RESUMO
OBJECTIVE: To define whether the rs9939609 FTO (fat mass and obesity associated) single nucleotide polymorphism (SNP) is associated with anthropometric measurements and its modulation by educational level in a Mediterranean population. METHODS: We studied 3 independent adult samples: a random sample (n = 1580) from the general population (GP), obese hospital patients (OHP) (n = 203) and elderly subjects (n = 1027) with high cardiovascular risk (HCR). Weight and height were directly measured. Education and physical activity (PA) were measured using questionnaires. RESULTS: The rs9939609 presented heterogeneous associations with BMI. In the GP, the minor A-allele was significantly associated with greater BMI, following a co-dominant pattern (P = 0.009), whereas in the OHP this association was recessive (P = 0.004). Conversely, we did not find a significant association with BMI in the HCR group (P < 0.596). In the GP we found a significant interaction between the FTO SNP and education (P = 0.048). In the stratified analysis, no association of the FTO SNP with greater BMI in university subjects was detected (P = 0.786), whereas the association was observed in non-university subjects (P = 0.001). The FTO × education interaction (P = 0.020) was also observed in determining obesity risk in the GP. A-allele carriers had a greater risk of being obese only if they had no university education (OR: 1.56; 95%CI: 1.09-2.23 for TA and OR: 2.01; 95%CI: 1.27-3.26 for AA subjects). The interaction of the FTO with education remained significant even after adjustment for PA. CONCLUSIONS: The association of the FTO SNP with greater BMI and obesity risk in the GP was strongly modulated by education.