Longitudinal plasma metabolomics of aging and sex

Abstract

Understanding how metabolites are longitudinally influenced by age and sex could facilitate the identification of metabolomic profiles and trajectories that indicate disease risk. We investigated the metabolomics of age and sex using longitudinal plasma samples from the Wisconsin Registry for Alzheimer's Prevention (WRAP), a cohort of participants who were dementia free at enrollment. Metabolomic profiles were quantified for 2,344 fasting plasma samples among 1,212 participants, each with up to three study visits. Of 1,097 metabolites tested, 623 (56.8%) were associated with age and 695 (63.4%) with sex after correcting for multiple testing. Approximately twice as many metabolites were associated with age in stratified analyses of women versus men, and 68 metabolite trajectories significantly differed by sex, most notably including sphingolipids, which tended to increase in women and decrease in men with age. Using genome-wide genotyping, we also report the heritabilities of metabolites investigated, which ranged dramatically (0.2-99.2%); however, the median heritability of 36.2% suggests that many metabolites are highly influenced by a complex combination of genomic and environmental influences. These findings offer a more profound description of the aging process and may inform many new hypotheses regarding the role metabolites play in healthy and accelerated aging.

Keywords: aging; genomics; heritability; lipidomics; metabolomics; sex.

Figure 1

Adjusted effects of a 10-year increase in age on the top 100 metabolites most strongly influenced by age. Positive values indicate the amount a metabolite increased over 10 years, whereas negative values indicate the amount a metabolite decreased over 10 years. Black vertical lines indicate standard errors.

Figure 2

Adjusted effects of the top 100 metabolites most strongly influenced by sex. Positive values indicate that the metabolite was higher in women, whereas negative values indicate that the metabolite was higher in men. Black vertical lines indicate standard errors.

Figure 3

Pinwheel plot of metabolite heritabilities. Each bar indicates the heritability of the corresponding metabolite. Heritability estimates can also be found in Table S1.