 95% CI: 0.59–0.72; P<0.001; Fig. 3 ), colorectal cancer (32%, HR_=_0.68, 95% CI: 0.55–0.86; P<0.001; Fig. 3 ), lung cancer (64%, HR_=_0.36, 95% CI: 0.27–0.47; P<0.001; Table S10 ), colon cancer (38%, HR_=_0.62, 95% CI: 0.47–0.80; P<0.001; Table S10 ), and bladder cancer (59%, HR_=_0.41, 95% CI: 0.24–0.70; P0.27; Table S10 ). The proteomic CRF signature was associated with a lower risk of type 2 diabetes in the discovery cohort (n = 4,235) (HR = 0.07, 95% CI: 0.02–0.23; P < 0.001; Fig. 4 ). No significant associations were found with all-cause mortality, cardiovascular disease ( Fig. 4 ), or any type of cancer ( Table S11 ). However, in the validation cohort (n = 29,961), the proteomic CRF signature was borderline associated with a 17% lower risk of all-cause mortality (HR = 0.83, 95% CI: 0.69–1.00; P = 0.05; Fig. 4 ), and was significantly associated with a 39% lower risk of type 2 diabetes (HR = 0.61, 95% CI: 0.47–0.80; P < 0.001; Fig. 4 ) and a 22% lower risk of cardiovascular disease (HR = 0.78, 95% CI: 0.66–0.92; P = 0.003; Fig. 4 ). No associations were found with any type of cancer ( Table S12 ). Download figure Open in new tab Figure 4: Associations between cardiorespiratory fitness (CRF) and the proteomic CRF signature with all-cause mortality, type 2 diabetes (T2D), cardiovascular disease (CVD), and colorectal cancer in the UK biobank population. Hazard ratios (HR) and 95% confidence intervals (CI) are presented across quintiles of the proteomic CRF signature. Results are shown separately for the discovery cohort (left panels) and the validation cohort (right panel). For the proteomic CRF signature in the discovery cohort (n = 4,235), quintiles were: Q1 (n = 847), Q2 (n = 847), Q3 (n = 847), Q4 (n = 847), and Q5 (n = 847). The number of events (%) for each disease is shown in Table S11. In the validation cohort (n = 29,961), the proteomic CRF signature was similarly divided: Q1 (n = 5,993), Q2 (n = 5,992), Q3 (n = 5,992), Q4 (n = 5,992), and Q5 (n = 5,992). The number of events (%) for each disease is shown in Table S12. All Cox proportional hazards models were adjusted for sex, ethnicity, socio-economic status (Townsend index and education), country, lifestyle behaviours (smoking, alcohol intake, physical activity, fruit and vegetable and saturated fat consumption), BMI category, family history of chronic diseases (diabetes, cardiovascular diseases or cancer), medication use (cholesterol, blood pressure and others) and menopausal status where applicable. Age was used as the underlying time scale. P-values indicate the significance of associations compared to the reference group. Discussion In this study, we identified robust metabolomic and proteomic signatures associated with CRF in a large sample of UK Biobank participants. In short, these signatures reflected that individuals with higher CRF showed downregulation of pathways related to inflammation, triglyceride metabolism, glycolysis, and vascular dysfunction, and upregulation of pathways related to cholesterol transport, apolipoprotein particle size, and cytoskeletal remodeling. Both signatures showed good predictive performance, with an R² of 0.49 for the metabolomic signature and R²=0.60 for the proteomic signature. Notably, the metabolomic CRF signature was strongly and consistently associated with lower risk of all-cause mortality and multiple non-communicable diseases, including type 2 diabetes, cardiovascular disease, and several cancers, whereas the proteomic CRF signature predicted all-cause mortality, type 2 diabetes and cardiovascular disease. Together, these findings suggest that circulating metabolites and proteins capture the physiological imprint of CRF and may serve as indirect biomarkers for predicting disease and mortality risk. Cardiorespiratory fitness is inversely associated with adverse metabolic and inflammatory signatures and positively linked to mitochondrial function In this study, we investigated the relationship between a broad panel of metabolites and proteins and CRF in 30,010 and 4,235 participants from the UK Biobank, respectively. Specifically, higher CRF was strongly associated with greater fat oxidation capacity and increased