Prediction model for post-retirement depression in the older population: A cross-sectional study

Prediction model for post-retirement deression in the elderly

Article information

J Korean Gerontol Nurs. 2025;27(1):21-34

Publication date (electronic) : 2025 February 28

doi : https://doi.org/10.17079/jkgn.2024.00500

Assistant Professor, School of Nursing, Dongyang University, Yeongju, Korea

Corresponding author: Myeunghee Han School of Nursing, Dongyang University, 145 Dongyangdae-ro, Punggi-eup, Yeongju 36040, Korea TEL: +82-54-630-1279 E-mail: dewdrop54@daum.net

Received 2024 June 12; Revised 2024 September 28; Accepted 2025 January 6.

Abstract

Purpose

The aim was to identify high-risk groups for postretirement depression among those aged 65 or older.

Methods

This study used the 9th Korean Longitudinal Study of Ageing to identify the prevalence of depression among 1,879 retirees aged 65 and above and to develop a predictive model. The decision-tree-analysis method was used to construct a predictive model.

Results

The highest proportion of depression, at 58.5%, was observed in males who do not engage in regular exercise and have low daily living performance. In the predictive model where age is fixed as the first branch, depression was highest at 55.4% among middle-old or old-old individuals with low oral-health-related quality of life and marital status of separated/divorced/widowed/never married. In a predictive model that first classifies by sex, the proportion of depression was highest at 53.7% for dependent males.

Conclusion

This study confirmed that age, oral health, marital status, heath status, instrumental activities of daily living, activities of daily living, and regular exercise have an impact on depression. However, these factors change over time, so longitudinal studies should be conducted to understand how depression changes with them.

Keywords: Depression; Aged; Decision trees; Retirement

INTRODUCTION

1. Background

Retirement marks the end of an individual’s professional career upon reaching a certain age [1]. While it can be seen as a positive transition that allows for a more leisurely lifestyle and greater social engagement [2], it also presents challenges. The loss of work and social belonging can lead to psychological and social crises, not only for retirees but also for their family members [1]. As retirement represents the second half of the life cycle for most individuals [2], the loss of social roles can negatively affect mental health [3]. In Korea, 75% of retirees report difficulty adjusting to life after retirement [4], and elderly individuals who are no longer economically active are at a higher risk of depression [1]. As a major life event, retirement can have profound effects on mental well-being [3]. Specifically, a study reported high rates of suicide attempts in retirement among retirees due to mental health challenges, demoralization, and other issues related to retirement [5]. Therefore, it is crucial to identify elderly individuals experiencing depression after retirement, determine the factors contributing to their condition, and take proactive measures to prevent depression. There is an urgent need to identify retirees at high risk for depression and develop targeted prevention programs tailored to their needs.

Previous studies have identified differences in depression among the elderly based on age [6,7], sex [1,8,9], region of residence [10,11], spousal status [9,12], education level [13], exercise [14,15], physical activity [16], oral health [17], and alcohol and tobacco use [11,18]. A study focusing solely on male retirees found that retirement stress, lack of optimism, and marital dissatisfaction were associated with higher depressive tendencies [19]. Additionally, higher social contact has been linked to lower depression levels [2], while positive family relationships have been shown to reduce anxiety and stress in retirees, ultimately mitigating depression [20]. Similarly, frequent interactions with children and friends, along with participation in social activities, have been found to lower depression in the elderly [12,21]. A study found that declining physical functioning due to aging contributes to depression by making daily life more challenging. Conversely, individuals who perceive themselves as healthier are less likely to experience depression [9].

However, despite the severity of post-retirement depression, few studies have examined the factors that influence it. To date, most research on retirement and depression has focused on males [2,22]. Therefore, this study aims to develop models to predict the presence of depression in retirees population aged 65 and older and to identify high-risk groups after retirement.

2. Purposes of Study

This study aims to develop models for predicting depression in retirees aged 65 and older. The specific objectives are as follows:

1. Compare the general characteristics of individuals with and without depression.

2. Compare health-related characteristics between those with and without depression.

3. Compare retirement-related characteristics between those with and without depression.

4. Develop a model to predict depression.

5. Develop a model to predict depression based on age.

6. Develop a model to predict depression based on sex.

METHOD

Ethic statement: The Aging Research Panel Survey used in this study consisted of secondary data that does not contain personally identifiable information, ensuring anonymity and confidentiality. As a result, the study was granted exemption from review by the Institutional Review Board (IRB) of Dongyang University (IRB No. 1041495-202409-HR-02-01).

1. Study Design

This study employed a descriptive design using empirical analysis of secondary data from the 9th Korean Longitudinal Study of Aging conducted in 2022. The study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (https://www.strobe-statement.org).

