Association Between Paid Work and Health-Related Quality of Life Among Community-Dwelling Older Adults: The Sukagawa Study

Abstract

In this study, we aimed to determine whether paid work has an impact on health-related quality of life (HRQOL) among older adults. Over three years, we longitudinally collected data from 5,260 community-dwelling older adults aged 75 years or older from a city in Japan. We assessed HRQOL using the Short-Form-8. We estimated the mean difference between the physical component summary (PCS) and the mental component summary (MCS) scores, which were stratified based on gender using multivariate, generalized estimating equation models. We further conducted a subgroup analysis based on the participants’ occupational backgrounds. Engagement in paid work was associated with increased MCS scores across both genders and with increased PCS scores among women. In the subgroup analysis, only women who had previously worked as managerial workers showed an inverse association with MCS scores. In this population, engagement in paid work may be a crucial factor associated with well-being.

Keywords

employment health-related quality of life gender socioeconomic status well-being

What this paper adds

This study investigated three-year longitudinal data of health-related quality of life among paid workers aged 75 years or older.

This study added evidence of the association between engagement in paid work and health-related quality of life among older adults.

This study explores the effect modification of occupational background on the association between engagement in paid work and health-related quality of life among older adults.

Applications of study findings

Paid work could be associated with better mental-health-related quality of life among older adults.

Occupational background could have effect modification on the association between paid work and quality of life among older adults.

Paid work could be a crucial factor affecting well-being among older adults.

Introduction

An aging society demands both the improvement of employability and the extension of working life among older adults (OECD, 2006). Currently, aging is an urgent global social issue. Globally, by the year 2050, one in six people will be older than 65 years (United Nations, 2019), and an increasingly larger proportion of the population will be prone to becoming dependent on long-term care or assistance. Moreover, a decline in the birthrate will exacerbate the concerns associated with an aging population (Tomáš, 2008). Therefore, many societies must tackle the twin challenge: increased demand for support and decreased supply of support. To address this double challenge, the inclusion of older adults in the work force after retirement is required across various societies (OECD, 2006). In Japan, since 2013, one in four people are older than 65 years (Ministry of Internal Affairs and Communications, 2014). Japan is among the countries leading in rapidly aging societies, extremely low birthrates, and high social demand for an older workforce. In 1986, the Japanese government implemented the Act of Stabilization of Employment among older adults and set the retirement age to 60 years to protect workers from long-term labor (Labor and Welfare Ministry of Health, 1989). However, in 2012, the aforementioned Act was amended to extend the retirement age to 65 years, and in 2021, it was further amended to extend the retirement age to 70 years (Otsuki & Ando, 2021). According to the national labor force survey conducted in 2020, one in ten older adults aged over 75 years are considered part of the labor force. Specifically, 18.2% of older adults aged between 75 and 79 years are included in the labor force.

Although the number of older adults engaging in various types of work has increased, evidence regarding the impact of engagement in paid work on health-related quality of life (HRQOL), especially among older adults aged 75 years or older, remains scarce. Although previous studies have demonstrated that retirement might be beneficial for mental health among older workers (Tuula et al., 2011), other studies have argued that engagement in paid work could be beneficial for older adults. For instance, some researchers reported that participation in productive activities, such as paid work, could prevent the occurrence of geriatric depression. This was based on the results of studies conducted in 14 European countries (Choi et al., 2013) as well as Japan (Sugihara et al., 2008). Moreover, among older adults, paid work could be a potential factor for preventing the decline of cognitive functions (Ye et al., 2019), improving physical functions, such as basic daily activities (Yoshinori et al., 2016), and it might reduce mortality ultimately (Hc, 2007). Additionally, in the context of well-being, many researchers argued in favor of productive activities, which include both volunteering and engagement in paid work, as well as expedited mental, physical, and social activities among older adults (Schwingel et al., 2009).

However, unlike volunteering, paid work often involves various obligations and responsibilities, which may be stressors among workers. Therefore, whether engaging in paid work is beneficial for well-being among older adults remains unclear. In this study, we investigated the association between engagement in paid work and HRQOL among older adults aged 75 years or older.