activity of the TCA cycle. It is well established that exercise modifies lipoprotein composition and size 6 , 7 . For instance, a single session of aerobic exercise reduces the triglyceride content within lipoprotein particles, suggesting that triglycerides are being used as an energy substrate during exercise 8 . Similarly, long-term exercise interventions have been shown to reduce atherogenic lipoproteins such as LDL, VLDL, and IDL, as well as other circulating lipid species 9 , highlighting a potential mechanism through which exercise lowers cardiovascular risk and contributes to longevity 7 . Our findings, from the largest study to date combining metabolomics and CRF, reinforce these observations. These results are consistent with previous studies showing that increased mitochondrial density and function are linked to improved cardiometabolic outcomes 10 . Mitochondria are the central organelles for energy production, hosting key metabolic processes such as β-oxidation, the TCA cycle, and lactate oxidation 11 . Furthermore, the proteomic CRF signature revealed downregulation of inflammatory, vascular, and senescence-related pathways, providing molecular evidence that higher CRF may counteract systemic inflammation and vascular aging. Taken together, these findings support the hypothesis that aerobic (endurance) exercise, through its well-documented ability to increase CRF, may extend lifespan by promoting profound metabolic adaptations. This idea is supported by compelling epidemiological evidence: a robust Finnish cohort study of elite athletes active between 1920 and 1965 showed that endurance athletes lived longer than both power athletes and controls, with significantly lower standardized mortality ratios from all causes and cardiovascular disease (43% and 44% lower, respectively) 12 . Further analyses revealed that these former elite endurance athletes also had a substantially reduced need for chronic disease medications, including a 30–60% lower probability of using drugs for hypertension and type 2 diabetes 13 . The molecular mechanisms underlying these associations are well characterized: aerobic exercise activates AMP-activated protein kinase (AMPK) signaling in skeletal muscle, leading to increased mitochondrial biogenesis 14 . This promotes more efficient utilization of fat and glucose as energy substrates, contributing to better health span and quality of life—and potentially explaining the increased longevity observed in endurance-trained individuals. This notion is further supported by the metabolomic and proteomic CRF signatures identified in the present study, which reflect enhanced mitochondrial function and substrate utilization. It is noteworthy that our proteomic CRF signature highlights similar biological pathways to those observed in the metabolomic CRF signature. Specifically, individuals with higher CRF showed enrichment in pathways related to energy metabolism, oxidative stress, inflammatory signaling, and cellular regulation, as well as mitochondrial function, angiogenesis, and immune homeostasis. Notably, several of these pathways were also reported in a previous study that identified a distinct proteomic CRF signature 15 . Interestingly, we observed overlap in both pathways and individual proteins, with six proteins—LEP, FABP4, ADM, TIMP4, FCGR2B, and ANGPTL2—consistently highlighted in both studies. Metabolomic and proteomic CRF signatures as markers of mortality and non-communicable diseases One of the most striking findings of our study is the strong and consistent association between the metabolomic CRF signature and reduced risk of all-cause mortality and multiple non-communicable diseases, including type 2 diabetes, cardiovascular disease, and several cancers. In both the discovery and validation cohorts—comprising over 230,000 participants—individuals in the highest quintile of the metabolomic CRF signature exhibited substantial reductions in mortality (34–39%), cardiovascular disease (35–39%), colorectal cancer (32%) and lung cancer (64–72%). Most notably, they showed up to a 91% lower risk of developing type 2 diabetes. This effect size exceeds those reported in previous