An important consideration is whether our finding of rs1659258 at THNSL2 in association with VAT in women represents a variant that is specific to central adiposity. Indeed, our strongest result is derived from VAT itself. However, we observed associations with SAT in women, and our validation data suggests this SNP is also associated with BMI in women. Once we adjusted the VAT trait for BMI, our SNP did not completely lose its statistical significance (p = 6.9*10E-05), suggesting that there is some specificity of our finding to fat deposition in the central abdominal compartment. This result is in contrast to our prior results for a SNP in NRXN3 in association with waist circumference: upon adjustment for BMI, statistical significance was attenuated [43]. Finally, our eQTL data reveals nominal expression in subcutaneous but not visceral adipose tissue.
