Over recent decades, the prevalence of obesity has increased dramatically worldwide. Although this epidemic is mainly attributable to modern (western) lifestyle, multiple twin and adoption studies indicate the significant role of genes in the individual's predisposition to becoming obese. As the hypothalamus plays a central role in controlling body weight, its regulatory circuits may represent a crucial system in the pathogenesis of the disorder. Genetic variations in genes in the hypothalamic pathways may therefore contribute to the susceptibility for obesity in humans and animals. We summarize current knowledge on the physiological role of the hypothalamus in body-weight regulation and review genetic studies on the hypothalamic candidate genes in relation to obesity. Together, data from functional and genetic studies as well as the new, common, obesity loci identified in genome-wide association scans support an important role for the hypothalamic genes in predisposing to obesity. However, findings are still inconclusive for many candidate genes. To improve our understanding of the genetic architecture of common obesity, we suggest that specific obesity phenotypes should be considered and different analytical approaches used. Such studies should consider multiple genes from the same physiological pathways, together with environmental risk factors.

Aims/hypotheses We recently reported significant associations between BMI and three TUB single nucleotide polymorphisms (SNPs) in two Dutch cohorts enriched for type 2 diabetes. Here, we attempted a replication of these associations in a large population-based cohort of female twins comprehensively phenotyped for measures of general and central obesity.Methods Two TUB SNPs (rs2272382, rs2272383) and a third (rs1528133), 22 kb distal to RIC3, were genotyped in 2694 Europid women from the St Thomas' UK Adult Twin Registry (Twins UK) (mean age±SD: 47.6±12.7 years; 42.8% postmenopausal). We explored the hypothesis that TUB is a candidate gene for late-onset obesity in humans through testing the interaction of the SNPs by menopausal status. Results In the whole cohort, none of the three SNPs showed a significant main effect on measures of general or central obesity. However, for central obesity the rs2272382 SNP showed a significant interaction with menopausal status (p=0.036). Postmenopausal women homozygous for the minor allele of rs2272382 showed significantly more general obesity (p=0.022) and central obesity (p=0.009) than carriers of the major allele. Differences (beta [95% CI]) between the two genotype groups were 0.92 kg/m 2 (0.03-1.81) for BMI (p=0.036), 2.73 cm (0.62-4.84) for waist circumference (p=0.013) and 2.43% (0.27-4.60) for per cent central fat (p=0.027). These associations were confirmed by a sibling transmission disequilibrium test for central obesity, waist circumference and per cent central fat. Conclusions/interpretation We have replicated associations of TUB SNP rs2272382 with measures of general and central obesity in normal postmenopausal women. These findings confirm TUB as a candidate gene for late-onset obesity in humans.

No abstract