Moreover, in this study, in an exploratory manner, we assessed the hypothesis that occupational background had effect modifications on the association between engagement in paid work and HRQOL. Currently, owing to physical declines resulting from aging and insufficient social arrangements among older adults engaging in paid work, the varieties of paid work in which older adults can engage remain limited. However, their occupational backgrounds are diverse and can affect their health in the future, regardless of their current engagements (James, 2017). Some studies reported that, among older adults, a history of engaging in physical labor involving high levels of physical stress was independently associated with decreased physical functions in their later life (Russo et al., 2006), and lifetime occupational cognitive requirements were associated with improved cognition in the later lives of older adults (Lindsay R et al., 2016). Other researchers argued that occupational history was associated with social exclusion in the later life of older adults (Xu & Feng, 2022). Additionally, social status, which was substantially determined through occupational background, was independently associated with future health among such individuals (Bassuk et al., 2002). Therefore, in this study, depending on the difference between physical labor, for example, agriculture, forestry, and fisheries, and administrative or managerial work, we assessed the effect modifications on the association between engagement in paid work and HRQOL among older adults.

Methods

Study Design and Setting

This was a retrospective-cohort study relying on annual individual data obtained from the Sukagawa Study, which was conducted between 2018 and 2020. The Sukagawa Study was a prospective population-based project that aimed to achieve healthy aging for community-dwelling older adults aged 75 years and older in Sukagawa City, Fukushima, Japan. Details of the Sukagawa Study have been described elsewhere (Toru et al., 2021). In 2018, the city had a population of 76,832 people, of which 21,176 (27.6%) were 65 years and older and 10,405 (13.5%) were 75 years and older.

Participants (Inclusion Criteria /Exclusion Criteria)

We distributed self-report questionnaires to the eligible citizens of Sukagawa city. At the baseline, the inclusion criteria involved the following aspects: independent older adults aged 75 years and older (this was determined based on older adults requiring lower than level 2 long-term care, as it pertains to the levels of long-term care insurance [LTCI]), older adults fitting the criteria who were not hospitalized, and older adults living in Sukagawa city. Long-term care insurance is a mandatory social insurance system run by the local municipalities to support long-term care for older adults in Japan (Labor and Welfare, Ministry of Health, 2016). The levels are categorized into seven ranks based on the levels of needs for daily living through a comprehensive evaluation conducted by an authorized care manager and the primary physician in charge (Labor and Welfare, Ministry of Health, 2016). Long-term care level 2 indicates that a person requires a limited amount of assistance in their basic activities of daily living, such as bathing and managing daily medicine or finances.

We included participants who responded to the 2018 questionnaires and excluded non-respondents from both the 2019 and 2020 questionnaires as well as those who did not complete the exposure variable in 2018 and the outcome variable in 2019 and 2020. We conducted the study in accordance with the Declaration of Helsinki and its future amendments. We obtained written informed consent for using the collected data for research purposes from all the respondents at the every-year surveys. Therefore, in this secondary-data-use study, we were exempt from obtaining any additional consent.

Measurements

The main exposure was engagement in paid work throughout 2018. In the baseline survey for 2018, we defined participants as paid workers if they answered “Yes” to the following question: “Do you currently engage in paid work?” In 2019 and/or 2020, the primary outcomes involved the physical and mental scores (PCS and MCS) contained in the Short-Form-8 (SF-8) (Fukuhara & Suzukamo, 2005), which were included in the self-administered questionnaires. These scores were standardized using the national reference of the 2017 Japanese version of the SF-8.

