The relationship between multiple chronic diseases and depressive symptoms among middle-aged and elderly populations: results of a 2009 korean community health survey of 156,747 participants

This study was performed using the data from the Korean Community Health Survey (KCHS) conducted by the Korean Centers for Disease Control and Prevention (KCDC) in 2009. The KCHS has been a national health survey with direct face-to-face interviewing by a trained interviewer, and 253 national public health centers have participated annually since 2008. The KCHS has public confidence and represents the entire population of Korea. It is secondary data produced by the Korean government for the purpose of public use and is designed as a cross sectional study [https://​chs.​cdc.​go.​kr/​chs/​intro.​html]. The aims of the KCHS are to calculate the health statistics of city and county units to establish regional health plans, to produce basic data to systematically assess the local health service, to expand the community surveillance infrastructure by establishing public cooperation, and to perform a standardized survey to compare the health of the population [22]. The KCHS 2009 was conducted from September to November 2009 on study participants aged 19 or older in each area who were selected by the probability proportional sampling method and the systematic sampling method. The KCHS collects detailed information, including sociodemographic characteristics, 15 types of chronic diseases in the individual’s past medical history data (defined as “diseases diagnosed by medical doctors”), lifestyle, behaviors, and health-related problems. The KCHS protocol is annually reviewed and approved by the institutional review board of the KCDC. Written informed consent was obtained from all participants in the KCHS. No additional institutional review board approval was required for this study because the data were intended for public purposes [23]. The 2009 KCHS included 230,715 participants [24]. We excluded 73,659 participants under the age of 40 years due to the low prevalence of chronic diseases in this population. In fact, as shown in previous studies, the prevalence of MCDs and chronic diseases, such as hypertension and diabetes, in individuals less than 40 years of age is relatively low [25, 26]. The prevalence of MCDs is only 0.4% in individuals in their 20s and 1.4% in those in their 30s. In contrast, the prevalence of MCDs is 11.9% in those aged 40–59 years and 46.3% in those aged ≥60 years in the 2009 KCHS. In addition, this study focused on comparing middle-aged and elderly populations. For these reasons, participants under the age of 40 years were excluded from the study. Moreover, 309 individuals who did not provide complete information on the questionnaire were excluded from this study (206 individuals had missing data on MCDs, and 103 individuals had a missing Center for Epidemiologic Studies Depression Scale (CES-D-K) score). Finally, a total of 156,747 participants (88,749 aged 40–59 years and 67,998 aged ≥60 years; 71,695 men and 85,052 women) were included in the study. The participant selection process is shown in Fig. 1.

The participants were divided into age groups, namely, 40–59 years and 60 years or over, for comparisons between the middle-aged and elderly populations. Marital status was classified as follows: married, separated, divorced, widowed, and never married. Employment was classified as yes/no. Occupational categories were classified into manual, non-manual, and unemployed/others. The category of other occupations included students and housewives. Educational status was classified as not educated, elementary school, middle school, high school, and college or more. The average monthly household family income of the participants was classified as 1 million won (Korean currency unit) or less, 2 million won or less, 3 million won or less, 4 million won or less, and 4.01 million won or more. It is the monthly average of the past one year of actual income of all household members except taxes. One thousand won roughly equals one US dollar. The residence locations were classified into four groups according to the Korean administrative districts, which are based on population size: (1) county (population less than 50,000), (2) small city (population more than 50,000), (3) medium city (population more than 500,000), and (4) metropolitan area (population more than 1,000,000). All questionnaires were conducted through one-on-one interviews.

The 2009 KCHS covered 15 chronic medical diseases: (1) diabetes, (2) dyslipidemia, (3) hypertension, (4) angina, (5) myocardial infarction, (6) stroke, (7) osteoporosis, (8) arthritis, (9) tuberculosis, (10) asthma, (11) hepatitis B, (12) atopic dermatitis, (13) allergic rhinitis, (14) cataract, and (15) depression. All chronic diseases were investigated with the following question: “Have you ever been diagnosed with any specific disease by a physician in your lifetime?” We used the number of comorbid chronic diseases to define the MCDs. MCDs were defined as the concurrent presence of two or more chronic diseases. Tuberculosis can be cured, but the treatment period can last for more than six months; therefore, tuberculosis affects the individual’s quality of life and is not only an endemic disease but also one of the highest public health priorities in Korea. Therefore, we included tuberculosis as an important medical disease despite it being treatable [27].

We used depressive symptoms as a dependent variable. Commonly, depressive symptoms include low mood, feeling fatigue, anxiety, reducing self-esteem, and sleep disturbance [28]. Also, the survey about depressive symptoms is an effective tool for screening depression [29].

The 2009 KCHS used the Korean-validated 20-item version of the CES-D-K to assess depressive symptoms [30] and the definition of depressive symptoms is equal to or greater than a CES-D-K score of 16 [31]. The CES-D was not originally developed to generate a clinical diagnosis of depression, but to detect individuals with depressive symptoms. The CES-D has been translated into Korean, and its reliability and validity have been confirmed [32]. Cho et al. suggested two optimal cut-off points: one is the universal cut-off point of 16, which effectively detects and assesses ‘probable’ depressive symptoms; the other is 25, which more accurately corresponds to the DSM-III-R classification of major depression and is regarded as the cut-off point for indicating severe, ‘definite’ depression symptoms [33].

This scale is designed to identify the existence of depressive symptoms and evaluate their severity. It is a 20-item measure consisting of 4 positively worded items and 16 negatively worded items. The positive items are reverse-coded so that the scores have a potential range of 0 to 60, with higher scores indicating the presence of more depressive symptoms [23]. The internal consistency of the Korean version of the CES-D-K was verified in 1993 by Cho et al. (Cronbach’s alpha = 0.9098) [32].

All analyses were stratified by age group for comparisons between the middle-aged and elderly populations. Chi-square tests were used to examine the significance of the differences in the sociodemographic characteristics, numbers of chronic diseases, presence of depressive symptoms and MCDs, and depression diagnosed by physicians between the age groups. The CES-D-K scores are expressed as the medians and interquartile ranges (IQRs) and compared by age group using Mann–Whitney U test. The scores were transformed to a natural logarithm because they did not follow a normal distribution.

We performed a multivariate logistic regression analysis to assess the relationship between depressive symptoms and each chronic disease, MCDs, and number of diseases. Multivariate logistic regression analyses were performed to calculate the adjusted risk ratios (RRs) and 95% confidence limits (CLs) by adjusting for variables affecting the depressive symptoms, such as age (years), gender (except when the results were stratified by gender), marital status, employment, occupational categorization, education, monthly income, and residence location [34], and each disease to correct for the influence of other diseases and analyze the depressive symptoms in each disease. P values <0.05 were considered statistically significant, and all statistical analyses were performed with SAS 9.4 (SAS Inc., Cary, NC, USA).