The association of obesity phenotypes and risk of cardiovascular disease using time-varying and time-invariant approaches: An 18-year follow-up cohort study.

Background and aim: Our aim was risk estimation for cardiovascular disease (CVD) across obesity phenotypes over 18 years of follow-up using both time-invariant and time-varying approaches.

Methods and results: This prospective cohort study included 9752 participants aged ≥30 years examined in the first and second phases of the Tehran Lipid and Glucose Study (1999-2001 and 2002-2005). Six phenotypes [i.e., metabolically healthy normal weight (MHNW), overweight (MHOW), and obese (MHO), as well as metabolically unhealthy normal weight (MUNW), overweight (MUOW), and obese (MUO)] were defined based on the body mass index (BMI) and metabolic status. Incident CVD was documented until March 2018. Time-invariant and time-varying Cox regression models were used to estimate CVD hazard ratio (HRs) for obesity phenotypes. Mean age of the participants was 46.6 ± 12.0 years, and 53.9 % of them were women. During 18 years of follow-up, 1083 new CVD events occurred. In metabolically unhealthy individuals, but not metabolically healthy people, multivariable-adjusted HRs for CVD events increased by BMI according to time-varying (HR = 1.6, 95 % CI = 1.13-2.26 for MUNW; HR = 1.92, 95 % CI = 1.43-2.58 for MUOW; HR = 1.94, 95 % CI = 1.4-2.68 for MUO) and time-invariant (HR = 1.85, 95 % CI = 1.01-3.39 for MUNW; HR = 2.75, 95 % CI = 1.63-4.63 for MUOW, and HR = 3.26, 95 % CI = 1.95-5.47 for MUO) models.

Conclusion: Metabolically unhealthy overweight and obese individuals are at increased risk of CVD and should be regularly screened to prevent possible cardiovascular events.