The covariates included were age, annual income, living alone (Hays, 2002), marital status (Robards et al., 2012), educational status (Kempen et al., 1999), instrumental activities of daily living (IADLs) (Tomioka et al., 2017), comorbidity (Nelis et al., 2019), financial distress (Huang et al., 2020), and scores of the SF-8 (MCS and PCS) at the baseline in 2018. Annual income was measured using an eight-category optional question that was later re-categorized into four larger categories. Living alone was based on the following question: “How many family members currently live together?” Marital status was defined through the experience of legal marriage, regardless of the current statuses, such as divorce or separation. Educational status, which is a binominal covariate, was defined as higher than high school, which generally took more than 12 years. IADLs were measured using the following five questions: “Can you go out using public transportation, such as a bus or train?,” “Can you go shopping to buy your daily necessities?,” “Can you prepare your daily meals?,” “Can you deal with an invoice or bill?,” and “Can you deposit money into a bank or postal savings and withdraw money from those?” The answers comprised three categories, which are as follows: independently able, able with support, and unable. The questions were based on the competence of instrumental self-maintenance as contained in the Tokyo Metropolitan Institute of Gerontology Index of Competence (Koyano et al., 1991). We defined the full IADL compliant participants as responders who could independently answer all the five questions (Fillenbaum, 1985). Comorbidity was measured using the following aspects: history of cancer diagnosis, cerebral ischemia or hemorrhage, myocardial infarction or angina pectoris, hypertension, diabetes mellitus, and depression. Multi-morbidity was defined as having more than two comorbidities (Johnston et al., 2019). Financial distress was defined based on the response to the following question: “Are you frustrated about your financial problems?” This question was based on the Comprehensive Survey of Living Conditions, which is a major national survey conducted by the Ministry of Health, Labor, and Welfare. As continuous variables, we used the baseline MCS and PCS scores for each model in 2018. Occupational background was measured using a standard international socio-economic index of occupational status (Ganzeboom et al., 1992). In the sub-group analysis, we investigated the potential differences resulting from engaging in physical labor or a high social status. We considered engagement in physical labor or a high social status, after which we categorized the previous occupational backgrounds of the participants into the following four groups: managerial jobs in the agriculture, forestry, and fisheries (AFF) industries and non-AFF jobs, including technocrats, engineering jobs, clerical work, services, and sales, and never employed as a regular worker (ICSE, 1993).

Statistical Analysis

To describe the baseline characteristics, participants were divided into two groups: paid workers and unemployed. The baseline characteristics were presented using means with standard deviations (SDs) and medians with interquartile range (IQR) for continuous variables as well as percentages for categorical variables. We analyzed all the estimations stratified by gender considering the historical gender difference in social roles among the participants’ generations. Because the data involved longitudinal repeated measures of outcome, we used multivariate generalized estimating equation models with an identity link function, assumed residuals in a Poisson distribution, exchanged the working correlation matrix, and used robust standard estimation (Liang & Zeger Scott, 1986). We used this model to estimate the mean differences and 95% confidence intervals (CIs) for the PCS and MCS scores, 2 years after the baseline survey was associated with the difference in paid workers and those who were unemployed. We adjusted for the baseline MCS and PCS scores as well as the aforementioned covariates throughout all the analyses. We conducted an exploratory subgroup analysis based on occupational backgrounds to identify the differences in the associations between previous occupations. Furthermore, we synthetized the overall associations and considered the heterogeneity of each occupational background using a random effect model. We used I² statistics to describe heterogeneity among each subgroup and considered I²< 40% not important (Higgins & Green, 2011). I² statistics refers to the percentage of the variability in the association of each subgroup owing to heterogeneity, rather than a sampling error. A high I² statistics score represents an increased level of heterogeneity. For the sensitivity analysis, we performed the same main analysis after excluding participants whose status of paid work had changed during the two years. In the preliminary analysis, we noticed that the analyzed dataset contained 2.3% of the missing data overall, which included, at most, 13.3% missing data in the variables of income. Therefore, we implemented 20 multiple imputations using chained equations for missing data of covariates, which were conducted under the assumption that the analyzed data were missing at random. We derived the association estimates and 95% CIs from the multiplied imputed data, and the mean value for each of the 20 estimates was achieved by combining the results from the datasets using Rubin’s rules (Rubin, 1987). P-values < .05 were considered statistically significant. All the data were analyzed using STATA, version 16.1 (Stata Corp., College Station, TX, USA).

Results

Baseline Characteristics

Out of the 10,267 citizens aged 75 years or older residing in Sukagawa city in March 2018, we delivered the baseline questionnaire to the 8,869 eligible participants. Of all the eligible participants, 5,620 responded in 2018, with a response rate of 63.4%. After excluding the participants who did not respond in 2019 and 2020 as well as the missing variables in exposure or outcome, 4,675 responses, with a follow-up rate of 83.2%, were analyzed, as shown in Figure 1. Among those, 3,453 (73.9%) participants responded in both 2019 and 2020, whereas 826 (17.7%) and 396 (8.5%) participants responded only in 2019 and 2020, respectively.

