%0 Historical Article %T Historical trends in suicide risk for the residents of mainland China: APC modeling of the archived national suicide mortality rates during 1987-2012. %A Chen X %A Sun Y %A Li Z %A Yu B %A Gao G %A Wang P %J Soc Psychiatry Psychiatr Epidemiol %V 54 %N 1 %D Jan 2019 %M 30171272 %F 4.519 %R 10.1007/s00127-018-1593-z %X OBJECTIVE: Distinctive and dramatic changes in the history of China with a rapid suicide decline in recent years present an opportunity to investigate the risk of suicide. In this study, we investigated suicide risk with a historical perspective with archived data to inform suicide research and prevention policies and strategies.
METHODS: Documented age-specific suicide mortality rates in 1987-2012 were decomposed into age, period, and cohort effect using APC-modeling method and intrinsic estimator (IE) technique. The estimated effects were further analyzed by numerical differentiation.
RESULTS: The data satisfactorily fit the constructed APC models. Cohort effect indicated that suicide risk in China fluctuated at very high levels during 1903-1967, followed by a sharp decline during 1968-1977, and reached the lowest level in 1983-1987 before increased again. Period effect confirmed the declining trend since 1987. Three sunny cohorts with reduced suicide risk and four cloudy cohorts with increased risk were, respectively, associated with significant cultural, social, political, and economic events in China since the 1900s.
CONCLUSIONS: The mega trends in the suicide risk at the population level are closely related to significant historical events in China. Suicide is anticipated to increase because of the growing risk for the young cohorts (particularly young females) as the country further develops. Study findings suggest the significance of national strategies for suicide prevention and control, including maintenance of social harmony and stability, provision of more opportunities for development, enhancement of social integration, and restriction of suicide facilitating factors.