Age-period-cohort model of ischemic and hemorrhagic stroke in Hong Kong

Abstract

Background: As an increasingly serious aging problem, stroke causes a heavy health burden. Ischemic and hemorrhagic stroke are two main categories of stroke with totally different underlying biological mechanisms and different incidence rates and mortality rates in males and females. The global stroke mortality has an overall downward trend with various patterns in different districts and countries. There is no research reporting the age, period and cohort effects of ischemic and hemorrhagic stroke separately in Hong Kong.

Objective: This dissertation aims to use age-period-cohort (APC) model to analyze relative risks (RRs) and age-standardized mortality of ischemic and hemorrhagic stroke and compare these two types of stroke to report the age, cohort and period effects and find out the possible reasons of the trends and propose targeted health prevention strategies.

Method: Ecological study is designed to analyze unit of age, period and cohort effects. The dataset of all registered deaths during 1998 to 2013 in Hong Kong by International Classification of Diseases-9 (ICD-9) and ICD-10 coding and the data of age and gender specific mid-year populations were offered by Census and Statistics Department (C&SD). Multiple imputation was used to estimate missing data and misclassification data. Age weighting was used to calculate the age-standardized mortality. APC and age-cohort model (AC model) was fit to estimate the RRs. R i386 3.1.1 was used to analyze the data.

Result: In terms of age-standardized mortality during 1998 to 2013 in Hong Kong, ischemic stroke showed a clear decline but hemorrhagic stroke had a fluctuated decline. Both stroke types had significant cohort effects of lower risk. In period effects, ischemic stroke had a downward trend recently in risk of death, while hemorrhagic stroke had a small increasing risk.

Conclusion: Ischemic stroke had a much stronger cohort effects during the past 10 decades, while hemorrhagic stroke did not have a clear decline recently and will continue to cause disease burden. The patterns in these two stroke types in male and female are different. More research is needed to evident, and targeted public health strategies should be considered to prevent these two stroke types.