A multivariant recall-by-genotype study of the metabolomic signature of BMI

Abstract

Objective: This study estimated the effect of BMI on circulating metabolites in young adults using a recall-by-genotype study design.

Methods: A recall-by-genotype study was implemented in the Avon Longitudinal Study of Parents and Children. Samples from 756 participants were selected for untargeted metabolomics analysis based on low versus high genetic liability for higher BMI defined by a genetic risk score (GRS). Regression analyses were performed to investigate associations between BMI GRS group and relative abundance of 973 metabolites.

Results: After correction for multiple testing, 29 metabolites were associated with BMI GRS group. Bilirubin was among the most strongly associated metabolites, with reduced levels measured in individuals in the high-BMI GRS group (β = -0.32, 95% CI: -0.46 to -0.18, Benjamini-Hochberg adjusted p = 0.005). This study observed associations between BMI GRS group and the levels of several potentially diet-related metabolites, including hippurate, which had lower mean abundance in individuals in the high-BMI GRS group (β = -0.29, 95% CI: -0.44 to -0.15, Benjamini-Hochberg adjusted p = 0.008).

Conclusions: Together with existing literature, these results suggest that a genetic predisposition to higher BMI captures differences in metabolism leading to adiposity gain. In the absence of prospective data, separating these effects from the downstream consequences of weight gain is challenging.

FIGURE 1

Study overview. This study involves the first‐generation offspring in the Avon Longitudinal Study of Parents and Children (ALSPAC) multigenerational cohort, in which 14,541 pregnant women, resident in the South West of England, were recruited in the 1990s. First, we constructed a genetic risk score (GRS) for BMI for all first‐generation offspring. Under the recall‐by‐genotype study design, we recalled the plasma samples (collected at the age‐24‐years clinic) of individuals with a low‐ (yellow) or high‐ (blue) BMI GRS for further analysis. Then metabolites in those plasma samples were quantified by Metabolon. Finally, we performed statistical analysis to compare the metabolite levels between the two BMI GRS groups. Our results are relevant to understanding the role of metabolites both as intermediates on the pathway to BMI and from BMI to disease. [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 2

Overview of statistical analysis. “Raw data” is the original scale data normalized in terms of raw area counts (as supplied by Metabolon). Data were prepared for statistical analysis by first filtering samples and metabolites based on a series of quality metrics and then applying imputation and rescaling procedures as appropriate. GRS, genetic risk score; QC, quality control. [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 3

Mean differences in BMI between the high‐ and low‐BMI GRS groups. Error bars represent the 95% confidence interval of the mean difference in BMI. Sample size ranges from 108 (at age 31 months) to 743 (at age 24 years). Test results are given for a Student (two‐sample, two‐sided) t test. ***p < 0.001; **p < 0.01. For full results see Supporting Information Table S2

FIGURE 4

Volcano plot depicting the association between circulating metabolites and BMI genetic risk score (GRS) group. Points are colored by superpathway. Log₂ median fold change calculated as the ratio of median abundance (untransformed and unimputed) in the high‐BMI GRS group divided by median abundance in the low‐BMI GRS group. P values used to derive −log₁₀(p) are those from the linear regression analysis. All points above the dashed line have a Benjamini‐Hochberg adjusted p < 0.05. Solid gray lines indicate the density of points. A representative selection of metabolites of known identity are labeled. *Indicates a compound that has not been confirmed based on a standard. [Color figure can be viewed at wileyonlinelibrary.com]

FIGURE 5

Relationship between selected BMI genetic risk score (GRS) group–associated metabolites and measured BMI. Based on measured BMI at age‐24‐years clinic visit. Yellow, low‐BMI GRS group; blue, high‐BMI GRS group. β _overall is the measured BMI effect (CI_95% = 95% CI), extracted from multivariate linear model fitted in all individuals (metabolite ~ BMI + BMI.GRS.group + sex + age). Where there was evidence that including an interaction term improved the fit of the model, the measured BMI effect (adjusted for age and sex) is given for each BMI GRS group separately (β _high.BMI.GRS, β _low.BMI.GRS). In the plots, solid lines denote the predicted univariate within GRS group relationship between BMI and metabolite with a 95% CI denoted by shading