Abstract
Background
While the impact of poor working conditions on workers' intention to leave the job is well-established, limited research has examined how the availability of pension benefits influences the association between adverse psychosocial work conditions and labor market exit for health reasons among middle-aged and older workers. This study explored the association of psychosocial and physical work conditions with labor market exit for health reasons among individuals with varying pension statuses.
Methods
This study utilized data from the Healthy Aging Longitudinal Study in Taiwan (HALST), which investigated reasons for labor market exit among 2,143 adults aged 55 and older. Work conditions were aggregated by occupation based on data from the Occupational Safety and Health Surveys, which included nationally representative employees, and subsequently linked to HALST data. We examined the differential impact of psychosocial and physical work conditions on health-related labor market exit, compared to old age retirement, among individuals with and without pension coverage.
Results
Among 2,143 study participants, 7.3% left the labor market due to health reasons, and 39.9% reported not having a pension. Individuals with low job control (adjusted odds ratio [aOR] = 2.23, 95% confidence interval [CI] = 1.05 to 4.73) and high physical demands (aOR = 2.72, 95% CI = 1.26 to 5.85) were more likely to exit the labor market for health reasons compared to old age retirement. Among participants without a pension, adverse work conditions were significantly associated with labor market exit for health reasons.
Conclusions
Adverse work conditions were associated with labor market exit for health reasons particularly among older adults without pension coverage. Implementing policies to improve psychosocial work conditions and enhance the pension system is warranted.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-025-22215-3.
Keywords: Pension, Population aging, Psychosocial work conditions, Retirement
Introduction
Population aging is a global phenomenon that poses multiple challenges to both the health and economic sustainability of societies. It is characterized not only by increasing healthcare and pension expenditures but also by shrinking workforces. According to the report by the World Bank [1], the East Asian and Pacific region is experiencing a particularly rapid aging process compared to other parts of the world. In Taiwan, for instance, the proportion of the population aged over 65 exceeded 14% in 2018 and is projected to exceed 20% by 2025. In response to these demographic and societal challenges, many European countries have implemented "extending working lives" policies since the 1990s [2–4]. In Taiwan, however, pension systems are less comprehensive, and policy reforms to address population aging have been implemented much later. The National Pension Insurance was established in 2008 to provide basic coverage for individuals not included in occupation-based pension schemes. In 2019, the Middle-aged and Elderly Employment Promotion Act was enacted to promote labor force participation among middle-aged and older individuals. Given Taiwan’s later but faster population aging, extending working lives has become an urgent policy challenge requiring innovative solutions beyond simply adopting European models.
The traditional concept of retirement is facing challenges due to the rise in nonstandard/temporary employment, and the rapid evolution of technology and required skills in the workplace. Raising the statutory retirement age has been one of the most common policies designed to foster longer working lives. Policies aimed at extending working lives need to go beyond merely postponing retirement ages [5]. In the context of improving work conditions, systematic reviews revealed that high physical demands, shift work, and low job control were associated with early retirement [6, 7]. However, there is high heterogeneity in the occupation backgrounds of study participants and variations in the measurements for work conditions.
Moreover, most existing studies have been conducted in Europe, where pension systems are generally more generous and comprehensive [8]. Only a few of studies have been conducted in countries with less developed pension systems. For instance, a study in China found that poor health is associated with labor market exit, particularly among farmers [9]. In Taiwan, pension systems are primarily occupation-based, supplemented by the National Pension Insurance. Generally, old-age pensions are not tied to recipients’ financial conditions, except for means-tested social assistance benefits. The design of pension systems can significantly influence work and retirement decisions and may contribute to income and health inequalities in old age [10]. In such case, extending working lives has raised concerns about forcing employees with poor health and low income to continue working to secure retirement benefits [5]. Some Taiwanese workers, particularly those in unstable jobs, small enterprises, or self-employment, do not receive pension benefits. These workers often endure adverse working conditions and a heightened risk of work-related illnesses and disabilities, which can impede their ability to work until old age. In contrast, employees with secure pension benefits tend to have stable jobs, better-regulated working conditions, lower health risks, and higher motivation to work until old age to become eligible for pension benefit upon retirement. However, to our knowledge, no research has examined how pension availability affects the labor market participation of older workers.
Health issues, whether physical or mental, emerge as a predominant obstacle to employment among older adults [11–14]. Higher level of work-related stress not only increased the likelihood of early labor market exit but also poor health [15]. Numerous studies have demonstrated the correlation between work conditions (e.g., high psychological job demands, low autonomy, limited influence, insufficient social support) and poor physical health in middle-aged and older adults [16, 17]. Notably, the significant effect of work-related stress on labor market exit disappears when accounting for workers' health status [18]. This suggests that health partially mediates the relationship between psychosocial work conditions and early labor market exit [19]. Nevertheless, the direct measurement of reasons for labor market exit remains limited, and obtaining such information would contribute to a more comprehensive knowledge that can inform the development of policies aimed at modifying work environment and thereby preventing labor market exit for health reasons among older employees.
