The association between 25-hydroxyvitamin D and 5-, 10-, or 15-year weight change were assessed in a population-based, prospective study conducted in Lausanne, Switzerland. Data from the first (2009−2012, N = 3527, 51.3% women), second (2014−2017, N = 3237, 53.8% women), and third (2018−2021, N = 2567, 54.2% women) follow-ups were used. A weighted genetic risk score (GRS) of 115 SNPs associated with vitamin D levels was constructed. At baseline, the GRS correlated positively with 25-hydroxyvitamin D levels based on a Spearman rank correlation and 95% confidence interval: 0.198 (0.166; 0.231), p < 0.001; and with body mass index: 0.036 (0.004; 0.068), p = 0.028. No association was found between quartiles of GRS and weight changes at 5, 10, or 15 years: multivariate-adjusted weight changes ± SEM at 5-years follow-up were 1.39 ± 0.17, 1.13 ± 0.17, 1.24 ± 0.17, and 1.00 ± 0.17 kg for the first to the fourth quartile of the GRS, respectively (p = 0.401). Two-step linear regression showed a significant but clinically meaningless association between GRS-derived vitamin D and weight change at 5- and 15-years: slope and 95% confidence interval for a 5 nmol/L increase in GRS-derived 25-hydroxyvitamin D levels: 0.082 (0.013; 0.150) and 0.130 (0.018; 0.243) kg, respectively. We conclude that there is little association between genetically determined 25-hydroxyvitamin D levels and weight gain.
Keywords: Switzerland; mendelian randomization; prospective study; vitamin D; weight change.