Figure 1.

Study flow diagram.

The baseline characteristics of the analyzed 4,675 participants, stratified by gender, are listed in Table 1. There were 2,654 women (56.8%) and 2,021 men (43.2%). The means (SD), and medians (IQR) of age, for the unemployed vs. paid workers among women and men were 81 (5), 80 (IQR 77–84) vs. 78 (4), 77 (IQR 75–80) among women and 80 (5), 80 (IQR 77–83) vs. 78 (4), 77 (IQR 75–80) among men. For both genders, paid workers tended to be younger and scored high HRQOL scores in both the PCS and MCS at the baseline. The dropout rate throughout the two years involving each unemployed and paid worker was 6.3% (461), compared to 7.0% (21) among women and 14.0% (250) compared to 6.1% (32) among men.

Table 1.

Characteristics of Participants by Gender.

		Women			Men
		Unemployed	Paid worker	Total	Unemployed	Paid worker	Total
		N=2,374	N=280	N=2,654	N=1,530	N=491	N=2,021
Age		81 (5)	78 (4)	80 (5)	80 (5)	78 (4)	80 (5)
Annual income in 2018	∼ ¥2 million	1,209 (50.9%)	136(48.6%)	1,345(50.7%)	585 (38.2%)	153(31.2%)	738 (36.5%)
	¥2∼4 million	331 (13.9%)	39 (13.9%)	370 (13.9%)	424 (27.7%)	104 (21.2%)	528 (26.1%)
	¥4∼7.5 million	276 (11.6%)	37 (13.2%)	313 (11.8%)	300 (19.6%)	100 (20.4%)	400 (19.8%)
	¥7.5 million∼	558 (23.5%)	68 (24.3%)	626 (23.6%)	221 (14.4%)	134 (27.3%)	355 (17.6%)
Living alone		449 (18.9%)	63 (22.5%)	512 (19.3%)	146 (9.5%)	35 (7.1%)	181 (9.0%)
Married experience		2,283 (98.6%)	265 (97.4%)	2,548 (98.5%)	1,497(98.6%)	479 (99.4%)	1,976(98.8%)
Education (higher than high school)		985 (41.5%)	133 (47.5%)	1,118 (42.1%)	785 (51.3%)	250 (50.9%)	1,035 (51.2%)
Full IADL		1,692 (71.3%)	246 (87.9%)	1,938 (73.0%)	1,174 (76.7%)	413 (84.1%)	1,587 (78.5%)
Full BADL		2,182 (91.9%)	267 (95.4%)	2,449 (92.3%)	1,433 (93.7%)	469 (95.5%)	1,902 (94.1%)
Multimorbidity (comorbidity >=2)		429 (18.1%)	44 (15.7%)	473 (17.8%)	395 (25.8%)	103 (21.0%)	498 (24.6%)
Financial distress		281 (11.8%)	30 (10.7%)	311 (11.7%)	227 (14.8%)	52 (10.6%)	279 (13.8%)
PCS score in 2018		46.7 (7.3)	50.5 (5.1)	47.1 (7.2)	48.0 (7.0)	50.3 (5.4)	48.6 (6.7)
MCS score in 2018		50.7 (6.2)	52.6 (4.7)	50.9 (6.1)	51.3 (5.6)	52.3 (4.6)	51.6 (5.4)
Occupational backgrounds	AFF	351 (14.8%)	60 (21.4%)	411 (15.5%)	216 (14.1%)	120 (24.4%)	336 (16.6%)
	Non-AFF	1,763(74.3%)	198 (70.7%)	1,961 (73.9%)	1,113 (72.7%)	314 (64.0%)	1,427 (70.6%)
	Managerial worker	25 (1.1%)	8 (2.9%)	33 (1.2%)	191 (12.5%)	52 (10.6%)	243 (12.0%)
	No experience as a regular employee	235 (9.9%)	14 (5.0%)	249 (9.4%)	10 (0.7%)	5 (1.0%)	15 (0.7%)

PCS: Physical Component Scale, MCS: Mental Component Scale.