While health issues are expected to influence labor market performance of older workers, few studies have examined the relationship between work conditions and labor market exit for health reasons. Additionally, there is a scarcity of studies available regarding the association between work conditions and labor market exit in East Asian countries, where population aging is occurring at a faster pace than elsewhere and pension system are less comprehensive. The study aim was to examine the influences of work conditions on labor market exit for health reasons, as opposed to old age retirement, among middle-aged and older adults with different pension statuses.
Methods
Study design
We employed data from the Healthy Aging Longitudinal Study in Taiwan (HALST), a nationally representative longitudinal cohort of older adults, and the Occupational Safety and Health Surveys (OSHS). This OSHS dataset was utilized to derive occupational-level work conditions for linkage with the HALST data. All data were de-identified before being handled to the authors.
The HALST study employed a systematic sampling method to enroll healthy community-dwellers aged 55 and above from seven selected districts in Taiwan. These seven locations encompassed both urban and rural areas, as well as to represent various ethnic groups, thereby reflecting the diverse sociodemographic characteristics of the Taiwanese population [20]. The initial wave of the survey was conducted from 2009 to 2013, followed by a second wave of data collection from 2014 to 2019. A total of 5,663 participants completed the first wave of the survey, while 4,233 participants completed the second wave. In this study, data from both waves of the survey were utilized.
The OSHS, conducted by the Ministry of Labor of Taiwan, has been administered to a nationwide working population every 3–5 years since 1988, with the goal of gaining insights into workplace safety and health conditions. For each survey of the OSHS, participants were selected using a two-stage random sampling process. In the first stage, all districts and villages throughout Taiwan were grouped into strata based on their levels of urbanization, and a random sample of districts and villages was selected from each stratum. In the second stage, households were then randomly selected within each district or village. Individuals who were currently employed at the time of the survey were identified from the sampled households. Participation was limited to individuals who were economically active, and those who were not economically active were deemed ineligible. The interviews were conducted face-to-face by trained interviewers. The sample's representativeness was carefully ensured, and detailed information regarding the survey's sampling methodology can be found elsewhere [21].
Measurement
Employment and reasons for labor market exit
In HALST, participants were asked about their work history and whether they were currently working at the time of the survey. Individuals who had never worked or were employers were excluded from wave 1 survey (N = 692 and 2091, Fig. 1). Those who had been employees self-reported their last primary occupation, defined as “the job you have worked at the longest or the main source of your income.” A total of 2307 participants in wave 1 and another 236 participants in wave 2 reported having worked previously but were not currently employed (Fig. 1). They were then asked to select the reason for their labor market exit from: old age retirement, health reasons, work-related issues (being laid off, difficulty adapting to work, or unsatisfactory pay), family-related reasons (marriage or caregiving burden), and other unspecified factors. To examine the work conditions as risk factors for labor market exit for health reasons compared to old age retirement, further exclusions were made for those who exited the labor market for reasons other than old age retirement or health reasons (n = 400), resulting in 2,143 participants for the analysis. If both old-age retirement and health reasons were selected (n = 25), participants were categorized into the old-age retirement group.
Fig. 1.
Flow chart for study sample selection from the Healthy Aging Longitudinal Study in Taiwan cohort
Work conditions
In the OSHS in 2007, the occupations were categorized according to the 5th edition of the Standard Occupational Classification by the Ministry of Labor of Taiwan, which aligns with the International Standard Classification of Occupations (ISCO-88) using two-digit major and sub-major groups. A job exposure matrix was created by aggregating work conditions for 35 occupations using data from 3,148 individuals [22], by averaging scores for each work condition among employees aged 50 and above (supplementary Table 1). The response rates for this survey wave was 86%. Participants provided information regarding their total working hours for the week preceding the survey. The psychosocial work conditions measured in OSHS surveys were psychological job demands, job control, workplace justice, and physical demands. Psychological job demands and job control were assessed using the validated Chinese version of the Job Content Questionnaire (JCQ) based on the job-demand-control (JDC) model developed by Karasek and Theorell [23, 24]. The OSHS questionnaires in 2007 included nine items for the job control scale (learning new things, non-repetitive work, creative work, allowing own decision, high level of skills, freedom to make decisions, various tasks, influential opinions, and the development of one’s abilities), as well as seven items for psychological job demands (fast work, hard work, excessive work, insufficient time, concentration on the job for a long time, hectic work, and insufficient manpower). The items were presented as statement, and respondents were asked to indicate their level of agreement on a four-point Likert scale ranging from 1 (strongly disagree) to 4 (strongly agree).
A seven-item scale for the assessment of workplace justice (trust, reliable information, fair work arrangements, fair rewards, fair performance evaluation, inclusion of information in decision-making, and respect) was utilized. This scale was modified from the original standard questionnaire and showed good psychometric properties [25, 26]. Responses for these items were recorded on a four-point Likert scale, ranging from one (strongly disagree) to four (strongly agree).