2. Study Population

The Aging Research Panel Survey was designed to collect data on the demographic, economic, and health characteristics of the elderly, providing a foundation for socioeconomic policy development. Data collection began in 2006 with 10,254 individuals aged 45 and older nationwide through computerized face-to-face interviews, with the 9th wave of data collected in 2022.

For this study, data were obtained through the Employment Survey Analysis System after completing the required registration process. To determine the appropriate sample size, we used G*Power 3.1.9.7, calculating a minimum of 557 subjects based on the general proportion method, with an odds ratio of 1.3, a proportion of .70, a significance level of .05, and a power (1-β) of 80.

The study was conducted on a sample of 1,879 subjects aged 65 and older who responded “retired (I have no intention to work further unless there is a special situation)” to the question “Which of the following states is closest to you with regard to working in the labor market?” and who also completed the depression section.

3. Study Instruments

1) General Characteristics

The general characteristics of the study participants included age (younger elderly: 65~69 years, mid-older elderly: 70~74 years, older elderly: 75~84 years, and very old elderly: 85 years and older) [14], sex (male, female), marital status (single, separated/divorced/widowed, married), education (elementary school or less, junior high school, high school, graduate, or higher), place of residence (urban, rural), whether participants had grandchildren they helped raise before the age of 10 (yes or no), the number of people in their household (including themselves), the number of living children, and the number of grandchildren.

Participation in social activities was categorized as “yes” or “no” for religious gatherings, social gatherings (e.g., rotating savings and credit associations, senior centers), leisure, cultural, or sports-related organizations (e.g., senior colleges), and alumni, hometown friendship, or clan associations. The frequency of participation in these activities was classified into four categories: “None,” “1~3 times per week,” “1~2 times per month,” and “1~6 times per year.”

2) Health-Related Characteristics

Health-related characteristics were categorized into three levels: good, fair, and poor health. The following variables were classified as “yes” or “no”: activity (work) limitations due to a health condition, physician-determined disability, current difficulty with daily activities due to vision impairment, regular use of hearing aids, regular use of dentures, difficulty with daily activities due to pain, regular exercise, current smoking, and current alcohol consumption.

Activities of daily living (ADL) and instrumental activities of daily living (IADL) were categorized as “independent” or “dependent,” while oral-related quality of life was classified as “poor” or “good.”

3) Retirement-Related Characteristics

Retirement-related characteristics included satisfaction with retirement, categorized as very satisfied, satisfied, or not at all satisfied, and a comparison of life before and after retirement, classified as better in retirement than before retirement, similar in retirement and before retirement, or not better in retirement than before retirement.

4) Depressive Characteristics

The depression scale used in this study was the Korean version of the Center for Epidemiological Studies-Depression Scale (CES-D)-10 (Boston form), a shortened version of the 20-item U.S. CES-D developed for the elderly and chronically ill. The scale included the following items: 1) People seemed to treat me coldly, 2) I felt sad, 3) I felt very depressed, 4) I felt like everything was hard, 5) I felt like I was doing relatively well, 6) I felt like people didn’t like me, 7) I didn’t sleep well, 8) I had no major complaints, 9) I felt alone in the world, and 10) I didn’t know what to do. Each item was scored as follows: always felt that way (5~7 days) (3 points), often felt that way (3~4 days) (2 points), sometimes felt that way (1~2 days) (1 point), and briefly felt that way or never felt that way (less than one day) (0 points), with a total score of 0~30. Based on previous research [23,24], a cutoff point of 10 was used, categorizing those scoring above 10 as “depressed” and those scoring below 10 as “non-depressed.”

4. Data Analysis

Data analysis for this study was conducted using SPSS ver. 29.0 (IBM Corp.). Descriptive statistical analysis, independent samples t-tests, and cross-tabulation were performed to analyze the data. A decision tree analysis was also conducted to develop models for predicting depression.

Since the study variables included both nominal and continuous data, we applied the χ²-test for discrete variables and chi-squared Automatic Interaction Detection, which uses the F-test for continuous variables [25,26]. The tree configuration was set to a maximum of three levels, with 100 parent nodes and 50 child nodes [25,26]. To verify the validity of the final tree model, split-sample validation was performed [26].

5. Ethical Considerations

The Aging Research Panel Survey used in this study consisted of secondary data that does not contain personally identifiable information, ensuring anonymity and confidentiality. As a result, the study was granted exemption from review by the Institutional Review Board (IRB) of Dongyang University (IRB No. 1041495-202409-HR-02-01).

RESULTS

1. General Characteristics

A comparison of the general characteristics of the participants is presented in Table 1. After retirement, the older elderly accounted for the highest proportion in both the non-depressed group (44.2%) and the depressed group (44.1%). The non-depressed group included 18.5% of the younger elderly and 21.1% of the mid-older elderly, while the depressed group had 11.9% and 17.6%, respectively. However, there was a significant difference in age distribution between the two groups (p<.001), with the proportion of very old elderly being lower in the non-depressed group (16.2%) than in the depressed group (26.4%).

