Age-specific effects of childhood body mass index on multiple sclerosis risk

Abstract

Objective: Higher body mass index (BMI) during early life is thought to be a causal risk factor for multiple sclerosis (MS). We used longitudinal Mendelian randomisation (MR) to determine whether there is a critical window during which BMI influences MS risk.

Methods: Summary statistics for childhood BMI (n ~ 28,000 children) and for MS susceptibility were obtained from recent large genome-wide association studies (GWAS) (n = 14,802 MS, 26,703 controls). We generated exposure instruments for BMI during four non-overlapping age epochs (< 3 months, 3 months-1.5 years, 2-5 years, and 7-8 years) and performed MR using the inverse variance weighted method with standard sensitivity analyses. Multivariable MR was used to account for effects mediated via later-life BMI.

Results: For all age epochs other than birth, genetically determined higher BMI was associated with an increased liability to MS: Birth [Odds Ratio (OR) 0.81, 95% Confidence Interval (CI) 0.50-1.31, Number of Single-Nucleotide Polymorphisms (N_SNPs) = 7, p = 0.39], Infancy (OR 1.18, 95% CI 1.04-1.33, N_SNPs = 18, p = 0.01), Early childhood (OR 1.31, 95% CI 1.03-1.66, N_SNPs = 4, p = 0.03), Later childhood (OR 1.34, 95% CI 1.08-1.66, N_SNPs = 4, p = 0.01). Multivariable MR suggested that these effects may be mediated by effects on adult BMI.

Conclusion: We provide evidence using MR that genetically determined higher BMI during early life is associated with increased MS risk. This effect may be driven by shared genetic architecture with later-life BMI.

Keywords: Body mass index; Causation; Epidemiology; Mendelian randomization; Multiple sclerosis.

Fig. 1

Forest plots showing the MR effect estimates during each epoch for the effect of BMI on MS susceptibility. Points represent beta estimates reflecting the predicted log odds ratio for MS risk per 1 unit increase in genetically determined standardised BMI at each time point. Error bars show 95% confidence intervals. The separate panels display the primary MR analytic method—the inverse variance weighted (IVW) method—followed by secondary sensitivity analyses