One item assessing physical demands (my job requires a lot of physical effort) was evaluated using a four-point Likert scale, with responses ranging from one (strongly disagree) to four (strongly agree). The scores for job control, psychological demands, workplace justice, and physical demands were summed, standardized, and subsequently divided into low and high groups based on the median.
Socioeconomic and health variables
In HALST, pension availability was evaluated by asking participants about their primary financial sources in the previous year. Participants were required to select only one option from the following: retirement pensions or National Pension, support from children or family members, rental or stock dividends, and social welfare benefits or others. Individuals who obtained retirement pensions or National Pension were classified as having a pension, regardless of other financial resources. Those who did not receive either type of pension were categorized as having no pension.
Participants self-reported their gender, birth year, educational years, marital status, household income, and pension availability. Education level was categorized into two groups: primary education or less, and secondary education or above. Marital status was grouped into married and other status (separated, divorced, widowed, or single). Participants provided information about their household income by selecting one of seven predefined categories. However, 42.2% of participants either left this information as unknown or chose not to disclose it. To address this, a dummy variable was generated to account for the missing household income data. Additionally, participants were categorized into two groups—high and low household income—using a threshold of 30,000 New Taiwan Dollars per month (equivalent to approximately 945 U.S. Dollars).
Mental health and physical health were both assessed through questionnaires. Depressive symptoms were assessed using the 20-item Center for Epidemiological Studies-Depression Scale (CES-D) [27]. Response options for each item ranged from 0 to 3. Scores on the scale could range from 0 to 60, with higher scores indicating a higher level of depressive symptoms. Physical illnesses were categorized based on whether participants reported a history of two or more of the major diseases: hypertension, diabetes, cardiovascular disease, stroke, or cancer.
Statistical analysis
We conducted Chi-square tests and Mann–Whitney tests to compare the differences between old age retirement and labor market exit for health reasons. Because HALST participants were nested within occupations, we employed a mixed regression analysis with fixed effects for work conditions on labor market exit for health reasons. This model included work conditions that have been reported to be associated with health, namely job control, psychological job demands, physical demands, workplace justice, and weekly working hours. The odds ratio (OR) and 95% confidence interval (CI) were estimated, and adjusted models were utilized, controlling for gender, age, education, marital status, household income, and pension availability. To investigate how pension availability influenced the impact of work conditions on labor market exit, we further stratified the participants by their pension availability and examined the associations between work conditions and labor market exit for health reasons separately. Data analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC, USA).
Results
Among the 2,143 study participants, 156 (7.3%) reported labor market exit for health reasons. Compared to old age retirement, participants who exited the labor market for health reasons were generally younger, had a higher proportion of female, and a lower level of education, being unmarried, had lower household income, and lacked pension coverage (Table 1). Additionally, they reported higher depressive symptom scores and a higher prevalence of having physical illness when compared to those who retired due to old age. Moreover, participants who exited the labor market for health reasons were more likely to have jobs characterized by low workplace justice, low job control, and higher physical demands. Notably, 60.15% of all participants reported having a pension.
Table 1.
Characteristics of individuals who exited labor market for health reasons or old age retirement
| All | Health reasons | Old age retirement | p | |
|---|---|---|---|---|
| (n = 2143) | (n = 156) | (n = 1987) | ||
| N (%) | N (%) | N (%) | ||
| Demographic characteristics | ||||
| Gender | < 0.001 | |||
| Female | 849 (39.62) | 92 (58.97) | 757 (38.10) | |
| Male | 1294 (60.38) | 64 (41.03) | 1230 (61.90) | |
| Age (year) | < 0.001 | |||
| < 65 | 579 (27.02) | 70 (44.87) | 509 (25.62) | |
| ≥ 65 | 1564 (72.98) | 86 (55.13) | 1478 (74.38) | |
| Education | < 0.001 | |||
| Primary or lower | 802 (37.44) | 106 (67.95) | 696 (35.05) | |
| Secondary or above | 1340 (62.56) | 50 (32.05) | 1290 (64.95) | |
| Marital status | < 0.001 | |||
| Married | 1624 (75.78) | 93 (59.62) | 1531 (77.05) | |
| Other status | 519 (24.22) | 63 (40.38) | 456 (22.95) | |
| Socioeconomic status | ||||
| Household income | 0.002 | |||