studies, where individual metabolomic predictors—such as branched-chain and aromatic amino acids, lipids, and acylcarnitines—have been associated with type 2 diabetes risk, typically with HR ranging from +44% to +400%, depending on the biomarker and cohort. For instance, the Framingham Offspring Study reported up to a fivefold increase in type 2 diabetes risk among individuals with elevated amino acid levels 16 , whereas the EPIC-Potsdam cohort observed a 69% risk reduction in participants with favourable metabolic profiles 17 . Surprisingly, in the UK Biobank cohort, CRF levels were borderline associated with cardiovascular disease risk 18 . Detailed analyses further showed no significant associations with stroke, ischemic heart disease, or atrial fibrillation 18 . In contrast, our metabolomic CRF signature, derived from 30 circulating metabolites, likely reflects the integration of multiple protective metabolic pathways and appears to outperform previously reported biomarkers. These findings underscore the potential utility of the metabolomic CRF signature as a biomarker of physiological fitness, capable of capturing molecular features that reflect both metabolic efficiency and long-term disease risk. Notably, this signature showed stronger and more consistent associations across multiple outcomes than traditional CRF measures, suggesting added predictive value. Despite the strong correlation between measured CRF and the proteomic CRF signature, we found that the proteomic signature was associated with a lower risk of all-cause mortality, type 2 diabetes and cardiovascular disease, but not with any type of cancer. This is consistent with our proteomic analyses, which showed that IGFBP2 and IGFBP1 19 , 20 —two proteins positively associated with CRF—are known to enhance insulin sensitivity. These findings align with those of a previous study with a larger discovery sample (n = 14,000) 15 , compared to our cohort of 4,000 participants. Importantly, both the metabolomic and proteomic CRF signatures should be validated in other cohorts as well as in intervention studies where CRF is modified through exercise training, to determine whether these molecular profiles are responsive to changes in fitness. Limitations Several limitations should be considered when interpreting our findings. First, CRF was estimated using submaximal exercise tests, which rely on extrapolating VO_max based on heart rate responses. This approach, while practical for large-scale cohorts such as the UK Biobank, may introduce variability due to individual differences in heart rate responses to exercise and factors like medication use, stress, or hydration status. Second, different CRF assessment protocols were applied across UK Biobank subgroups (e.g., varying ramp rates and cycle ergometer loads), which could introduce measurement heterogeneity and limit comparability. Third, not all participants with CRF measurements had available metabolomic or proteomic data. These omics datasets are being generated progressively, and current analyses are limited to a subset of participants—particularly for proteomics, where only ∼4,000 individuals were included in the discovery cohort, potentially limiting statistical power and generalizability. Fourth, outcomes in the UK Biobank were ascertained through national health record linkage. While this method is objective and minimizes selective loss to follow-up, it is susceptible to delayed case detection, often capturing only advanced or severe cases. This may introduce bias toward disease severity and lead to underestimation of associations for milder or earlier disease stages. Despite these limitations, the large sample size, rigorous statistical modeling, and replication of findings in independent cohorts strengthen the validity and robustness of our conclusions. Conclusions We identified novel metabolomic and proteomic signatures strongly associated with CRF levels. The metabolomic CRF signature emerged as a robust predictor of all-cause mortality and multiple non-communicable diseases, including type 2 diabetes, cardiovascular disease, and several cancers. In addition, the proteomic CRF signature was associated with