IADL: instrumental activities of daily living, BADL: basic activities of daily living.

AFF: agriculture, forestry, and fisheries, non-AFF: includes technocrats, engineering jobs, clerical work, service, sales, and others.

Primary outcome

The adjusted mean differences of the PCS and MCS scores are listed in Tables 2 and 3, respectively. The results revealed that paid workers had independently higher PCS scores among women (mean difference in PCS scores among women 0.67, 95% CI: 0.04–1.30) and independently higher MCS scores in both genders (mean difference in MCS scores among women 0.90, 95% CI: 0.30–1.49 and men MCS 0.48, 95% CI: 0.02–0.95).

Table 2.

Mean Differences of Engagement in Paid Work for PCS Scores According to the SF-8 Estimated Through GEE Among Women/Men (Aged > = 75 Years) Throughout the Two-Year Follow-Up Period.

		Women			Men
		Mean differences	95% CIs		Mean differences	95% CIs
Paid work		0.67*	0.04	1.30	0.07	–0.46	0.60
Age (per year)		–0.17*	–0.23	–0.12	–0.17	–0.23	–0.10
Annual income	∼¥2 million	ref	—	—	ref	—	—
	¥2∼4 million	0.04	–0.55	0.64	0.40	–0.19	0.99
	¥4∼7.5 million	0.17	–0.49	0.83	0.61	–0.03	1.25
	¥7.5 million∼	0.10	-0.47	0.66	1.01*	0.24	1.78
Living alone		0.49	–0.06	1.04	0.36	–0.52	1.24
Married		0.39	–1.19	1.96	–0.37	–2.97	2.23
Education (higher than high school)		0.29	–0.27	0.86	0.10	–0.60	0.80
Full IADL		2.28*	1.65	2.91	2.18*	1.38	2.98
Multimorbidity (comorbidity >=2)		–0.68*	–1.27	–0.09	–1.11*	–1.68	–0.55
Financial distress		–0.82*	–1.57	–0.06	–0.97*	–1.77	–0.17
PCS in 2018 (per Unit)		0.60*	0.56	0.65	0.58*	0.51	0.66

PCS: Physical Component Scale, MCS: Mental Component Scale, IADL: instrumental activities of daily living, SF-8: Short-Form-8, GEE: generalized estimating equation, CI: confidence interval, *: p-value < 0.05.

Adjusted for age, annual income, living alone, married, education, full IADL, multimorbidity, financial distress, and PCS in 2018.

Table 3.

Mean Differences of Engagement in Paid Work for MCS Scores According to the SF-8 Estimated Through GEE Among Women/Men (Aged > = 75 Years) Throughout the Two-Year Follow-Up Period.

		Women			Men
		Mean differences	95% CIs		Mean differences	95% CIs
Paid work		0.90*	0.30	1.49	0.48*	0.02	0.95
Age (per year)		0.01	–0.04	0.06	0.01	-0.04	0.05
Annual income	∼ ¥2 million	ref	—	—	ref	—	—
	¥2∼4 million	–0.39	–0.99	0.22	0.22	–0.33	0.76
	¥4∼7.5 million	–0.17	–0.81	0.47	0.19	–0.39	0.77
	¥7.5 million∼	–0.21	–0.73	0.31	0.37	–0.23	0.97
Living alone		0.37	–0.17	0.92	–0.78	–1.62	0.05
Married		1.08	–1.18	3.33	0.85	–1.39	3.10
Education (higher than high school)		0.20	–0.34	0.74	0.17	–0.40	0.73
Full IADL		0.94*	0.36	1.51	1.52*	0.90	2.15
Multimorbidity (comorbidity >=2)		–0.69*	–1.28	–0.09	0.50	–0.46	0.56
Financial distress		–1.30*	–2.04	–0.57	–1.12*	–1.83	–0.41
PCS in 2018 (per unit)		0.51*	0.46	0.55	0.54*	0.49	0.59

Adjusted for age, annual income, living alone, married, education, full IADL, multimorbidity, financial distress, and MCS in 2018.