Table 1.

General Characteristics (N=1,879)

The non-depressed group had a higher proportion of married individuals, 73.5% compared to the depressed group (61.0%) (p<.001). Education level also showed a significant difference (p<.001), with the highest proportion of individuals in both groups having an elementary school education or less, including 41.2% in the non-depressed group and 51.8% in the depressed group. Additionally, the non-depressed group was significantly more likely to live in urban areas (75.3%) compared to the depressed group (70.0%) (p=.027).

The mean number of household members was 2.07 in the non-depressed group and 1.94 in the depressed group (p=.006). The number of grandchildren was 4.50 in the non-depressed group and 4.87 in the depressed group (p=.049). Participation in religious gatherings was significantly higher in the non-depressed group at 17.5% compared to 8.1% in the depressed group (p<.001). Similarly, participation in social gatherings (p<.001), leisure/cultural/sports-related organizations (p<.001), and alumni associations/hometown friendship associations/clan gatherings (p=.006) were all higher in the non-depressed group than in the depressed group.

The proportion of individuals with no participation in any gatherings was 33.8% in the non-depressed group compared to 47.4% in the depressed group (p<.001). Regarding participation frequency in social gatherings, the non-depressed group had the highest proportion (55.0%) attending 1~2 times per month, while the depressed group had equal proportions (38.8%) attending either 1~2 times per month or 1~6 times per year (p<.001).

There were no significant differences between the two groups in terms of sex (p=.128), whether they had grandchildren they helped raise before the grandchildren were 10 years old (p>.99), number of living children (p=.077), participation in religious gatherings (p=.185), participation in leisure/cultural/sports organizations (p=.509), or participation in alumni associations/hometown friendship associations/clan gatherings (p=.487).

2. Comparison of Health-Related Characteristics

The results of comparing health-related characteristics by depression status are shown in Table 2. There was a significant difference in self-reported health status, with 69.3% of non-depressed respondents and 84.4% of depressed respondents reporting poor health (p<.001). The proportion of individuals reporting activity (work) limitations due to a health condition was 44.3% in the non-depressed group compared to 59.3% in the depressed group (p<.001). The proportion of individuals without a disability was significantly higher in the non-depressed group (99.6%) compared to the depressed group (97.5%) (p<.001).

Table 2.

Health-Related Characteristics (N=1,879)

Regarding hearing aid use, 95.5% of the non-depressed group and 91.6% of the depressed group reported not using their usual hearing aids (p=.003). The proportion of individuals with good oral health was significantly higher in the non-depressed group (87.0%) compared to the depressed group (73.1%) (p<.001). Difficulty with daily activities due to pain was also significantly different, with 67.3% of the non-depressed group reporting no difficulty, compared to only 46.6% in the depressed group (p<.001).

Regular exercise was significantly more common in the non-depressed group (55.8%) than in the depressed group (35.0%) (p<.001). Both groups had a high proportion of current drinkers, with 27.3% in the non-depressed group and 18.9% in the depressed group (p<.001). For ADLs and IADLs, the proportion of individuals able to perform them independently was significantly higher in the non-depressed group (93.9% and 79.6%) than in the depressed group (82.4% and 64.1%) (p<.001). There were no significant differences between the two groups in terms of whether current vision made daily activities difficult (p=.254), whether they usually wore dentures (p=.953), or whether they were current smokers (p=.321).

3. Retirement-Related Characteristics

When comparing satisfaction with retirement and life before and after retirement based on depression status, both groups were more likely to report being satisfied with retirement. However, satisfaction was significantly higher in the non-depressed group (75.7%) compared to the depressed group (62.6%) (p<.001).

Regarding perceptions of life before and after retirement, both groups most commonly reported that their lives were similar before and after retirement. This response was significantly higher in the non-depressed group (60.9%) compared to the depressed group (53.3%) (p<.001).

4. The Depression Prediction Model for the Retired Elderly

The model for predicting depression among retired seniors is shown in Figure 1. Regular exercise at least once a week was found to have a significant effect on depression (F=59.41, p<.001). Among those who exercised regularly (Node 1), 16.7% were depressed. Additionally, oral health-related quality of life had a significant impact on depression (F=9.86, p=.002). Among individuals with good oral health (Node 3), 15.4% were depressed, whereas 28.3% of those with poor oral health (Node 4) experienced depression.

Figure 1.

Prediction model of depression in retired elderly. Adj.=Adjusted; ADL=Activities of daily living; df=Degree of freedom.

For those with good oral health-related quality of life, daily life limitations due to health conditions significantly influenced depression levels (F = 4.62, p=.032). Among those without limitations (Node 7), 13.5% were depressed, while 19.1% of those with limitations (Node 8) experienced depression.