all-cause mortality, type 2 diabetes and cardiovascular disease. Together, these findings suggest that circulating metabolites and proteins can capture the physiological imprint of CRF and may serve as indirect biomarkers for predicting mortality and non-communicable disease risk. Funding The UK Biobank was supported by the Wellcome Trust, Medical Research Council, Department of Health, Scottish Government, the Northwest Regional Development Agency, the Welsh Assembly Government and the British Heart Foundation. This research has been conducted using the UK Biobank resource under application number 124330. SMM is supported by a postdoctoral fellowship (FPU19/01137) from the Spanish Ministry of Universities and by a short-term research stay grant from the University of Granada. The authors also acknowledge Instituto de Salud Carlos III for personal funding of A.T.M.: i-PFIS contract: IIS doctorates-company in health sciences and technologies of the Strategic Health Action (IFI22/00013). C.P. is funded by RYC2020-028818-I (MCIN/AEI/10.13039/501100011033 and “ESF Investing in your future”, Ministry of Science and Innovation, Spain), which also paid for the access to the data. BMT is supported by the Grant RYC2022-036473-I by MCIN/AEI/ 10.13039/501100011033 and by ESF. This research was designed, conducted, analysed, and interpreted by the authors entirely independently of the funding sources. Methods Study Design and Population We used data from the UK biobank study, a large-scale prospective cohort that recruited 502,134 participants aged 40-69 years between 2006 and 2010 across the UK (England, Scotland and Wales). At the baseline assessment, participants reported information on sociodemographic factors, medical history and health behaviours through self-completed touchscreen questionnaires. Participants also underwent an examination and a verbal interview with a nurse who collected relevant medical history, blood pressure, height and weight measures as well as blood, saliva and urine samples. In addition, the UK Biobank has established linkages with the NHS Hospital Episode Statistics (HES), cancer registries, and the death registries from the Office of National Statistics (ONS) 5 . The latest UK Biobank protocol is openly available elsewhere ( https://www.ukbiobank.ac.uk/wp-content/uploads/2025/01/Main-study-protocol.pdf ). The UK Biobank study was conducted according to the Declaration of Helsinki, and ethical approval was granted by the Northwest Multi-centre Research Ethics Committee (reference number 06/MRE08/65). At recruitment, all participants gave informed consent to participate and be followed up through data linkage. Metabolomic and proteomic assessment Metabolomic and proteomic profiling was undertaken in non-fasting blood samples collected at baseline (between 2006-2010). Frozen plasma was extracted from the blood samples collected in EDTA blood collection tubes and used to determine metabolomics using high-throughput nuclear magnetic resonance (NMR) spectroscopy (Nightingale Health Plc, Finland) and proteomic signatures (Olink Analysis Service, Sweden). Metabolomic data is currently available for a subset of around 275,000 UK Biobank participants. Technical details have been previously described 21 ( https://biobank.ndph.ox.ac.uk/showcase/ukb/docs/NMR_companion_phase2.pdf ). Targeted metabolomics was used to perform a quantitative analysis, which provided simultaneous quantification of metabolic biomarkers in a single assay expressed in molar concentration units. A total of 251 metabolites were included in the final analyses, including fatty acids and fatty acid composition and lipoproteins in 14 subclasses, amino acids, ketone bodies and glycolysis-related metabolites. Proteomics data was obtained through the UK Biobank Pharma Proteomics Project, which measured a random sample of 53,058 baseline participants who were representative of the overall UK Biobank population. Information on data processing and quality control can be found in the UK Biobank online resource ( https://biobank.ndph.ox.ac.uk/showcase/ukb/docs/Olink_proteomics_data.pdf ). Plasma samples were stored at −80 °C and shipped to Olink Analysis