Subgroup Analysis

The adjusted mean differences of the PCS and MCS scores based on occupational background for both genders are shown in Figure 2. The results revealed that non-AFF jobs and no experience as a regular employee among women were associated with independently higher MCS scores (mean difference in MCS scores among non-AFF women was 0.80, 95% CI: 0.11–1.50; mean difference in MCS scores for no experience as a regular employee was 3.40, 95% CI: 0.66–6.14 among women). Non-AFF jobs among men were associated with independently higher MCS scores (mean difference in MCS scores for non-AFF jobs: 0.62, 95% CI: 0.06–1.19 among men). Conversely, although the sample size was limited, women who were managerial workers had independently lower MCS scores (mean difference in MCS scores among women who were managerial workers: −2.82, 95% CI: −5.08–0.55). Overall, across both genders, the PCS and MCS scores of paid workers were higher than those of the unemployed individuals (mean difference 0.38, 95% CI: −0.13–0.90 and mean difference 0.58, 95% CI: 0.22–0.95), with insignificant heterogeneity (I² = 23.45% and I² = 0.00%), respectively.

Figure 2.

Subgroup analysis based on occupational backgrounds.

Sensitivity Analysis

Between 2018 and 2019, 65 (23.2%) and 107 (21.8%) paid workers quit their jobs, and 67 (2.8%) and 67 (4.4%) unemployed women and men, respectively, started new jobs. During the period between 2019 and 2020, 89 (31.8%) and 134 (27.3%) paid workers quit their jobs, and 41 (1.7%) and 34 (2.2%) unemployed women and men, respectively, started new jobs. After excluding all the participants who changed engagement in paid work, the adjusted mean difference in the PCS and MCS scores among women was 1.08 (95% CI: 0.58–1.58) and 0.77 (95% CI: 0.29–1.25), as shown in Supplementary Figures 1 and 2, respectively. The adjusted mean difference in the PCS and MCS scores among men was 0.58 (95% CI: 0.13–1.03) and 0.44 (95% CI: 0.08–0.80), as shown in Supplementary Figures 1 and 2, respectively.

Discussion

Summary of Findings

Current engagement in paid work was associated with high levels of HRQOL among community-dwelling older adults aged 75 years or older during the two-year follow up. Engagement in paid work had an independent association with high score in the physical component among women and with high scores in the mental component for both genders. The subgroup analysis suggests effect modifications based on occupational backgrounds. Unlike other groups, especially among women, paid workers with a history of managerial work had a negative association with the MCS scores of HRQOL.

Overall Positive Influence of Paid Work Among Older Adults

The overall findings of this study showing that engagement in paid work could be beneficial for older adults, both mentally and physically, are concordant with those of many previous studies. This study successfully added new evidence regarding HRQOL among older workers to existing knowledge. Previous studies revealed that productive activities, such as paid work, could be beneficial in ensuring mental health among older adults (Choi et al., 2013; Sugihara et al., 2008). Engagement in paid work might also be associated with improved physical status (Yoshinori et al., 2016) and enhanced cognitive functions (Ye et al., 2019) among older adults. Although most previous studies report substantial variances in the effects of outcomes pertaining gender, urban and rural residence, and socioeconomic status, the positive impacts of paid work on mental and physical functions could potentially explain the results of this study, whereby paid workers showed elevated levels of HRQOL. A study on the population suffering from rheumatoid arthritis also revealed that engagement in paid work is positively associated with improved HRQOL (Grønning et al., 2010). Although the study was based on a cross-sectional design, researchers argued that the positive association of paid work with increased levels of HRQOL remained after adjusting for disease severity among the population suffering from this chronic and physically disabling condition. The findings of their study are in line with our results.

