Abstract P203: Healthy Lifestyle Plasma Metabolite Profile and Risk of Mortality in US Prospective Cohort Studies
Anne-Julie Tessier,Fenglei Wang,Liming Liang,Clemens Wittenbecher,Danielle E Haslam,A. Heather Eliassen,Qi Sun,Deirdre K Tobias,Jun Li,Oana Zeleznik,Alberto Ascherio,Meir J Stampfer,Francine Grodstein,Kathryn M Rexrode,Miguel A Martinez-Gonzalez,Clary Clish,Jorge E Chavarro,Frank B Hu,Marta Guasch
DOI: https://doi.org/10.1161/circ.147.suppl_1.p203
IF: 37.8
2023-02-28
Circulation
Abstract:Background: A healthy lifestyle is associated with a lower risk of premature death. Metabolic pathways of a healthy lifestyle and their association with mortality remain to be understood. This study aimed to identify the metabolomic profile of a healthy lifestyle score and examine its prospective association with all-cause and cause-specific mortality, including death from cardiovascular disease (CVD) and cancer. Methods: The population included 12,146 participants from the Nurses’ Health Study (NHS), NHS II and Health Professionals Follow-Up Study (HPFS)(83% women, 97% white, aged 55±9y). Plasma metabolites were profiled using high-throughput liquid chromatography mass-spectrometry at baseline (NHS:1989-1990; NHSII:1996-1999; HPFS:1993-1995). The healthy lifestyle score was computed by summing the total number of healthy lifestyle factors participants adhered to from validated questionnaires at baseline: healthy diet (Alternative Healthy Eating Index, upper 40%), moderate alcohol intake (women: 5-15 g/d; men: 5-30 g/d), moderate-to-vigorous physical activity (≥30min/d), never smoking and normal BMI (18.5-24.9kg/m 2 ). Deaths were ascertained with death certificates and medical records. The metabolite profile was identified using elastic net regressions with train test validation split (70-30%). Metabolic pathways were determined using Metabolite Set Enrichment Analysis (MSEA). Multivariable-adjusted Cox proportional hazards regressions were used to estimate hazard ratios and 95% confidence intervals (HR[CI]) per unit of score of the healthy lifestyle metabolite profile with mortality risk. Results: The identified profile included 88 metabolites and correlated with the healthy lifestyle score (Pearson r=0.43-0.44; p<0.001). Triglyceride and diglyceride metabolite sets were inversely associated with the healthy lifestyle score, whereas cholesteryl ester and phosphatidylcholine plasmalogen sets were directly associated (p<0.001). Among individual lifestyle factors, the profile was most strongly correlated with normal BMI (r pb =0.43; p<0.001). Over 32y of follow-up, there were 3,851 deaths, including 749 deaths from CVD and 994 from cancer. Participants with a higher healthy lifestyle metabolite profile score had lower risk of all-cause (HR=0.79[0.73, 0.85]) and CVD mortality (HR=0.77[0.58, 0.95]), but not cancer (HR=0.91[0.77, 1.05]). Significant associations persisted after further adjustment for the healthy lifestyle score. Conclusions: In US adults, we identified a metabolite profile related to a healthy lifestyle largely reflecting lipid metabolism pathways. A higher metabolite score was associated with lower subsequent all-cause mortality risk, specifically from CVD. Findings provide novel insights into potential metabolic pathways underlying the association between a healthy lifestyle and lower premature mortality.
cardiac & cardiovascular systems,peripheral vascular disease