Service (Sweden) for analysis. Relative concentrations of 2923 unique proteins were determined using the antibody-based Olink ExploreTM Proximity Extension Assay, in combination with Next-Generation Sequencing. Measurements are expressed as normalized protein expression values (log2-transformed) 22 . All protein levels were standardized in the analyses. After excluding three proteins with a percentage of missing values of >30%, a total of 2920 proteins were included in the proteomic analyses. We excluded 11,680 participants with a percentage of missing values >10%. The remaining missing values were singly imputed using a quantile regression approach for the imputation of left-censored missing data with the imputeLCMD package in R. Exposure and Outcome definition The main exposure of interest comprised the metabolomic and proteomic signatures associated with cardiorespiratory fitness (CRF). CRF in the UK Biobank was derived via estimated VO_max from a submaximal cycle ergometer test. The protocol involved a 6-minute steady-state cycling exercise during which heart rate was continuously recorded. The estimated VO_max was derived using a multivariable equation that incorporated heart rate response, workload, age, sex, and body weight 18 . This non-invasive approach allowed for large-scale CRF assessment in a population-based setting, enabling the inclusion of CRF as a continuous trait in epidemiological analyses. The primary outcome of interest was all-cause mortality. Other outcomes of interest included the incidence of fatal and non-fatal cardiovascular disease (CVD), cancer and type 2 diabetes ascertained through linked HES episodes and cancer registries. ICD-10 codes were used to define the outcomes: 1) CVD: coronary heart disease (CHD; I20–I25 and K40-46, K49-50, K75), congestive heart failure or cardiomyopathy (CHF; I50, I50.1, 150.9, I11.0, I13.0, I13.2, I42, I43.1, I43.2, I43.8), and total stroke (I60–I64). Fatal CVD events were measured by I00–I25, I27–I88, and I95–I99; 2) Cancer: breast (C50), colon (C18), colorectal (C18-C20), endometrial (C54.1), bladder (C67), lung (C34) and prostate (C61); 3) type 2 diabetes (E11.0-E11.9), malnutrition-related diabetes (E12.0-E12.9), other specified diabetes (E13.0-E13.9) and unspecified diabetes (E14.0-E14.9). Covariate definition Further details on the derivation and classification of covariates can be found in Table S13 . Covariates were defined as follows: sex (male, female); ethnicity (grouped as White, Black, Asian, Other); Townsend index of deprivation, a composite measure of deprivation based on unemployment, non-car ownership, non-home ownership and household overcrowding, categorized in quintiles 1-5 (higher quintiles indicating more deprivation); education (grouped as higher degree, any school degree, vocational qualification, unknown); country (England, Scotland, Wales); smoking status (grouped as never, current, previous, unknown); alcohol consumption (grouped as none, occasional, moderate, heavy, unknown); physical activity classified as meeting recommendations (≥150_min of moderate-intensity activity and/or ≥75_min of vigorous activity per week), not meeting recommendations, unknown; dietary quality proxied via meeting fruit and vegetable recommendations (classified as meeting five portions a day, not meeting five portions a day, unknown); body mass index (BMI) categorised as <18.5, 18.5-24.9, 25-29.9, 30-34.9, ≥35_kg/m 2 ; family history of CVD, cancer, diabetes (grouped as yes, no, unknown); medication use (grouped as none, lipid-lowering, blood pressure medication); menopause grouped as yes, no, unknown. Statistical and bioinformatic analysis Participant characteristics were studied through descriptive analysis. Categorical variables are reported as total numbers and percentages, whereas continuous variables are reported as means, standard deviations, medians and interquartile ranges. To compare categorical and continuous variables, Pearson’s Chi-square and Wilcoxon rank sum tests were performed, respectively. Retrospective analyses were conducted separately for the metabolomic and proteomic datasets using both univariable and multivariable approaches. Iterative