Age–period–cohort analysis of the suicide rate in Korea

• We presented the trends in Korea’s suicide rate using Joinpoint analysis and APC modeling.
• Within the same period, the suicide rate was higher among older than younger age.
• Within the same birth cohort, the suicide rate was higher among older than younger age.
• Within the same age, the suicide rate was higher among younger than older cohorts.
• In the APC modeling, old age, recent period, and the cohort born before 1924 showed increased suicide rate.

BackgroundThe suicide rate has been increasing in Korea, and the country now has the highest rank in the world. This study aimed to present the long-term trends in Korea's suicide rate using Joinpoint analysis and age–period–cohort (APC) modeling.MethodsThe population and the number of suicides for each five-year age group were obtained from the National Statistical Office for the period 1984–2013 for Koreans aged 10 years and older. We determined the changes in the trends in age-standardized mortality rates using Joinpoint. APC modeling was performed to describe the trends in the suicide rate using the intrinsic estimator method.ResultsThe age-standardized suicide rate in men rapidly increased from 1989 to 2004, and slightly increased thereafter, whereas the suicide rate in women increased from 1989 to 2009 and then decreased thereafter. Within the same period, the suicide rate was higher among the older age groups than in the younger groups. Within the same birth cohort, the suicide rate of the older groups was also higher than that in the younger groups. Within the same age group, the suicide rate of the younger cohorts was higher than it was in the older cohorts. In the APC modeling, old age, recent period, and having been born before 1924 were associated with higher suicide rates.LimitationsThe accuracy and completeness of the suicide rate data may lead to bias.ConclusionsThis study showed an increasing trend in the suicide rates for men and women after 1989. These trends may be mainly attributed to cohort effects.