| High | 643 (30.00) | 33 (21.15) | 610 (30.70) | |
| Low | 596 (27.81) | 61 (39.10) | 535 (26.93) | |
| Missing | 904 (42.18) | 62 (39.74) | 842 (42.38) | |
| Pension | < 0.001 | |||
| No | 854 (39.85) | 88 (56.41) | 766 (38.55) | |
| Yes | 1289 (60.15) | 68 (43.59) | 1221 (61.45) | |
| Health conditions | ||||
| CES-D scores (mean and SD) | 4.05 (6.22) | 8.83 (9.67) | 3.68 (5.70) | < 0.001 |
| Physical illness | 0.004 | |||
| Yes | 592 (27.62) | 59 (37.82) | 533 (26.82) | |
| No | 1551 (72.38) | 97 (62.18) | 1454 (73.18) | |
| Work conditions | ||||
| Psychological job demands | 0.158 | |||
| High | 1085 (50.63) | 70 (44.87) | 1015 (51.08) | |
| Low | 1058 (49.37) | 86 (55.13) | 972 (48.92) | |
| Job control | < 0.001 | |||
| High | 1041 (48.58) | 27 (17.31) | 1014 (51.03) | |
| Low | 1102 (51.42) | 129 (82.69) | 973 (48.97) | |
| Physical demands | < 0.001 | |||
| High | 1064 (49.65) | 130 (83.33) | 934 (47.01) | |
| Low | 1079 (50.35) | 26 (16.67) | 1053 (52.99) | |
| Workplace justice | < 0.001 | |||
| High | 1063 (49.60) | 52 (33.33) | 1011 (50.88) | |
| Low | 1080 (50.40) | 104 (66.67) | 976 (49.12) | |
| Working hours (mean and SD) | 42.25 (3.59) | 42.38 (3.39) | 42.24 (3.60) | 0.990 |
In the adjusted regression model (Table 2), study participants who had experienced low job control (adjusted OR = 2.23, 95% CI = 1.05 to 4.73, p = 0.037) and high physical demands (adjusted OR = 2.72, 95% CI = 1.26 to 5.85, p = 0.012) were more likely to exit the labor market for health reasons rather than due to old age retirement. Having no pension was associated with labor market exit for health reasons as well (adjusted OR = 1.74, 95% CI = 1.19 to 2.54, p = 0.005).
Table 2.
The association of working conditions and pension availability with labor market exit for health reasons
| Crude model | Adjusted model | p | ||
|---|---|---|---|---|
| Independent variables | OR (95%CI) | OR (95%CI) | ||
| Psychological Job demands | 0.773 | 0.398 | ||
| Low | 1 | 1 | ||
| High | 0.89 (0.41–1.97) | 0.80 (0.47–1.36) | ||
| Job control | < 0.001 | 0.037 | ||
| High | 1 | 1 | ||
| Low | 5.07 (2.64–9.72) | 2.23 (1.05–4.73) | ||
| Physical demands | < 0.001 | 0.012 | ||
| Low | 1 | 1 | ||
| High | 4.98 (2.70–9.20) | 2.72 (1.26–5.85) | ||
| Workplace justice | 0.071 | 0.534 | ||
| High | 1 | 1 | ||
| Low | 2.04 (0.94–4.43) | 0.82 (0.44–1.55) | ||
| Working hours | 0.98 (0.87–1.10) | 0.741 | 0.95 (0.87–1.03) | 0.194 |
| Pension | 0.001 | 0.005 | ||
| Yes | 1 | 1 | ||
| No | 1.97 (1.37–2.84) | 1.74 (1.19–2.54) | ||
The model’s outcome was labor market exit for health reasons, as opposed to old age retirement. Odds ratios (OR) and 95% confidence intervals (CI) were shown in univariate crude models and mutually adjusted models, with additional adjustments for gender, age, education, marital status, and household income
Compared to those who had pension, participants without pension had higher proportion of low job control, high physical demands, and low workplace justice (Supplementary Table 2). We further stratified the participants by their pension status in the logistic regression analysis (Table 3). Among older adults who had no pension coverage, those who had experienced low job control (OR = 4.58, 95% CI = 1.80 to 11.71, p = 0.002) and high physical demands (OR = 3.03, 95% CI = 1.28 to 7.17, p = 0.013) were more likely to exit the labor market for health reasons rather than due to old age retirement, but psychological job demands showed a negative association (OR = 0.52, 95% CI = 0.30 to 0.90, p = 0.020). Among those with pension, work conditions were not significantly associated with labor market exit for health reasons.
Table 3.
Association between working conditions and labor market exit for health reasons, stratified by pension availability
| No pension (N = 854) | Pension (N = 1289) | ||||
|---|---|---|---|---|---|
| Work conditions | OR (95%CI) | p | OR (95%CI) | p | |
| Psychological Job demands | 0.020 | 0.693 | |||
| Low | 1 | 1 | |||
| High | 0.52 (0.30–0.90) | 1.17 (0.53–2.60) | |||
| Job control | 0.002 | 0.783 | |||
| High | 1 | 1 | |||
| Low | 4.58 (1.80–11.71) | 0.87 (0.30–2.49) | |||
| Physical demands | 0.013 | 0.117 | |||
| Low | 1 | 1 | |||
| High | 3.03 (1.28–7.17) | 2.43 (0.79–7.47) | |||
| Workplace justice | 0.052 | 0.316 | |||
| High | 1 | 1 | |||
| Low | 0.51 (0.26–1.01) | 1.61 (0.62–4.18) | |||
| Working hours | 0.93 (0.84–1.02) | 0.117 | 0.96 (0.85–1.09) | 0.543 | |
The model’s outcome was labor market exit for health reasons, as opposed to old age retirement. Models were mutually adjusted for gender, age, education, marital status, and household income
Given that participants who exited the labor market for health reasons were younger than the age of those who reported old age retirement, we stratified the participants by age (< 65 and ≥ 65 years old) and examined the association in each group separately (Table 4). High physical demands were associated with labor market exit for health reasons in the older group (OR = 3.21, 95% CI = 1.15 to 8.93, p = 0.027), while low job control was associated with labor market exit for health reasons in the younger group (OR = 3.68, 95% CI = 1.28 to 10.63, p = 0.018).