linear models were applied to assess the individual associations between each metabolite and CRF. Nominal p-values were corrected for multiple testing using the false discovery rate according to the Benjamini–Hochberg procedure. Subsequently, a regularized generalized linear model was used to derive the CRF-associated omic signatures. This multivariable approach, based on elastic net regularization—which combines L1 and L2 penalties—requires a statistical framework to optimize feature selection and predictive performance. To this end, 10-fold cross-validation was employed to determine the hyperparameters that minimized the root mean squared error (RMSE). The final model was then fitted using the entire dataset and the selected hyperparameters. The output from this model was scaled and referred to as the CRF-associated signature, expressed as z-scores. In both univariable and multivariable analyses, models were adjusted for age, sex, and ethnicity. All analyses were performed in R using the caret and glmnet packages. For the prospective analysis of all-cause mortality and health outcomes, we estimated hazard ratios (HR) with 95% CIs for the CRF-associated metabolomic and proteomic signatures (with scores modelled as quintiles) and the outcomes using multivariable Cox proportional hazard models with age (years) as the underlying time covariate. Person-time of follow-up was calculated from the age at the baseline assessment visit, until the age at which participants were censored, using the date when the outcome occurred. For death, the latest available data was 31 May 2024 for England and Wales; 31 December 2023 for Scotland. For HES-related outcomes, the latest available data was 31 October 2022 for England, 31 May 2022 for Wales, and 31 August 2022 for Scotland. For cancer, the latest available data was 31 December 2020 for England, 31 December 2016 for Wales, and 30 November 2021 for Scotland. Models were sequentially adjusted starting from a minimally adjusted model with age, sex and ethnicity; socio-economic status (Townsend index and education) and country; health behaviours (smoking, alcohol, diet, BMI group); relevant medical history (family history, menopausal status, medication use). The proportional hazards assumption using Schoenfeld residuals’ plots for the exposures was not violated for the variables of interest in the adjusted model (P >_0.05). Heterogeneity in the associations between omics signatures and outcomes was explored for sex, sex, age group and BMI group. The survival package in R was used and a two-sided p value of <_0.05 was defined as statistically significant. Data Availability All data produced are available online at UK biobank Footnotes ↵ * These authors shared first authorship References 1. ↵ Ross R , Blair SN , Arena R , Church TS , Després JP , Franklin BA , Haskell WL , Kaminsky LA , Levine BD , Lavie CJ , Myers J , Niebauer J , Sallis R , Sawada SS , Sui X , Wisløff U . Importance of Assessing Cardiorespiratory Fitness in Clinical Practice: A Case for Fitness as a Clinical Vital Sign: A Scientific Statement from the American Heart Association . Circulation 2016 ; 134 : e653 – e699 . OpenUrl Abstract / FREE Full Text 2. ↵ Blair SN , Kampert JB , Kohl HW , Barlow CE , Macera CA , Paffenbarger RS , Gibbons LW . Influences of cardiorespiratory fitness and other precursors on cardiovascular disease and all-cause mortality in men and women . JAMA 1996 ; 276 : 205 – 210 . OpenUrl CrossRef PubMed Web of Science 3. ↵ Blair SN , Kohl HW , Paffenbarger RS , Clark DG , Cooper KH , Gibbons LW . Physical fitness and all-cause mortality. A prospective study of healthy men and women . JAMA 1989 ; 262 : 2395 – 2401 . OpenUrl CrossRef PubMed Web of Science 4. ↵ Laukkanen JA , Isiozor NM , Kunutsor SK . Objectively Assessed Cardiorespiratory Fitness and All-Cause Mortality Risk: An Updated Meta-analysis of 37 Cohort Studies Involving 2,258,029 Participants . Mayo Clin Proc 2022 ; 97 : 1054 – 1073 . OpenUrl CrossRef PubMed 5. ↵ Sudlow C , Gallacher J , Allen N , Beral V , Burton P , Danesh J , Downey P , Elliott P , Green J , Landray M , Liu B , Matthews P , Ong G , Pell J , Silman A , Young A , Sprosen T , Peakman T , Collins R . UK Biobank: An Open Access Resource for Identifying the Causes of a Wide Range of Complex Diseases of Middle and Old Age . PLoS Med 2015 ; 12 . 