Among individuals who did not engage in regular exercise (Node 2), the ability to perform ADL had a significant impact on depression (F=29.42, p<.001). Among those who were independent in performing daily activities (Node 5), 28.4% were depressed, while among those who were dependent (Node 6), the proportion was significantly higher at 51.4%.

For those with independent ADL, marital status had a significant effect on depression (F=24.25, p<.001). Depression was reported by 22.5% of married individuals (Node 9) and 39.2% of those who were separated, divorced, widowed, or single (Node 10). Among those who were dependent in performing daily activities (Node 6), sex differences were observed in depression levels (F=3.95, p=.047). The proportion of depressed females (Node 11) was 41.7%, while among males (Node 12), it was 58.5%.

The lowest rate of depression (13.5%) was found among individuals who exercised regularly, had high oral health-related quality of life, and experienced no daily life limitations due to health conditions. In contrast, the highest rate of depression (58.5%) was observed among those who did not exercise regularly, were dependent, and male.

5. The Depression Prediction Model for the Retired Elderly by Age

The depression prediction model for retired seniors by age is shown in Figure 2. Depression levels varied significantly by age (F=28.89, p<.001). Among the younger elderly (Node 1), 17.0% were depressed, with a significant difference based on self-reported health status (F=15.26, p<.001). Among those in fair or good health (Node 4) 6.6% were depressed whereas 23.5% of those in poor health (Node 5) experienced depression. Among those in poor health, depression levels also differed based on daily life limitations due to the health conditions (F=5.31, p=.021), with 17.8% of those without limitations (Node 10) and 32.1% of those with limitations (Node 11) experiencing depression.

Figure 2.

Prediction model of depression by age in retired elderly. Adj.=Adjusted; df=Degree of freedom.

Among the mid-older or older elderly (Node 2), 23.1% were depressed, with a significant difference in depression levels based on oral health-related quality of life (F=30.09, p<.001). Depression was reported by 20.5% of those with good oral health-related quality of life (Node 6) and 39.6% of those with poor oral health-related quality of life (Node 7). Among those with good oral health-related quality of life, depression levels varied depending on regular exercise (F=21.45, p<.001), with 15.4% of those who exercised regularly (Node 12) and 27.0% of those who did not (Node 13) experiencing depression.

Among the very old elderly (Node 3), 34.2% were depressed. Among those who exercised regularly (Node 8), 20.2% experienced depression, while among those who did not exercise regularly (Node 9), 40.1% were depressed. Among individuals who did not exercise regularly, depression levels also differed based on whether pain made daily activities difficult (F=8.93, p=.008). Among those without difficulty due to pain (Node 16), 31.0% experienced depression, while 49.6% of those with difficulty were depressed.

The lowest rate of depression (6.6%) was observed among younger elderly individuals in fair or good health. The highest rate of depression (55.4%) was found among mid-older or older elderly individuals with poor oral health-related quality of life who were separated, divorced, widowed, or single.

6. The Depression Prediction Model for the Retired Elderly by Gender

The prediction model of depression by sex among retired seniors are shown in Figure 3. There was no significant difference in depression between males and females (F=2.34, p=.126). Among females (Node 1), 25.9% were depressed, with a significant difference based on IADL performance (F=30.46, p<.001). Among those who were independent in IADLs (Node 3), 21.5% were depressed, whereas 42.2% of those who were dependent (Node 4) experienced depression. For those who were independent in IADL, depression levels varied depending on regular exercise (F=12.83, p<.001). Among those who exercised regularly (Node 7), 15.7% were depressed, compared to 27.4% of those who did not exercise. For those who were dependent on IADL, oral health-related quality of life had a significant effect on depression. Among those with good oral health-related quality of life (Node 9) 36.7% were depressed, compared to whereas among those with poor oral health-related quality of life, the proportion was significantly higher at 54.7%.

Figure 3.

Prediction model of depression by sex in retired elderly. Adj.=Adjusted; ADL=Activities of daily living; df=Degree of freedom; IADL=Instrumental activities of daily living.

Among males, 22.8% were depressed, with a significant difference based on the ability to perform ADL (F=56.26, p<.001). Among those who were independent in performing ADL (Node 5), marital status was a significant factor influencing depression (F=19.23, p<.001). Among married individuals (Node 11), 17.7% were depressed, whereas 34.7% of those who were separated, divorced, or widowed (Node 12) experienced depression.

Among males who were dependent in ADL (Node 6), 53.7% were depressed.

In this study, females who were instrumentally independent and exercised regularly had the lowest likelihood of depression at 15.7%, whereas males who were instrumentally dependent had the highest likelihood of depression at 53.7%.

7. Validation of Depression Prediction Models for the Retired Elderly

The validity analysis of the prediction models in this study showed that the training data of the prediction model without a first node had a Risk Estimate (RE) of .235, indicating a 76.5% probability of correct classification. The model in which age was set as the first node had an RE of .244 on the training data, indicating a 75.6% probability of correct classification. The model with sex as the first node had an RE of .226 on the training data, indicating a 77.4% probability of correct classification.

