Relation of Regional difference in Mental health state in Seoul using multilevel data analysis

Abstract

Introduction : Importance of mental health on our life has shown through lots of studies. Many results of research said the poor mental health makes one's life unhealthy. In Korea, however, prevalence of mental illness has been growing according to 2011 mental disorder survey. On the other hand, a lots of studies has shown the Neighborhood effect(Contextual effect) on mental health. Community level features influence one's mental health state. This study is conducted to examine the difference of mental health state by small area in Seoul and find-out association between small area level factors and individuals mental health state.

Methods : The data for this study came from KCHS(Korean Community Health Study) of 2011, 2012, 2013. Adults living in Seoul(25 districts) are subjects of this study. Independent variables are determined by two level(Individual and Community) factors. Individual-level variables are including smoking, drinking, physical activity, self-perceived obese state, experience of disease, accident or poisoning. Community-level variables are calculated by proportion of female, elderly, specific marital status, under high-school educational level, manual worker, low household income(100man-won and less) population. Dependent variable(Mental health state) is evaluated by degree of stress and depression experience. Dependent variable(Mental health state) is evaluated by degree of stress and depression experience. To find-out the differences of mental health state by region, frequency analysis is performed. And finally generalized linear mixed model logistic analysis is performed to find-out the effect of community-level factors by using SAS program version 9.3.

Results : The total number of subjects for this study is 39,380 people. 12789 people of subjects answered they felt high-stress and 2255 people of them experienced depression. Community in Seoul can be divided into 25 gu and 424 dong. Average proportion of high-stress by community, named gu, is 29.3%. And depression experienced is 7.1% by each district. There is significant(p<0.05) association between mental health state and almost individual-level features including demographics and health-related factors. The results of generalized linear mixed model analysis, there are differences between region to region. Especially, experience of depression adjusted for all individual's and dong level factors has big percentage change of variance.

Conclusion : There is association between regional features and individuals mental health by comparing the differences between small area in Seoul and figuring the effect factors on mental health state. Especially, in the case of depression, there is a big association with area features. Therefore, it is necessary that establish community level approach and management to improve individuals mental health.