Table 4.
Association between working conditions and labor market exit for health reasons, stratified by age
| Age < 65 (N = 579) | Age ≥ 65 (N = 1564) | ||||
|---|---|---|---|---|---|
| Work conditions | OR (95%CI) | p | OR (95%CI) | p | |
| Psychological Job demands | 0.345 | 0.551 | |||
| Low | 1 | 1 | |||
| High | 0.69 (0.32–1.51) | 0.83 (0.44–1.57) | |||
| Job control | 0.018 | 0.464 | |||
| High | 1 | 1 | |||
| Low | 3.68 (1.28–10.63) | 1.43 (0.54–3.78) | |||
| Physical demands | 0.079 | 0.027 | |||
| Low | 1 | 1 | |||
| High | 2.58 (0.89–7.47) | 3.21 (1.15–8.93) | |||
| Workplace justice | 0.410 | 0.500 | |||
| High | 1 | 1 | |||
| Low | 0.69 (0.27–1.73) | 0.77 (0.36–1.68) | |||
| Working hours | 0.88 (0.78–1.00) | 0.050 | 0.99 (0.89–1.09) | 0.765 | |
The model’s outcome was labor market exit for health reasons, as opposed to old age retirement. Models were mutually adjusted for gender, education, marital status, and household income
Discussion
Compared to old age retirement, low job control and high physical demands were risk factors for labor market exit for health reasons. Approximately 40% of older adults did not have any pension at the time of labor market exit, and no pension coverage was associated with labor market exit for health reasons. Additionally, work conditions were found to be associated with labor market exit for health reasons specifically among older adults who had no pension, but not those who were receiving pension benefits.
As expected, this study reveals a correlation between high physical demands and labor market exit for health reasons, particularly among older workers. As workers age, physical demands increasingly become a barrier to extended working lives. Previous studies also demonstrated that senior workers engaged in primarily physical demanding jobs tend to exhibit higher stress scores [28]. Furthermore, research indicates that a favorable change in physical workload lowered the risk of exiting paid employment [29]. Beyond physical demands, psychosocial work conditions play a crucial role in maintaining older workers in the workplace. Even for senior workers engaged in physically demanding jobs, psychosocial work stress has been associated with increased perceived stress and early retirement [30–33]. According to the JDC model, the combination of low job control and low job demands characterizes passive work, which is associated with adverse mental and physical health [34, 35]. This finding echoes our observation that low job control and low psychological demands were associated with labor market exit for health reasons. With aging, the importance of job control, including skill use and decision authority, in predicting prolonged working lives increased [36]. This study found that low job control significantly impacted participants younger than 65 years old, a group that often occupies high occupational positions and typically experiences high job control. Previous research has demonstrated an association of low job control with poor physical and mental health outcomes [37, 38], which may consequently impede an individual's ability to work until traditional retirement age. Furthermore, decision authority, a component of job control, buffers the negative impact of heavy physical demands on working longer among older men [39, 40]. Therefore, to retain older workers in the labor market, implementing work redesign strategies focused on enhancing skill utilization and fostering an environment that promotes decision authority may be beneficial.
While high psychological job demands are commonly perceived as work stress, this study found an association with a lower risk of labor market exit for health reasons among individuals without pension support. Similarly, a previous study also noted that a decrease in psychological job demands increased the likelihood of exiting paid employment [29]. It has been proposed that when an older worker experiences physical functional limitation, psychological job demands may become the last resource to prevent exiting from paid employment [29]. High psychological demands, a component of active jobs within the JDC model, may protect workers from dementia through cognitive stimulation [34], thereby reducing the risk of labor market exit for health reasons. Furthermore, psychological job demands were generally more pronounced for workers with higher occupational positions and educational levels [41–45]. In this study, participants without pensions were younger, had lower levels of education, and reported higher physical demands compared to those with pensions. Among participants without pensions, jobs with low psychological demands often involved high physical demands, potentially increasing the risk of labor market exit due to health reasons.
