Modifiable pathways for longevity: a Mendelian randomization analysis
Xiaolin Ni,Huabin Su,Yuan Lv,Rongqiao Li,Lei Liu,Yan Zhu,Ze Yang,Caiyou Hu
DOI: https://doi.org/10.1016/j.clnu.2023.04.026
IF: 7.643
2023-05-07
Clinical Nutrition
Abstract:Background A variety of factors, including diet and lifestyle, obesity, physiology, metabolism, hormone levels, psychology, and inflammation, have been associated with longevity. The specific influences of these factors, however, are poorly understood. Here, possible causal relationships between putative modifiable risk factors and longevity are investigated. Methods A random effects model was used to investigate the association between 25 putative risk factors and longevity. The study population comprised 11 262 long-lived subjects ( ≥ 90 years old, including 3484 individuals ≥ 99 years old) and 25 483 controls ( ≤60 years old), all of European ancestry. The data were obtained from the UK Biobank database. Genetic variations were used as instruments in two-sample Mendelian randomization to reduce bias. The odds ratios for genetically predicted SD unit increases were calculated for each putative risk factor. Egger regression was used to determine possible violations of the Mendelian randomization model. Results Thirteen potential risk factors showed significant associations with longevity (≥90th) after correction for multiple testing. These included smoking initiation (OR:1.606; CI: 1.112-2.319) and educational attainment (OR:2.538, CI: 1.685-3.823) in the diet and lifestyle category, systolic and diastolic blood pressure (OR per SD increase: 0.518; CI: 0.438-0.614 for SBP and 0.620; CI 0.514-0.748 for DBP) and venous thromboembolism (OR:0.002; CI: 0.000-0.047) in the physiology category, obesity (OR: 0.874; CI: 0.796-0.960), BMI (OR per 1-SD increase: 0.691; CI: 0.628-0.760), and body size at age 10 (OR per 1-SD increase:0.728; CI: 0.595-0.890) in the obesity category, type 2 diabetes (T2D) (OR:0.854; CI: 0.816-0.894), LDL cholesterol (OR per 1-SD increase: 0.743; CI: 0.668-0.826), HDL cholesterol (OR per 1-SD increase: 1.243; CI: 1.112-1.390), total cholesterol (TC) (OR per 1-SD increase: 0.786; CI: 0.702-0.881), and triglycerides (TG) (OR per 1-SD increase: 0.865; CI: 0.749-0.998) in the metabolism category. Both longevity (≥90th) and super-longevity (≥99th), smoking initiation, body size at age 10, BMI, obesity, DBP, SBP, T2D, HDL, LDL, and TC were consistently associated with outcomes. The examination of underlying pathways found that BMI indirectly affected longevity through three pathways, namely, SBP, plasma lipids (HDL/TC/LDL), and T2D (p<0.05). Conclusion BMI was found to significantly affect longevity through SBP, plasma lipid (HDL/TC/LDL), and T2D. Future strategies should focus on modifying BMI to improve health and longevity.
nutrition & dietetics