Gender-specific associations between ADIPOQ gene polymorphisms and adiponectin levels and obesity in the Jackson Heart Study cohort

BACKGROUND: Despite the important role of adiponectin in regulating general metabolic homeostasis, analysis of genetic determinants of adiponectin and the related cardio-metabolic traits in African American population has been limited and inconsistent. Considering the high genetic admixture of African Americans and thus the important population stratification that may confound the genetic-trait associations, the objective of this work was to perform a comprehensive analysis of the associations between ADIPOQ variants and adiponectin levels and obesity phenotypes in a large African American population from the Jackson Heart Study (JHS) cohort. METHODS: Genotype data was available for 2968 JHS participants (1131men; 1837women). Single Nucleotide Polymorphisms (SNPs) were selected by a Tag-SNP Approach and literature review. The genotype imputation was performed using IMPUTE2 software and reference phased data from the 1000G project. PLINK software was used for the genetic analysis. Plasma specimens were analyzed by ELISA for adiponectin levels. All analyses were controlled for population stratification assessed by Individual Proportions of European Ancestry (PEA) estimates calculated in HAPMIX using ancestry informative markers (AIMs). RESULTS: We found a gender-dependent association of some ADIPOQ variants and adiponectin levels. In women four of the studied polymorphisms (rs6444174, rs16861205, rs1403697, rs7641507) were associated with adiponectin levels after Bonferroni correction and controlling for the percentage of PEA, age, annual household income and smoking. These results were consistent with the haplotype analysis. The association between the rs12495941 variant and obesity is modulated by the PEA, so that the relationship between the G allele and a higher incidence of obesity was present in those individuals within the lower PEA group. In addition we found an effect modification of obesity on the association between the ADIPOQ rs6444174 SNP and BMI so that the presence of the T allele was negatively and significantly associated with BMI only in participants with a normal weight. CONCLUSIONS: In this large African American cohort, ADIPOQ variants were associated with adiponectin levels in a gender-dependent manner and the relationship of some of these variants with obesity and BMI was modulated by the PEA and obesity status respectively. This suggests that the effects of these polymorphisms on adiponectin and obesity phenotypes are subject to a strong interaction with genetic and environmental factors in African American population. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12881-015-0214-x) contains supplementary material, which is available to authorized users.