The ADRB3 Trp64Arg variant and BMI: a meta-analysis of 44 833 individuals

Background: The beta-3 adrenergic receptor gene (ADRB3) is part of the adrenergic system, which is known to play a key role in energy metabolism. The association between the Trp64Arg variant in the ADRB3 and body mass index (BMI) has been widely examined, but previous studies have been small and results have been inconsistent. Methods: We assessed the association between the ADRB3 Trp64Arg variant and BMI in a large UK population-based cohort of 4854 middle-aged men and women. We also performed a meta-analysis of 97 studies, involving 44 833 individuals, to place our findings in context. Results: Although we found no significant difference in BMI (0.20 kg/m(2), P = 0.40) between the Trp64Trp homozygotes and Arg64 allele carriers in our UK population-based cohort, the meta-analysis showed significant association between the Arg64Trp variant and BMI, with Arg64-allele carriers having a 0.24 kg/m(2) (P = 0.0002) higher BMI compared with noncarriers. However, we also found substantial heterogeneity among the studies (P = 2.2 x 10(-14)). The difference in East Asians (0.31 kg/m(2), P = 0.001) was 3.9 times larger than that in Europeans in whom no significant association was observed (0.08 kg/m(2), P = 0.36). This was consistent with the chronological cumulative decrease in the effect size, which decreased steadily in Europeans and reached nonsignificance after 11 studies in 1996. In East Asians, the cumulative effect size decreased after the first reports, but reached a steady state at a significant effect size of 0.24 kg/m(2) in 2000. Although the funnel plot indicated no apparent publication bias, smaller studies tended to report greater differences in BMI, compared with larger studies. Conclusions: Collectively, these data suggest that the Trp64Arg ADRB3 genetic variant might be associated with BMI in East Asians, but not Europeans. More generally, our study shows the importance of meta-analyses in the field of genetic association studies for common traits. Each genetic variant makes only a small contribution to variation in BMI, and large sample sizes are needed to reliably assess and interpret gene-phenotype associations.