In this study, having no pension was associated with labor market exit for health reasons, likely because pension eligibility is tied to old age retirement. Additionally, adverse work conditions were associated with labor market exit for health reasons in the no-pension group. A cumulative social disadvantage in their work life was evident in individuals without pension coverage who also experienced poor health. In other words, individuals experiencing greater marginalization in the labor market are less likely to reap benefits from the pension system in their later years. Previous studies have revealed a similar predicament among low-educated workers, who experienced more lost working years due to unemployment, lack of income, and disability benefits [44, 46]. As the occurrence of chronic diseases was highest among low educated workers, the probabilities of exiting paid employment through disability benefits were high among this group [47]. The associations between physical demands, autonomy, and job strain and health were also strongest in the lower educated workers [48]. The findings of this study suggest an accumulation of disadvantages of adverse work conditions, poor health, and financial insecurity in old age.
The strength of this study is that we used data from a nationally representative cohort of middle-aged and older adults. Using self-reported reasons for labor market exit has both pros and cons. It is subject to recall bias, especially among those who retired a substantial time before the survey. Nevertheless, we captured employees who were not accessible in registered database of labor statistics, such as informal employment and irregular employment. This is evidenced by the high proportion of participants receiving no pension, and the well-being of marginalized employees deserves more attention. This study has limitations. First, because work conditions were not collected in the HALST cohort, they were aggregated by occupation from another national survey (the OSHS). There could be variations across employees in the same occupation, which introduce bias. Extreme work conditions may have been eliminated during the aggregation process, thus limiting the variations. Second, work stress was evaluated using the JDC model in this study, omitting other useful work stress models. For example, health mediated the association between work stress, measured by effort-reward imbalance, and early labor market exit [18]. Including more work stress measurements in this study would provide a more comprehensive picture of the association between work conditions and labor market exit. Third, the cross-sectional design of this study poses limitations on information including the precise timing of labor market exits and exposure to work conditions, as well as whether the exits were temporary or permanent.
In conclusion, to mitigate early labor market exit for health reasons and to encourage the retention of workers in their later years, organizational interventions and labor policies should strive to promote active job design while reducing physical demands. This study also underscores the necessity for targeted policies for individuals lacking secure pension coverage, such as older adults with fragmented work tenure and non-standard employment, to support for continued employment and to establish an inclusive pension system.
Supplementary Information
Acknowledgements
Not applicable.
Abbreviations
- CES-D
Center for Epidemiological Studies-Depression Scale
- CI
Confidence Interval
- HALST
Healthy Aging Longitudinal Study in Taiwan
- ISCO
International Standard Classification of Occupations
- JCQ
Job Content Questionnaire
- OR
Odds Ratio
- OSHS
Occupational Safety and Health Surveys
Authors’ contributions
WJ Cheng and HY Chiang contributed to the study's conception and drafted the paper. YC Yang gained access to the data and performed statistical analyses. CC Hsu supervised the data analyses. Y Cheng and HM Hasselhorn critically provided critical revisions for significant intellectual content. All authors have approved the submitted version.
Funding
This work was supported by National Health Research Institutes [grant number 12A1-CGGP06-051]. The funders have no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Data availability
The data that support the findings of this study are available from the National Health Research Institutes of Taiwan. The availability of these data is subject to restrictions, and their use in this study was permitted under license.
Declarations
Ethics approval and consent to participate
This study has been approved by the ethics committee of the National Health Research Institutes (EC1090903-R4). Informed consents were obtained from all participants of the Occupational Safety and Health Surveys and the Healthy Aging Longitudinal Study in Taiwan. This study adhered to the principles of the Declaration of Helsinki.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.World Bank, International Monetary Fund: Global Monitoring Report 2015/2016: Development Goals in an Era of Demographic Change. In. Washington, DC; 2016.
- 2.Hasselhorn HM. Social inequality in the transition from work to retirement. In: Theorell T, editor. Handbook of Socioeconomic Determinants of Occupational Health - From Macro-level to Micro-level Evidence. Berlin, New York: Springer Heidelberg; 2020. p. 105–30. 10.1007/978-3-030-31438-5_32.
- 3.Foster L, Walker A. Active and successful aging: A European policy perspective. Gerontologist. 2015;55(1):83–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hasselhorn HM, Apt W, editors. Understanding employment participation of older workers: Creating a knowledge base for future labour market challenges. Federal Ministry of Labour and Social Affairs (BMAS) and Federal Institute for Occupational Safety and Health (BAuA). Berlin: BMAS/BAuA; 2015. 118 pages (Research Report). https://www.jp-demographic.eu/wp-content/uploads/2015/07/JPIUEP_Brochure1.pdf.
- 5.Phillipson C. ‘Fuller’ or ‘extended’ working lives? Critical perspectives on changing transitions from work to retirement. Ageing Soc. 2019;39(3):629–50. [Google Scholar]
- 6.Browne P, Carr E, Fleischmann M, Xue B, Stansfeld SA. The relationship between workplace psychosocial environment and retirement intentions and actual retirement: a systematic review. Eur J Ageing. 2019;16(1):73–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Scharn M, Sewdas R, Boot CRL, Huisman M, Lindeboom M, van der Beek AJ. Domains and determinants of retirement timing: A systematic review of longitudinal studies. BMC Public Health. 2018;18(1):1083. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.International Labour Organization: World Employment and Social Outlook 2015: The Changing Nature of Jobs. In. Geneva; 2015.