DISCUSSION

This study used the 9th Aging Research Panel Survey from 2022 to assess the depression status of the elderly population after retirement and develop prediction models to identify high-risk groups for depression among retired seniors. In the prediction model without a first node, regular exercise at least once a week was identified as the primary factor significantly affecting depression. Those who exercised regularly, had a high oral health-related quality of life, and experienced no daily living limitations were the least likely to be depressed. Conversely, individuals who did not engage in regular exercise, relied on daily routines, and were male were found to be the most likely to experience depression.

Previous studies examining the effects of regular exercise on depression in elderly populations [14,15] have found that older adults who engage in regular exercise exhibit lower levels of depression than those who do not, supporting the findings of this study. In particular, a study on leisure sports participation among male retirees [27] found that engaging in sporting activities had a positive impact on resocialization after retirement. However, research utilizing European Health, Ageing, and Retirement Data found lower levels of depression among elderly individuals, particularly those who participated in moderate or vigorous exercise [17]. Therefore, further studies are needed to explore the specific types, intensity, frequency, and long-term effects of exercise on depression among retirees undergoing major emotional and social transitions.

Consistent with our findings, a previous study found that limitations in daily activities significantly impacted depression in the elderly [16] and that better oral health was associated with lower levels of depression [17]. Additionally, prior research has shown that males in good health who continue to build their careers tend to have a higher quality of retirement and greater life satisfaction [8]. These findings align with the results of this study, as we found that males in poor health who do not exercise and rely on others for daily living are the most likely to experience depression.

In the prediction model with age as the first node, younger elderly individuals aged 65~69 years who were in fair health were the least likely to be depressed. In contrast, mid-older elderly individuals aged 70~74 years or older elderly aged 75~84 years with poor oral health-related quality of life and who were separated, divorced, widowed, or single were most likely to be depressed. Although not specifically focusing on retired seniors, a previous study found that better health was associated with lower levels of depression in the elderly [28], lower levels of and another study reported that oral health deteriorates with age and is linked to depression [17], which aligns with our findings. However, given the results of this study, it appears necessary to further categorize elderly individuals over the age of 65. Instead of treating them as a single group, subdividing them into more specific age categories may provide deeper insight into their depression levels, the factors influencing their mental health, and how these findings can be applied to depression management strategies.

Additionally, as physical and mental health status and oral health decline with age, longitudinal studies are needed to examine how these changes over time impact depression. Since marital status evolves over time and its effects on depression may vary by sex, further research is recommended to establish a clearer relationship between these variables.

In the predictive model with sex as the first node, there was no significant difference in depression based on sex. However, within the model, females who were independent in IADL and exercised regularly were the least likely to be depressed, while males who were dependent in performing IADL were the most likely to experience depression. Previous research has shown that male retirees have more difficulty adjusting to life after retirement compared to female retirees [22], which supports these findings. In particular, male retirees are more likely to suffer from depression in retirement, despite nearly twice as many female retirees being diagnosed with depression [29]. Despite previous studies indicating that depression in the elderly varies by sex [8,9], this study found no significant difference in depression by sex when the prediction model was implemented with sex as the first node. Therefore, further research is needed to determine whether the effect of sex on depression differs specifically among retired elderly individuals.

Additionally, a previous study on the impact of regular exercise on depression in the elderly found that those who exercised regularly were less likely to be depressed, supporting our findings that elderly individuals who did not engage in regular exercise were more likely to be diagnosed with depression [14].

Retired seniors have more time to engage in regular exercise than non-retired seniors, and they should be encouraged to do so as a means of reducing depression. However, it is essential to provide support in prescribing and implementing exercise programs that consider individual physical characteristics, taking into account factors such as health conditions and economic circumstances.

Consistent with the findings of this study, previous research has shown that dependence in IADL or ADL is significantly associated with depression [16], reinforcing the relevance of these factors in understanding depression among the retired elderly.

According to a study, elderly individuals tend to become more dependent on performing daily living and instrumental daily living activities as they age. When these abilities become limited, they are more susceptible to depression. Therefore, it is essential to assess depression levels alongside the decline in physical function among retired seniors and implement support programs to assist them in daily life. Additionally, ongoing monitoring of mood changes and performance levels is crucial to providing timely support and interventions.

Retirees experience a loss of role as they leave the workforce and cease to be economically and socially active. This loss of role can lead to a diminished sense of identity previously conferred by their professional position, negatively impacting psychological and emotional well-being [8]. Additionally, the loss of various social roles following retirement can result in feelings of worthlessness, which may undermine self-esteem and contribute to depressive symptoms, ultimately leading to difficulties in social adjustment [1].

