Background: The prevalence of diabetes mellitus is rapidly increasing in China, but the secular trends in incidence and mortality remain unknown. This study aims to examine time trends from 1990 to 2017 and the net age, period, and cohort effects on diabetes incidence and mortality. Methods: Incidence and mortality rates of diabetes (1990⁻2017) were collected for each 5-year age group (from 5⁻9 to 80⁻84 age group) stratified by gender from the Global Burden of Disease 2017 Study. The average annual percentage changes in incidence and mortality were analyzed by joinpoint regression analysis; the net age, period, and cohort effects on the incidence and mortality were estimated by age-period-cohort analysis. Results: The joinpoint regression analysis showed that age-standardized incidence significantly rose by 0.92% (95% CI: 0.6%, 1.3%) in men and 0.69% in women (95% CI: 0.3%, 1.0%) from 1990 to 2017; age-standardized mortality rates rose by 0.78% (95% CI: 0.6%, 1.0%) in men and decreased by 0.12% (95% CI: -0.4%, 0.1%) in women. For age-specific rates, incidence increased in most age groups, with exception of 30⁻34, 60⁻64, 65⁻69 and 70⁻74 age groups in men and 25⁻29, 30⁻34, 35⁻39 and 70⁻74 age groups in women; mortality in men decreased in the younger age groups (from 20⁻24 to 45⁻49 age group) while increased in the older age groups (from 50⁻54 to 80⁻84 age group), and mortality in women decreased for all age groups with exception of the age group 75⁻79 and 80⁻84. The age effect on incidence showed no obvious changes with advancing age while mortality significantly increased with advancing age; period effect showed that both incidence and mortality increased with advancing time period while the period trend on incidence began to decrease since 2007; cohort effect on incidence and mortality decreased from earlier birth cohorts to more recent birth cohorts while incidence showed no material changes from 1982⁻1986 to 2012⁻2016 birth cohort. Conclusions: Mortality decreased in younger age groups but increased in older age groups. Incidence increased in most age groups. The net age or period effect showed an unfavorable trend while the net cohort effect presented a favorable trend. Aging likely drives a continued increase in the mortality of diabetes. Timely population-level interventions aiming for obesity prevention, healthy diet and regular physical activity should be conducted, especially for men and earlier birth cohorts at high risk of diabetes.
Keywords: age-period-cohort effect; diabetes mellitus; incidence; joinpoint regression analysis; mortality; trends.