What Factors Affect Mortality over the Age of 40?

Park, Jong Ku; Koh, Sang Baek; Kim, Chun Bae; Kang, Myung Guen; Park, Kee Ho; Wang, Seung Jun; Chang, Sei Jin; Sin, Soon Ae

Abstract

OBJECTIVES
This study was conducted to identify the factors influencing the mortality of Koreans over the age of 40 by a nested case-control study.
METHODS
The cohort consisted of the beneficiaries of Korea Medical Insurance Corporation for Government Employees & Private School Teachers and Staff(KMIC) who received health examinations of KMIC in 1992 and 1993 retrospectively. At that time, they were more than 40 years old. The cases were 19,258 cohort members who had died until December 31, 1997. The controls were 19,258 cohort members who were alive until December 31, 1997. Controls were matched with age and sex distribution of the cases. The data used in this study were the funeral expenses requesting files, and the files of health examinations and health questionnaires gathered in 1992 and 1993. To assess the putative risk factors of death, student t-test, chi-square test, multiple logistic regression analysis were used.
RESULTS
In multiple logistic regression analysis, independent risk factors of death were as follows; systolic blood pressure, diastolic blood pressure, blood glucose, AST, urine glucose, urine protein, alcohol drinking(frequency), cigarette smoking and perceived health status, intake of restoratives and blood transfusion showed positive associations with death; coffee consumption showed negative associations with death; and body mass index and serum total cholesterol showed J-shaped association with death.
CONCLUSIONS
Regarding the direction of association, the result of analysis on the data restricted to '96-'97 was same as that of '93-'97. But in some variables such as obesity, serum cholesterol, the odds ratios of death in the data of '96-'97 were higer than those of '93-'94, which suggested that the data of '93-'94 was bearing effect-cause relationship. We concluded that it suggested further researches using long-term follow-up data to be needed in this area.