6. ↵ Smart NA , Downes D , Touw T van der , Hada S , Dieberg G , Pearson MJ , Wolden M , King N , Goodman SPJ . The Effect of Exercise Training on Blood Lipids: A Systematic Review and Meta-analysis . Sports Medicine . 7. ↵ Sarzynski MA , Burton J , Rankinen T , Blair SN , Church TS , Després JP , Hagberg JM , Landers-Ramos R , Leon AS , Mikus CR , Rao DC , Seip RL , Skinner JS , Slentz CA , Thompson PD , Wilund KR , Kraus WE , Bouchard C . The effects of exercise on the lipoprotein subclass profile: A meta-analysis of 10 interventions . Atherosclerosis 2015 ; 243 : 364 – 372 . OpenUrl PubMed 8. ↵ Søndergaard E , Poulsen MK , Jensen MD , Nielsen S . Acute changes in lipoprotein subclasses during exercise . Metabolism 2014 ; 63 : 61 – 68 . OpenUrl PubMed 9. ↵ Zhang Y , Zhang Z , Martinez-Tellez B , Di X , Kindt A , Kohler I , Osuna-Prieto FJ , Clark C , Drouin N , Harms A , Hankemeier T , Ruiz JR , Jurado-Fasoli L . Moderate-intensity exercise training uniquely modulates circulating lipid species beyond classical lipid levels in humans . EBioMedicine 2025 ; 118 . 10. ↵ Mølmen KS , Almquist NW , Skattebo Ø . Effects of Exercise Training on Mitochondrial and Capillary Growth in Human Skeletal Muscle: A Systematic Review and Meta-Regression . Sports Medicine . 11. ↵ Jacobs RA , Meinild A-K , Nordsborg NB , Lundby C . Lactate oxidation in human skeletal muscle mitochondria . Am J Physiol Endocrinol Metab 2013 ; 304 : 686 – 694 . OpenUrl 12. ↵ Kujala UM , Tikkanen HO , Sarna S , Pukkala E , Kaprio J , Koskenvuo M . Disease-specific mortality among elite athletes . JAMA 2001 ; 285 : 44 – 45 . OpenUrl CrossRef PubMed Web of Science 13. ↵ Laine MK , Eriksson JG , Kujala UM , Wasenius NS , Kaprio J , Bäckmand HM , Peltonen M , Mertsalmi TH , Sarna S . A former career as a male elite athlete – Does it protect against type 2 diabetes in later life? Diabetologia 2014 ; 57 : 270 – 274 . OpenUrl CrossRef PubMed 14. ↵ Hwang JH , Kim KM , Oh HT , Yoo GD , Jeong MG , Lee H , Park J , Jeong K , Kim YK , Ko YG , Hwang ES , Hong JH . TAZ links exercise to mitochondrial biogenesis via mitochondrial transcription factor A . Nat Commun 2022 ; 13 . 15. ↵ Perry AS , Farber-Eger E , Gonzales T , Tanaka T , Robbins JM , Murthy VL , Stolze LK , Zhao S , Huang S , Colangelo LA , Deng S , Hou L , Lloyd-Jones DM , Walker KA , Ferrucci L , Watts EL , Barber JL , Rao P , Mi MY , Gabriel KP , Hornikel B , Sidney S , Houstis N , Lewis GD , Liu GY , Thyagarajan B , Khan SS , Choi B , Washko G , Kalhan R , Wareham N , Bouchard C , Sarzynski MA , Gerszten RE , Brage S , Wells QS , Nayor M , Shah R V . Proteomic analysis of cardiorespiratory fitness for prediction of mortality and multisystem disease risks . Nat Med 2024 ; 30 : 1711 – 1721 . OpenUrl PubMed 16. ↵ Wang TJ , Larson MG , Vasan RS , Cheng S , Rhee EP , McCabe E , Lewis GD , Fox CS , Jacques PF , Fernandez C , O’Donnell CJ , Carr SA , Mootha VK , Florez JC , Souza A , Melander O , Clish CB , Gerszten RE . Metabolite profiles and the risk of developing diabetes . Nat Med 2011 ; 17 : 448 – 453 . OpenUrl CrossRef PubMed Web of Science 17. ↵ Floegel A , Stefan N , Yu Z , Mühlenbruch K , Drogan D , Joost HG , Fritsche A , Häring HU , Angelis MH De , Peters A , Roden M , Prehn C , Wang-Sattler R , Illig T , Schulze MB , Adamski J , Boeing H , Pischon T . Identification of serum metabolites associated with risk of type 2 diabetes using a targeted metabolomic approach . Diabetes 2013 ; 62 : 639 – 648 . OpenUrl Abstract / FREE Full Text 18. ↵ Gonzales TI , Westgate K , Strain T , Hollidge S , Jeon J , Christensen DL , Jensen J , Wareham NJ , Brage S . Cardiorespiratory fitness assessment using risk-stratified exercise testing and dose–response relationships with disease outcomes . Sci Rep 2021 ; 11 . 