Quantifying the contributions of cardiovascular risk factors to cardiovascular disease trends in 21st century Japan: a microsimulation study

Fig. 1. Overview of study design, including counterfactual analysis and IMPACTNCD-JPN structure. Abbreviations: CVD, cardiovascular diseases; GAMLSS, Generalised Additive Models for Location, Scale, and Shape; CHD, coronary heart disease; QALYs, quality-adjusted life years; CPPs, cases prevented or postponed; CYPPs, case-years prevented or postponed; DPPs, deaths prevented or postponed; SBP, systolic blood pressure; LDL-c, low density lipoprotein cholesterol; BMI, body mass index; PA, physical activity; FV, fruits and vegetables. 1Cigarette consumption for current smokers. 2Detailed age- and sex-stratified distributions are presented in Supplementary Figure S1 in Supplementary Results. IMPACTNCD-JPN simulated synthetic individuals with specific age (1-year), sex, year (1-year), and values for seven CVD risk factors: SBP, smoking status, LDL-c, HbA1c, BMI, PA, and FV consumption. Each CVD risk factor was modelled based on age, sex, year, and, where applicable, selected relevant CVD risk factors among the seven. A brief structure of IMPACTNCD-JPN is also presented in the Technical Appendix “Introduction” in Supplementary Methods.

Fig. 2. Annual incidence of CHD and stroke, and their cumulative CPPs from 2001 to 2019 in Japan, under the base-case scenario and counterfactual scenarios where each modelled CVD risk factor, and all combined, are fixed at 2001 levels. Abbreviations: CHD, coronary heart disease; CVD, cardiovascular disease; PA, physical activity; FV, fruit and vegetable consumption; BMI, body mass index; LDL-c, low-density lipoprotein cholesterol; SBP, systolic blood pressure; CPPs, cases prevented or postponed. 1SBP, LDL-c, and HbA1c were adjusted for medication use of antihypertensive, cholesterol-lowering, and diabetes treatments, respectively.

Fig. 3. Cumulative CYPPs for CHD and stroke, DPPs, net gained QALYs, net saved direct and indirect costs for CHD and stroke from 2001 to 2019 in Japan, assessed by comparing counterfactual scenarios where each modelled CVD risk factor, and all combined, are fixed at 2001 levels with the base-case scenario. Abbreviations: CHD, coronary heart disease; CVD, cardiovascular disease; PA, physical activity; FV, fruit and vegetable consumption; BMI, body mass index; LDL-c, low-density lipoprotein cholesterol; SBP, systolic blood pressure; CYPPs, case-years prevented or postponed; DPPs, deaths prevented or postponed. Results are summarised using medians (points) and 95% uncertainty intervals (error bars). Since the scenarios share common parameters, the results are correlated; therefore, overlapping UIs do not necessarily indicate a lack of statistical significance. For details, please refer to Technical Appendix “Model outputs” in Supplementary Methods. 1SBP, LDL-c, and HbA1c were adjusted for medication use of antihypertensive, cholesterol-lowering, and diabetes treatments, respectively.