However, depression in the elderly is often underestimated by society, and little research has been conducted specifically on depression among retired seniors. Since older adults may associate a depression diagnosis with having a mental illness, they may be reluctant to acknowledge their depressive symptoms. As a result, depression in the elderly may be more severe than social perceptions and existing research suggest, and it is likely to be even more pronounced among retired seniors undergoing drastic life changes. Therefore, continued research on depression in the retired elderly is necessary. Rather than relying solely on cross-sectional studies that provide a temporal snapshot, longitudinal studies should be conducted to track changes in depression and related factors over time, as depression is highly influenced by variables that evolve throughout the aging process.

This study had several limitations. First, the data were self-reported by the subjects, which may affect reliability and limit the accuracy of result interpretation. Second, while some participants may have experienced depression before retirement, this study could not account for the timing of depression onset.

Third, although previous studies have identified factors such as education level and sex as significant contributors to depression in the elderly, these variables were not significant in this study. Therefore, continued research on depression in retired seniors is needed to clarify these relationships.

Finally, this study focused on retired seniors but did not consider whether they were engaged in other economic activities for financial or personal reasons post-retirement, which represents a limitation in understanding the broader impact of retirement on depression.

CONCLUSION

This study aimed to develop prediction models for depressive status among retired seniors aged 65 and older and to identify high-risk individuals. The findings indicate that age, oral health, marital status, health status, IADL, ability to perform ADL, and regular exercise were all associated with depression. Specifically, older age, poor oral health, and being separated, divorced, widowed, or single were linked to higher rates of depression. Additionally, males who were dependent on others for daily living were more likely to experience depression.

Based on these findings, we recommend the development and implementation of community-based management programs to prevent and manage depression among the retired elderly. Furthermore, given that the elderly population undergoes continuous changes in age, physical health, mental health, and marital status, longitudinal studies are needed to examine the long-term effects of depression and its influencing factors over time.

Notes

Authors' contribution

MH contributed to the design and implementation of the research, to the analysis of the results and to the writing of the manuscript.

Conflict of interest

No existing or potential conflict of interest relevant to this article was reported.

Funding

None.

Data availability

Data of this study can be obtained from the website https://survey.keis.or.kr/klosa/klosa04.jsp.

Acknowledgements

None.

References

1. Um SR, Kim SK, Kim JW, Shin HR, Kim YS. The relationship between retirement and depression of elderly men: analysis of mediating effect of social contact and mediated moderation effect of utilizing electronic devices. Health and Social Welfare Review 2020;40(3):248–85. https://doi.org/10.15709/hswr.2020.40.3.248.

2. Yoon HS, Lee MJ. Gender, residential areas, retirement transitions, and psychological well-being of the elderly. Korean Journal of Social Welfare 2006;58(4):197–217.

3. Butterworth P, Gill SC, Rodgers B, Anstey KJ, Villamil E, Melzer D. Retirement and mental health: analysis of the Australian national survey of mental health and well-being. Social Science & Medicine 2006;62(5):1179–91. https://doi.org/10.1016/j.socscimed.2005.07.013.

4. Park JY, Yoo SY. Evaluation of preparing retirement in the households of adult children cohabiting with the working parents. Financial Planning Review 2018;11(3):57–77. https://doi.org/10.36029/FPR.2018.08.11.3.57.

5. Shin BK, Lee SJ. Factors influencing the length of time that it took for elderly to do suicidal ideation, suicidal plan and suicidal attempt after retirement in Korea. Social Welfare Policy 2016;43(2):1–33. https://doi.org/10.15855/swp.2016.43.2.1.

6. Lee SJ, Yun GH, Lee JY. Factors influencing period from retirement to occurrence of depression of elderlies. Journal of Humanities and Social Science 2017;8(5):905–26. https://doi.org/10.22143/HSS21.8.5.51.

7. Choi BJ, Lee JM. Effects of individual status and the community environment on geriatric depression: a multi-level analysis. The Korean Journal of Health Service Management 2021;15(3):51–65. https://doi.org/10.12811/kshsm.2021.15.3.051.

8. Quick HE, Moen P. Gender, employment, and retirement quality: a life course approach to the differential experiences of men and women. Journal of Occupational Health Psychology 1998;3(1):44–64. https://doi.org/10.1037/1076-8998.3.1.44.

9. Lim JC, Ju KH, Lim BW. Effects of work participation on the depression of the elderly; focusing on the mediating role of self-esteem. Journal of Welfare for the Aged 2012;57:29–50. https://doi.org/10.21194/kjgsw..57.201209.29.

10. Jung SR, Hur JS. The effects of community environment on depression among the elderly: focusing on comparison of metropolitan, medium-sized city, and rural area. Journal of Community Welfare 2022;80:49–91. https://doi.org/10.15300/jcw.2022.80.1.49.

