The Lancet regional health. Western Pacific
Projections of diabetes mellitus in Japan through 2040: a bayesian age-period-cohort analysis of the national health and nutrition survey population
By Mondal R, Ogata S, Kiyoshige E and 9 others.
Abstract
Background
Accurate projection of the burden of diabetes mellitus (DM) requires consideration of age, period, and cohort (APC) effects underlying prevalence trends. We estimated projected trends in DM prevalence and case numbers in Japan from 2020 until 2040, by sex, using the Bayesian APC (BAPC) model.
Methods
We employed Bayesian age-period-cohort modeling and prediction (BAMP), using observed populations (aged 40 years or older) and DM cases from the National Health and Nutrition Survey from 2003 to 2019, to project national-level DM prevalence and cases from 2020 to 2040 using the national-level projected population until 2040. We further compared BAPC's predictive performance with that of conventional prediction model 1 with a constant prevalence rate and conventional prediction model 2 with a monotonically changing prevalence rate, using root mean square error (RMSE) between observed and projected case numbers for model validation during 2015-2019.