Difference in Gender and Occupational Backgrounds

Similar to previous studies, the results of this study also indicated differences between genders (Damman & Henkens, 2020; Orfila et al., 2006). In our observations, women’s HRQOL is slightly but consistently lower than men’s HRQOL. These results are consistent with those of previous studies regarding HRQOL among older adults (Orfila et al., 2006). Regarding the mental component, both genders showed an independent and positive association between paid work and MCS scores. Conversely, regarding the physical component, men did not show a significant association between paid work and PCS scores. In the subgroup analysis of the PCS, only men from the non-AFF background showed lower PCS scores, and although the results were insignificant, they were slightly different from the trends of other occupational backgrounds. Such tendencies may be partially explained by the fact that non-AFF background workers tend to experience a job change after mandatory retirement, unlike the other groups. However, according to previous reports, lower socioeconomic status was related to decreased leisure-time physical activity and increased occupational physical activity (Beenackers et al., 2012). Therefore, men with a non-AFF background could have potentially lower leisure-time physical activities than managerial workers, who generally earn high levels of income and achieve high educational status. Additionally, because older adults with a non-AFF background engage in non-physical labor unlike those with an AFF background, they could experience relatively increased levels of physical activity. Therefore, the potential low levels of physical activity before retirement might explain the opposite influence of paid work on PCS scores among older adults with a non-AFF background. Previous studies also reveal that AFF experiences are positively associated with late-life independence (Haruyama et al., 2020) and longevity (Thelin et al., 2009). Thus, under specific types of past occupational backgrounds, there should be a non-negligible possibility for an effect modification of paid work on physical components among older adults (Beenackers et al., 2012). However, further detailed investigations on occupational backgrounds are required to verify the speculations on such impacts.

Furthermore, our results revealed an effect modification of paid work on MCS scores among older adults with managerial work backgrounds, especially among women. Although the limited statistical power resulting from the sample size reduced robustness in this study, women who were formerly engaged in managerial work and who are currently engaged in paid work showed significantly lower MCS scores. The results of the MCS scores among men formerly engaged in managerial work similarly showed a lower influence of paid work on the MCS scores. In comparison, women who had no experience as regular employees (regular employees can only be terminated for specific reasons) tended to show significantly higher MCS and PCS scores than women with other occupational backgrounds. These heterogeneities across occupational backgrounds can be explained by the change of social status in later life among such individuals. A previous study reported that subjective social status was independently associated with HRQOL after adjusting for objective social status (Choi et al., 2015). Therefore, the unintentional decline of social status in working experience throughout later life might reverse the benefit of engagement in paid work among this population. Meanwhile, although the subgroup analysis showed heterogeneity depending on the occupational backgrounds, the results only referred to the variances associated with this classification. Therefore, interpretations of the exploratory analysis should be approached with caution.

Healthy Worker Survivor Effect

Although we carefully adjusted for maximum potential confounders, we might be unable to completely eliminate all the residual confounder influences arising from what we term as the “healthy worker survivor effect,” which is a bias resulting from the continuous selection process, whereby those who remain employed tend to be healthier than those who leave employment (Arrighi & Hertz-Picciotto, 1994). However, unless they had lost independence in their daily lives, our survey followed up on paid workers after they quit their jobs. Therefore, selection bias induced by excess lost to follow-up rate among unhealthy workers was expected to be minimized. The results of the sensitivity analysis and follow-up rates for each group partially explain and support the arguments presented in this study. This is because paid workers who quit their jobs are likely to have decreased HRQOL levels, and excluding such workers would probably induce an overestimation of the associations in the group of paid workers, as shown in the results of the sensitivity analysis.

Clinical Implications of the Impact of the Mean Difference

Although the minimum clinically important difference (MCID) is different depending on the target population and the types of interventions or exposures, in the population involved in this study, the MCID of MCS and PCS throughout the two years is conventionally regarded as 2–5 points (Dorcas et al., 2002). From the short-term perspective, the mean differences between the results of this study and those of other studies would be too small according to the common sense of the MCID level. However, apart from the results of this study, there is no written evidence of whether engagement in paid work is longitudinally associated with HRQOL in this population. Therefore, considering the findings of this study, we might deduce that at least 4 to 6 years of longitudinal observation is required to identify the meaningful effect of paid work in future studies.