19. ↵ Haywood NJ , Cordell PA , Tang KY , Makova N , Yuldasheva NY , Imrie H , Viswambharan H , Bruns AF , Cubbon RM , Kearney MT , Wheatcroft SB . Insulin-Like Growth Factor Binding Protein 1 Could Improve Glucose Regulation and Insulin Sensitivity Through Its RGD Domain . Diabetes 2017 ; 66 : 287 – 299 . OpenUrl Abstract / FREE Full Text 20. ↵ Hedbacker K , Birsoy K , Wysocki RW , Asilmaz E , Ahima RS , Farooqi IS , Friedman JM . Antidiabetic Effects of IGFBP2, a Leptin-Regulated Gene . Cell Metab 2010 ; 11 : 11 – 22 . OpenUrl CrossRef PubMed Web of Science 21. ↵ Würtz P , Kangas AJ , Soininen P , Lawlor DA , Davey Smith G , Ala-Korpela M . Quantitative Serum Nuclear Magnetic Resonance Metabolomics in Large-Scale Epidemiology: A Primer on –Omic Technologies . Am J Epidemiol 2017 ; 186 : 1084 – 1096 . OpenUrl CrossRef PubMed 22. ↵ Sun BB , Chiou J , Traylor M , Benner C , Hsu YH , Richardson TG , Surendran P , Mahajan A , Robins C , Vasquez-Grinnell SG , Hou L , Kvikstad EM , Burren OS , Davitte J , Ferber KL , Gillies CE , Hedman ÅK , Hu S , Lin T , Mikkilineni R , Pendergrass RK , Pickering C , Prins B , Baird D , Chen CY , Ward LD , Deaton AM , Welsh S , Willis CM , Lehner N , Arnold M , Wörheide MA , Suhre K , Kastenmüller G , Sethi A , Cule M , Raj A , Kang HM , Burkitt-Gray L , Melamud E , Black MH , Fauman EB , Howson JMM , Kang HM , McCarthy MI , Nioi P , Petrovski S , Scott RA , Smith EN , Szalma S , Waterworth DM , Mitnaul LJ , Szustakowski JD , Gibson BW , Miller MR , Whelan CD . Plasma proteomic associations with genetics and health in the UK Biobank . Nature 2023 ; 622 : 329 – 338 . OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted August 11, 2025. Download PDF Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Metabolomic and proteomic signatures of cardiorespiratory fitness for predicting all-cause mortality and non-communicable disease risk: a prospective study in the UK Biobank Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Metabolomic and proteomic signatures of cardiorespiratory fitness for predicting all-cause mortality and non-communicable disease risk: a prospective study in the UK Biobank Sergio Miras-Moreno , Álvaro Torres-Martos , Jonatan R. Ruiz , Jennifer Carter , Carolina Abreu de Carvalho , Concepción M Aguilera , Carmen Piernas , Borja Martinez-Tellez medRxiv 2025.08.07.25332939; doi: https://doi.org/10.1101/2025.08.07.25332939 Share This Article: Copy Citation Tools Metabolomic and proteomic signatures of cardiorespiratory fitness for predicting all-cause mortality and non-communicable disease risk: a prospective study in the UK Biobank Sergio Miras-Moreno , Álvaro Torres-Martos , Jonatan R. Ruiz , Jennifer Carter , Carolina Abreu de Carvalho , Concepción M Aguilera , Carmen Piernas , Borja Martinez-Tellez medRxiv 2025.08.07.25332939; doi: https://doi.org/10.1101/2025.08.07.25332939 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Epidemiology Subject Areas All Articles Addiction Medicine (568) Allergy and Immunology (863) Anesthesia (300) Cardiovascular Medicine (4435) Dentistry and Oral Medicine (444) Dermatology (382) Emergency Medicine (608) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1509) Epidemiology (15228) Forensic Medicine (30) Gastroenterology (1124) Genetic and Genomic Medicine (6599) Geriatric Medicine (668) Health Economics (997) Health Informatics (4536) Health Policy (1368) Health Systems and Quality Improvement (1613) Hematology (540) HIV/AIDS (1264) Infectious Diseases (except HIV/AIDS) (15916) Intensive Care and Critical Care Medicine (1103) Medical Education (623) Medical Ethics (146) Nephrology (667) Neurology (6599) Nursing (346) Nutrition (998) Obstetrics and Gynecology (1144) Occupational and Environmental Health (957) Oncology (3332) Ophthalmology (974) Orthopedics (369) Otolaryngology (420) Pain Medicine (436) Palliative Medicine (130) Pathology (663) Pediatrics (1693) Pharmacology and Therapeutics (691) Primary Care Research (711) Psychiatry and Clinical Psychology (5447) Public and Global Health (9231) Radiology and Imaging (2198) Rehabilitation Medicine and Physical Therapy (1370) Respiratory Medicine (1196) Rheumatology (593) Sexual and Reproductive Health (712) Sports Medicine (530) Surgery (712) Toxicology (99) Transplantation (289) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a005ba203d8a8e2e',t:'MTc3OTU1Njc4Mg=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();