- 9.Li X, Zhang W, Sun H, Anis AH. Effects of health status on work exit and absenteeism among the older working population in China: a secondary analysis of a cohort sample. BMJ Open. 2019;9(9): e024115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Finley M: Getting Taiwanese to Engage in Retirement Planning. In: Taiwan Business Topics. Taipei, Taiwan: American Chamber of Commerce in Taiwan; 2019.
- 11.Ayalon L, Perel-Levin S, Georgantzi N, Lima CM. Participation of Older Persons With Mental Health Conditions and Psychosocial Disabilities in the Labor Market. Am J Geriatr Psychiatry. 2021;29(10):1033–7. [DOI] [PubMed] [Google Scholar]
- 12.van den Berg T, Schuring M, Avendano M, Mackenbach J, Burdorf A. The impact of ill health on exit from paid employment in Europe among older workers. Occup Environ Med. 2010;67(12):845–52. [DOI] [PubMed] [Google Scholar]
- 13.van den Berg TI, Elders LA, Burdorf A. Influence of health and work on early retirement. J Occup Environ Med. 2010;52(6):576–83. [DOI] [PubMed] [Google Scholar]
- 14.van Rijn RM, Robroek SJ, Brouwer S, Burdorf A. Influence of poor health on exit from paid employment: a systematic review. Occup Environ Med. 2014;71(4):295–301. [DOI] [PubMed] [Google Scholar]
- 15.Thorsen SV, Jensen PH, Bjorner JB. Psychosocial work environment and retirement age: a prospective study of 1876 senior employees. Int Arch Occup Environ Health. 2016;89(6):891–900. [DOI] [PubMed] [Google Scholar]
- 16.Leijten FR, van den Heuvel SG, van der Beek AJ, Ybema JF, Robroek SJ, Burdorf A. Associations of work-related factors and work engagement with mental and physical health: a 1-year follow-up study among older workers. J Occup Rehabil. 2015;25(1):86–95. [DOI] [PubMed] [Google Scholar]
- 17.Burr H, Hasselhorn HM, Kersten N, Pohrt A, Rugulies R. Does age modify the association between psychosocial factors at work and deterioration of self-rated health? Scand J Work Environ Health. 2017;43(5):465–74. [DOI] [PubMed] [Google Scholar]
- 18.Toczek L, Peter R. Investigating the influence of work-related stress on early labour market exit: the role of health. Eur J Ageing. 2023;20(1):31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Toczek L, Bosma H, Peter R. Early retirement intentions: the impact of employment biographies, work stress and health among a baby-boomer generation. Eur J Ageing. 2022;19(4):1479–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hsu CC, Chang HY, Wu IC, Chen CC, Tsai HJ, Chiu YF, Chuang SC, Hsiung WC, Tsai TL, Liaw WJ, et al. Cohort Profile: The Healthy Aging Longitudinal Study in Taiwan (HALST). Int J Epidemiol. 2017;46(4):1106–1106j. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Cheng Y, Chen CW, Chen CJ, Chiang TL. Job insecurity and its association with health among employees in the Taiwanese general population. Soc Sci Med. 2005;61(1):41–52. [DOI] [PubMed] [Google Scholar]
- 22.Descatha A, Fadel M, Sembajwe G, Peters S, Evanoff BA. Job-Exposure Matrix: A Useful Tool for Incorporating Workplace Exposure Data Into Population Health Research and Practice. Front Epidemiol. 2022;2: 857316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Karasek R. Theorell Tr: Healthy work : stress, productivity, and the reconstruction of working life. New York: Basic Books; 1990. [Google Scholar]
- 24.Cheng Y, Luh WM, Guo YL. Reliability and validity of the Chinese version of the Job Content Questionnaire in Taiwanese workers. Int J Behav Med. 2003;10(1):15–30. [DOI] [PubMed] [Google Scholar]
- 25.Cheng Y, Huang HY, Li PR, Hsu JH. Employment insecurity, workplace justice and employees’ burnout in Taiwanese employees: a validation study. Int J Behav Med. 2011;18(4):391–401. [DOI] [PubMed] [Google Scholar]
- 26.Donovan MA, Drasgow F, Munson LJ. The Perceptions of Fair Interpersonal Treatment Scale: development and validation of a measure of interpersonal treatment in the workplace. J Appl Psychol. 1998;83(5):683–92. [DOI] [PubMed] [Google Scholar]
- 27.Radloff LS. The CES-D Scale: A Self-Report Depression Scale for Research in the General Population. Appl Psychol Meas. 1977;1(3):385–401. [Google Scholar]
- 28.Vinstrup J, Sundstrup E, Andersen LL. Psychosocial stress and musculoskeletal pain among senior workers from nine occupational groups: Cross-sectional findings from the SeniorWorkingLife study. BMJ Open. 2021;11(3): e043520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Schram JLD, Robroek SJW, Ots P, Brouwer S, Burdorf A, van Zon SKR, Oude Hengel KM. Influence of changing working conditions on exit from paid employment among workers with a chronic disease. Occup Environ Med. 2020;77(9):628–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Vinstrup J, Meng A, Sundstrup E, Andersen LL: The Psychosocial Work Environment and Perceived Stress among Seniors with Physically Demanding Jobs: The SeniorWorkingLife Study. Int J Environ Res Public Health 2021, 18(14). [DOI] [PMC free article] [PubMed]
