Background: Higher childhood body mass index (BMI) has been associated with an increased risk of multiple sclerosis (MS).
Objective: To evaluate whether childhood BMI has a causal influence on MS, and whether this putative effect is independent from early adult obesity and pubertal timing.
Methods: We performed Mendelian randomization (MR) using summary genetic data on 14,802 MS cases and 26,703 controls. Large-scale genome-wide association studies provided estimates for BMI in childhood (n = 47,541) and adulthood (n = 322,154). In multivariable MR, we examined the direct effects of each timepoint and further adjusted for age at puberty. Findings were replicated using the UK Biobank (n = 453,169).
Results: Higher genetically predicted childhood BMI was associated with increased odds of MS (odds ratio (OR) = 1.26/SD BMI increase, 95% confidence interval (CI): 1.07-1.50). However, there was little evidence of a direct effect after adjusting for adult BMI (OR = 1.03, 95% CI: 0.70-1.53). Conversely, the effect of adult BMI persisted independent of childhood BMI (OR = 1.43; 95% CI: 1.01-2.03). The addition of age at puberty did not alter the findings. UK Biobank analyses showed consistent results. Sensitivity analyses provided no evidence of pleiotropy.
Conclusion: Genetic evidence supports an association between childhood obesity and MS susceptibility, mediated by persistence of obesity into early adulthood but independent of pubertal timing.
Keywords: Mendelian randomization; Multiple sclerosis; genetic epidemiology; obesity.