11. Moon SJ, Lee GR, Nam EW. Related factors of depression according to individual attributes and regional environment: using multi-level analysis. Health Policy and Management 2020;30(3):355–65. https://doi.org/10.4332/KJHPA.2020.30.3.355.

12. Jeon GS, Jang SN, Park S. Social support, social network, and frailty in Korean elderly. Journal of the Korean Geriatrics Society 2012;16(2):84–94. https://doi.org/10.4235/jkgs.2012.16.2.84.

13. Xu X, Zhou Y, Su D, Dang Y, Zhang X. Does education influence life-course depression in middle-aged and elderly in China? Evidence from the China Health and Retirement Longitudinal Study (CHARLS). International Journal of Environmental Research and Public Health 2023;20(2):1256. https://doi.org/10.3390/ijerph20021256.

14. Nam I, Yoon H, Hyun D, Choi A, Yeom S. Effect of regular exercise on depression among Korean older adults. Korean Journal of Social Welfare 2014;66(4):5–23. https://doi.org/10.20970/kasw.2014.66.4.001.

15. Kang HJ, Lee JH, Yoon YJ. A systematic review study of the effects of elderly people’s participation in sports on depression. Journal of Sport and Leisure Studies 2022;87:243–54. https://doi.org/10.51979/KSSLS.2022.01.87.243.

16. Zhao L, Wang J, Deng H, Chen J, Ding D. Depressive symptoms and ADL/IADL disabilities among older adults from low-income families in Dalian, Liaoning. Clinical Interventions in Aging 2022;17:733–43. https://doi.org/10.2147/CIA.S354654.

17. Chun JY, Jung ES. Relationship between health factor, oral health factor and prevalence of depression in Korean elderly. Journal of Korean Society of Dental Hygiene 2015;15(6):963–71. https://doi.org/10.13065/JKSDH.2015.15.06.963.

18. Lee JH, Park KS, Kim RB, Kim BJ, Chun JH. The influence of individual-level social capital on depression. Journal of Agricultural Medicine and Community Health 2011;36(2):73–86. https://doi.org/10.5393/JAMCH.2011.36.2.073.

19. Kim SY, Yoon HY. Effects of retirement stress, character strengths, and marital dissatisfaction on post-retirement depression. Korean Journal of Culture Social Issues 2013;19(4):533–76.

20. Lee S, Lee J. The effect of retirement-related variables and satisfaction with family relationships on life satisfaction after retirement among the middle-aged and older adults: with a focus on gender differences. Korean Journal of Human Ecology 2021;30(2):259–76. https://doi.org/10.5934/kjhe.2021.30.2.259.

21. Lee JS, Kim HA. The impact family relationship on life of satisfaction of the elderly: the mediation effect of depression. Journal of the Korean Society for Wellness 2017;12(2):27–36. https://doi.org/10.21097/ksw.2017.05.12.2.27.

22. Kim MC, Park MK, Song MH. A study on the affected variables to the adjustment of retired men. Journal of the Korean Home Economics Association 2005;43(2):143–58.

23. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement 1977;1(3):385–401. https://doi.org/10.1177/014662167700100306.

24. Jang HE, Kim YJ, Heo MH, Jong SW, Noh JW. The relationship between social support and depression in older adults by gender and residence. The Korean Journal of Health Service Management 2022;16(3):101–13. https://doi.org/10.12811/kshsm.2022.16.3.101.

25. Seo J, Kim M. A prediction model for quality of life by resilience in disaster female victims. Korean Journal of Adult Nursing 2021;33(6):639–48. https://doi.org/10.7475/kjan.2021.33.6.639.

26. Choi JH, Seo DS. Decision trees and its applications. Journal of the Korean Official Statistics 1999;4(1):61–83.

27. Son JY. The correlation among the retired male seniors’ motivation to participate in leisure sports activities, re-socialization and their intention to maintain participation. Journal of the Korea Contents Association 2022;22(12):744–54. https://doi.org/10.5392/JKCA.2022.22.12.744.

28. Yeum D, Jung J. Research on subjective health status and satisfaction with life of elderly: focused on the mediating effect of depression and the moderating effect of social networking. Social Science Research Review 2014;30(4):27–50.

29. Lee A. Impact and implications of retirement on mental health and cognitive function. Health and Welfare Issue & Focus 2018;357:1–8.