- 31.Gyllensten K, Toren K, Hagberg M, Soderberg M. A sustainable working life in the car manufacturing industry: The role of psychosocial factors, gender and occupation. PLoS ONE. 2020;15(5): e0233009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Stansfeld SA, Carr E, Smuk M, Clark C, Murray E, Shelton N, Head J. Mid-life psychosocial work environment as a predictor of work exit by age 50. PLoS ONE. 2018;13(4): e0195495. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hintsa T, Kouvonen A, McCann M, Jokela M, Elovainio M, Demakakos P. Higher effort-reward imbalance and lower job control predict exit from the labour market at the age of 61 years or younger: evidence from the English Longitudinal Study of Ageing. J Epidemiol Community Health. 2015;69(6):543–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Kivimaki M, Walker KA, Pentti J, Nyberg ST, Mars N, Vahtera J, Suominen SB, Lallukka T, Rahkonen O, Pietilainen O, et al. Cognitive stimulation in the workplace, plasma proteins, and risk of dementia: three analyses of population cohort studies. BMJ. 2021;374: n1804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Lerner DJ, Levine S, Malspeis S, D’Agostino RB. Job strain and health-related quality of life in a national sample. Am J Public Health. 1994;84(10):1580–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Stengard J, Leineweber C, Virtanen M, Westerlund H, Wang HX. Do good psychosocial working conditions prolong working lives? Findings from a prospective study in Sweden. Eur J Ageing. 2022;19(3):677–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Smith PM, Frank JW, Mustard CA, Bondy SJ. Examining the relationships between job control and health status: a path analysis approach. J Epidemiol Community Health. 2008;62(1):54–61. [DOI] [PubMed] [Google Scholar]
- 38.Too LS, Leach L, Butterworth P. Is the association between poor job control and common mental disorder explained by general perceptions of control? Findings from an Australian longitudinal cohort. Scand J Work Environ Health. 2020;46(3):311–20. [DOI] [PubMed] [Google Scholar]
- 39.Stengard J, Leineweber C, Wang HX. Prolonging working life among blue-collar workers: The buffering effect of psychosocial job resources on the association between physically demanding and hazardous work and retirement timing. SSM Popul Health. 2023;22: 101372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Lahelma E, Laaksonen M, Lallukka T, Martikainen P, Pietilainen O, Saastamoinen P, Gould R, Rahkonen O. Working conditions as risk factors for disability retirement: a longitudinal register linkage study. BMC Public Health. 2012;12:309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Dragano N, Wahrendorf M, Muller K, Lunau T. Work and health inequalities: The unequal distribution of exposures at work in Germany and Europe. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz. 2016;59(2):217–27. [DOI] [PubMed] [Google Scholar]
- 42.Borg V, Kristensen TS. Social class and self-rated health: can the gradient be explained by differences in life style or work environment? Soc Sci Med. 2000;51(7):1019–30. [DOI] [PubMed] [Google Scholar]
- 43.Schram JL, Oude Groeniger J, Schuring M, Proper KI, van Oostrom SH, Robroek SJ, Burdorf A. Working conditions and health behavior as causes of educational inequalities in self-rated health: an inverse odds weighting approach. Scand J Work Environ Health. 2021;47(2):127–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Robroek SJ, Rongen A, Arts CH, Otten FW, Burdorf A, Schuring M. Educational Inequalities in Exit from Paid Employment among Dutch Workers: The Influence of Health, Lifestyle and Work. PLoS ONE. 2015;10(8): e0134867. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Rohrbacher M, Hasselhorn HM. Social inequalities in early exit from employment in Germany: a causal mediation analysis on the role of work, health, and work ability. Scand J Work Environ Health. 2022;48(7):569–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Robroek SJ, Nieboer D, Jarvholm B, Burdorf A. Educational differences in duration of working life and loss of paid employment: working life expectancy in The Netherlands. Scand J Work Environ Health. 2020;46(1):77–84. [DOI] [PubMed] [Google Scholar]
- 47.Oude Hengel K, Robroek SJW, Eekhout I, van der Beek AJ, Burdorf A. Educational inequalities in the impact of chronic diseases on exit from paid employment among older workers: a 7-year prospective study in the Netherlands. Occup Environ Med. 2019;76(10):718–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.de Breij S, Huisman M, Deeg DJH. Work characteristics and health in older workers: Educational inequalities. PLoS ONE. 2020;15(10): e0241051. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the National Health Research Institutes of Taiwan. The availability of these data is subject to restrictions, and their use in this study was permitted under license.