Chronic medical conditions based obesity phenotypes: A two-step cluster analysis of a representative sample of obese American adults.

Although obesity is a heterogeneous disease, little is known regarding chronic medical conditions (CMCs) that defines variability in obese populations. The characterization of obese populations using CMCs rather than categorization using BMI alone can advance understanding of obesity. The aims of this study are to phenotype obesity in a large representative sample of non-Hispanic White (NHW), non-Hispanic Black (NHB) and Mexican American (MA) obese adults using CMCs, and assess relationship between resulting phenotypes and self-rated health (SRH).Sex-specific two-step cluster analysis was used to phenotype obese participants (n = 12,547) to CMC-based clusters. The prevalence of CMCs and lifestyle risk factors in each cluster was assessed. Sex and race/ethnic specific association between cluster membership and SRH was determined using odds ratio (OR) from logistic regression analysis.Distinct subgroups of obese men and women were observed: moderate dyslipidemic healthy young obese men, hypertensive-dyslipidemic middle-age obese men, hypertensive young obese men, hypertensive-dyslipidemic-asthmatic middle-age obese men, and syndemic elderly obese men, healthy young obese women, hypertensive-dyslipidemic middle-age obese women, dyslipidemic young adult obese women, syndemic middle-age obese women, and syndemic elderly obese women. Participants in the more CMCs symptomatic clusters reported high rates of behavioral risk factors and showed significantly greater odds of poor SRH than participants in the less symptomatic clusters. Compared to obese persons who are asymptomatic for CMCs, syndemic elderly obese and women men had much higher increased ORs for poor SRH with values of 3.88 [95% CI = 2.41-6.26], 3.96 [95% CI = 1.86-8.30] and 7.25 [95% CI = 2.41-9.6] for NHW, NHB and MA men, respectively. The corresponding ORs for women are 4.08 [95% CI = 2.71-6.14], 4.01 [95% CI = 2.40-6.69], and 2.62 [95% CI = 1.32-5.19], respectively.Obesity treatment and intervention should consider heterogeneity within obese persons and pay greater attention to obesity related co-morbidities and metabolic manifestations.