Article information Continued

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

Characteristic	Category	No-depression (n=1,425, 75.8%)	Depression (n=454, 24.2%)	p-value
Age^*	Young-old	264 (18.5)	54 (11.9)	<.001
	Middle-old	300 (21.1)	80 (17.6)
	Old-old	630 (44.2)	200 (44.1)
	Super-old	231 (16.2)	120 (26.4)
Sex	Male	824 (57.8)	244 (53.7)	.128
	Female	601 (42.2)	210 (46.3)	.128
Marital status	Single	4 (0.3)	5 (1.1)	<.001
	Separation/divorce/widow	374 (26.2)	172 (37.9)
	Married	1,047 (73.5)	277 (61.0)
Education level	Elementary school	587 (41.2)	235 (51.8)	<.001
	Middle school	245 (17.2)	73 (16.1)
	High school	398 (27.9)	105 (23.1)
	University	195 (13.7)	41 (9.0)
Resident area	City	1,073 (75.3)	318 (70.0)	.027
	Province	352 (24.7)	136 (30.0)	.027
Care of grandchildren^†	No	6 (20.0)	1 (14.3)	>.99
	Yes	24 (80.0)	6 (85.7)	>.99
Number of family members		2.07±0.86	1.94±0.94	.006
Number of children		2.89±1.33	3.02±1.49	.077
Number of grandchildren		4.50±3.38	4.87±3.60	.049
Participation in religious gatherings	No	1,175 (82.5)	417 (91.9)	<.001
	Yes	250 (17.5)	37 (8.1)	<.001
Participation in social gatherings (senior citizen centers, etc.)	No	692 (48.6)	253 (55.7)	<.001
	Yes	733 (51.4)	201 (44.3)	<.001
Participation in group meetings (leisure, sports-related organizations, senior college, etc.)	No	1,279 (91.0)	442 (97.4)	<.001
	Yes	128 (9.0)	12 (2.6)	<.001
Participation in alumni and family meetings	No	1,240 (87.0)	417 (91.9)	.006
	Yes	185 (13.0)	37 (8.1)	.006
No meetings attended	No	944 (66.2)	239 (52.6)	<.001
	Yes	481 (33.8)	215 (47.4)	<.001
Frequency of participation in religious gatherings	None	5 (2.0)	3 (8.1)	.185
	1~3/week	21 (8.4)	3 (8.1)
	1~2/month	27 (10.8)	5 (13.5)
	1~6/year	197 (78.8)	26 (70.3)
Frequency of participation in social gatherings (senior citizen centers, etc.)	None	2 (0.3)	2 (1.0)	<.001
	1~3/week	56 (7.6)	43 (21.4)
	1~2/month	403 (55.0)	78 (38.8)
	1~6/year	272 (37.1)	78 (38.8)
Frequency of participation in group meetings (leisure, sports-related organizations, senior college, etc.)	None	0 (0.0)	0 (0.0)	.509
	1~3/week	1 (0.8)	0 (0.0)
	1~2/month	28 (21.9)	1 (8.3)
	1~6/year	99 (77.3)	11 (91.7)
Frequency of participation in alumni and family meetings	None	2 (1.1)	0 (0.0)	.487
	1~3/week	129 (69.7)	30 (81.1)
	1~2/month	51 (27.6)	7 (18.9)
	1~6/year	3 (1.6)	0 (0.0)

Characteristic	Category	No-depression (n=1,425, 75.8%)	Depression (n=454, 24.2%)	p-value
Health status	Bad	987 (69.3)	383 (84.4)	<.001
	Average	372 (26.1)	66 (14.5)
	Good	66 (4.6)	5 (1.1)
Restriction of daily life by health status	No	794 (55.7)	185 (40.7)	<.001
	Yes	631 (44.3)	269 (59.3)	<.001
Disability^*	No	1,275 (99.6)	394 (97.5)	<.001
	Yes	5 (0.4)	10 (2.5)	<.001
Loss of vision	No	1,355 (95.6)	425 (94.2)	.254
	Yes	63 (4.4)	26 (5.8)	.254
Hearing aid use	No	1,361 (95.5)	416 (91.6)	.003
	Yes	64 (4.5)	38 (8.4)	.003
Use dentures	No	989 (69.4)	314 (69.2)	.953
	Yes	436 (30.6)	140 (30.8)	.953
Oral health	Bad	185 (13.0)	122 (26.9)	<.001
	Good	1,240 (87.0)	332 (73.1)	<.001
Difficulty in daily life by pain	No	663 (67.3)	159 (46.6)	<.001
	Yes	322 (32.7)	182 (53.4)	<.001
Regular physical activity (over 1/week)	No	630 (44.2)	295 (65.0)	<.001
	Yes	795 (55.8)	159 (35.0)	<.001
Smoking	No	1,329 (93.3)	430 (94.7)	.321
	Yes	96 (6.7)	24 (5.3)	.321
Drinking	No	1,036 (72.7)	368 (81.1)	<.001
	Yes	389 (27.3)	86 (18.9)	<.001
ADL	Independent	1,338 (93.9)	374 (82.4)	<.001
	Dependent	87 (6.1)	80 (17.6)	<.001
IADL	Independent	1,134 (79.6)	291 (64.1)	<.001
	Dependent	291 (20.4)	163 (35.9)	<.001