Genetic epidemiological aspects of obesity and related traits. Human studies in Swedish and North-American cohorts

Aims: To investigate two obesity candidate genes, the melanocortin-4 receptor (MC4R) and the adiponectin hormone (APM1), by screening their genomic sequences. To search the human genome for quantitative trait loci (QTLs) related to resting metabolic rate (RMR), and respiratory quotient (RQ). To investigate whether the aetiology of the obesity epidemic might have a genetic component via assortative mating for obesity.Methods: The coding and flanking sequence of MC4R was screened in a total of 433 Whites and 96 Blacks (94% females) obese and normal-weight subjects from the Swedish Obese Subjects (SOS) study, the HERITAGE Family Study, and a Memphis cohort. The translated exons and part of intron 2 of APM1 was sequenced in 96 obese and 96 normal-weight Swedish women from the SOS study. Regression-based and variance-components-based genome-wide autosomal scans on RMR and RQ phenotypes, obtained from indirect calorimetry, were performed in 169 families ascertained to the Quebec Family Study via an obese proband or from the general population. Measured and validated questionnaire data on family members from 7834 obese Swedish probands, and from 829 subjects randomly ascertained from the general Swedish population were available for analysis.Results: Twelve missense MC4R mutations were detected. None of them were associated with obesity phenotypes. Moreover, two heterozygous deletions were found in two obese women: a 65_ 64delTG mutation, and a 171delC frameshift mutation, predicted to result in a truncated non-functional receptor. Three previously know APM1 variants were found in equal frequencies among obese and controls. Among them, T45T and G276G were associated with features of the metabolic syndrome in obese carriers. Evidence for linkage to RMR was detected on chromosomes 3q26.1 (LOD = 2.74), 1q21.2 (2.44), and 22q12.3 (1.33). QTLs influencing RQ were found on 12q13 (1.65) and 14q22 (1.83). Considerable locus heterogeneity within this population was suggested, as the majority of families were unlinked to any one QTL. Spouse correlations in body mass index (BMI) were strongest among couples with the shortest duration of cohabitation, suggesting assortative mating. Obesity concordance in parents was associated with an obesity prevalence of 20.1% in adult offspring, compared to 1.4% if parents were concordantly non-obese (odds ratio (OR): 18.3, 95% confidence interval: 9.0 37.4). No association was found between rearing parents and non-biological offspring s BMI.Conclusions: Pathogenic mutations in the MC4R are rare in the general obese population. Synonymous and intronic APM1 variants may influence components of the metabolic syndrome in obese carriers. All three RMR linkages overlapped regions previously linked to the metabolic syndrome, and a significant association between RMR and the metabolic syndrome found in the present cohort reinforces this relationship. The QTL on 1q21 may be a replication of one previously reported. Assortative mating for obesity confers a higher risk for obesity in the offspring generation and may therefore contribute to the obesity epidemic. Parental obesity concordance is a strong, easily identifiable genetic risk factor that should be considered in designing primary prevention programs