Investigating the paradox of increasing obesity and declining heart disease mortality in the United States: Age-period-cohort model

BACKGROUND: Obesity as a risk factor of heart disease (HD) is confirmed through observational, laboratory, and intervention studies. However, it cannot explain why HD declines, but obesity increases in the United States in recent decades. This study attempted to understand this paradox. METHODS: Annual data of national HD mortality (1999–2018) were derived from Wide-Ranging Online Data for Epidemiologic Research, biannual obesity data (1999–2016) from the National Health and Nutrition Examination Survey, and smoking data (1965–1990) were from the National Health Interview Survey. Age-period-cohort method was used to decompose HD mortality into age, period, and cohort effects, and to estimate age-cohort adjusted mortality rates. To explain the paradox, age-cohort adjusted rates were associated with obesity rates to verify the positive obesity-HD relationship, while smoking rates were associated with cohort effects to explain the current declines in HD mortality. RESULTS: During 1999–2018, the prevalence of obesity increased while the crude HD mortality rate declined for both sex and all races. After controlling for the curvilinear age effect and consistent declining cohort effect, the age-cohort adjusted HD mortality sustained stable in 1999–2007 and increased thereafter. The age-cohort adjusted rate in 1999–2018 (per 100,000) increased from 189.31 to 238.56 for males, 67.23 to 90.28 for females, 115.54 to 157.39 for White, 246.40 to 292.59 for Black, 79.79 to 101.40 for Hispanics, and 49.95 to 62.86 for Asian. The age-cohort adjusted HD mortality rates were positively associated with obesity rates (r = 0.68 for males, 0.91 for females, 0.89 for White, and 0.69 for Hispanic, p < 0.05), but not significant for Black and Asian. Further, during 1965–1990, the estimated cohort effect showed a decline in HD risk and was positively associated with smoking rates (r = 0.98 for both sex, 0.99 for White, and 0.98 for Black, p < 0.01). CONCLUSION: Study findings reveal potential increase of HD risk and support the positive relationship between obesity and HD risk. Declines in HD mortality in the past two decades are primarily due to tobacco use reduction and this protective effect was entangled in the mortality rates as cohort effect.