Limitations

This study has several limitations. First, this study did not measure the status of unpaid work. Some of the unemployed individuals might engage in unpaid work, such as volunteering and housekeeping, which is expected to enhance the health status of older adults (Hayley & Liana, 2018). In this study, participants who were involved only in volunteering and not in paid work were incorporated into the unemployed group. Given the similarities of the background mechanisms associated with the impact of volunteering on the health of older adults, the classifications presented in this study might underestimate the impact of paid work. Therefore, we considered that these potential misclassifications would not alter the overall trends of the results. A previous study also supports our assumption (Schwingel et al., 2009). Second, this study faced a lack of information regarding the types, intensity, and length of current paid work. Considering the wide variation of older adults aged 75 years and older (Bangsbo et al., 2019), further studies on the ways in which to measure work among older populations should be conducted. Additionally, other factors, such as leisure activities (Cho et al., 2021), migration in later life (Xu & Feng, 2022), and status of social support (Pindek & Segel-Karpas, 2021), that potentially affect well-being among older workers should be considered in future studies. Third, owing to the design of this study, only relatively independent participants were included, which may have resulted in biased results. The inclusion criteria potentially tended to exclude participants with poor HRQOL status, thereby resulting in selection bias, which could overestimate the impact of paid work on HRQOL, especially in situations where paid workers with poor HRQOL levels are more likely to drop out, compared to unemployed individuals with poor HRQOL levels. However, the dropout rate of unemployed individuals was almost similar among women and slightly higher among men. Therefore, the possibility of the overestimate being induced by this type of selection bias would be low. Finally, in this observational study, we only investigated the association, not causality, by comparing the following two groups: older adults who had engaged in paid work and those who had not throughout the duration of the baseline survey. Furthermore, the data collected in this study did not include the length and types of engagement in paid work. Therefore, the effect of engagement in paid work on unemployment as an intervention was never inferred through the results of this study. From an ethical perspective, considering the difficulties of direct intervention through engagement in paid work, a longer observation period involving a new user design or time-varying exposure analysis would be preferable to overcome the aforementioned limitations under the restrictions of an observational study.

Conclusion

Current engagement in paid work was independently associated with increased mental and physical components of HRQOL among women and with a higher mental component among men for individuals aged 75 years or older. This study also suggested the potential effect modification of occupational backgrounds on the association between paid work and HRQOL. Engagement in paid work in this population may be a crucial factor affecting well-being among individuals within this community. Policymakers should consider the backgrounds and experiences of older adults to ensure that such individuals have access to appropriate employment opportunities.

Supplemental Material

Supplemental MaterialAssociation Between Paid Work and Health-Related Quality of Life Among Community-Dwelling Older Adults: The Sukagawa Study

Supplemental Material for Association Between Paid Work and Health-Related Quality of Life Among Community-Dwelling Older Adults: The Sukagawa Study by Atsushi Takayama, Taro Takeshima, Kenji Omae, Takashi Yoshioka, Hiroaki Nakagawa, Akihiro Ozaka, Sei Takahashi, Toru Naganuma, Sugihiro Hamaguchi, Shunichi Fukuhara, and The Sukagawa Study Group in Journal of Applied Gerontology.

Footnotes

Acknowledgments

This study has been conducted as part of the project “Sukagawa Healthy longevity project.” The authors deeply thank the participants of the Sukagawa Study, the Sukagawa city local government officers in the Department of Healthcare and Welfare, and Mr. Naoaki Yamaga for data administration and technical support.

Author Contributions

Atsushi Takayama: conceptualization, writing—original draft, methodology, and formal analysis.

Taro Takeshima: conceptualization and writing—review and editing.

Kenji Omae: conceptualization, writing—review and editing, and methodology; Takashi Yoshioka: data curation and writing—review and editing.

Hiroaki Nakagawa: data curation and writing—review and editing.

Akihiro Ozaka: data curation and writing—review and editing.

Sei Takahashi: data curation and writing—review and editing.

Toru Naganuma: data curation and writing—review and editing.

Sugihiro Hamaguchi: supervision and project administration.

Shunichi Fukuhara: supervision and project administration.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical Approval

The Institutional Ethics Review Board of Fukushima Medical University School of Medicine approved this study and its protocols (registered approval number: 2975).

ORCID iD

Atsushi Takayama

Supplemental Material

Supplemental material for this article is available online.

References

Arrighi

H. M.

Hertz-Picciotto

(1994). The evolving concept of the healthy worker survivor effect. Epidemiology (Cambridge, Mass.), 5(2), 189–196, https://doi.org/10.1097/00001648-199403000-00009

Bangsbo

Blackwell

Boraxbekk

C. J.

Caserotti

Dela

Evans

A. B.

Jespersen

A. P.

Gliemann

Kramer

A. F.

Lundbye-Jensen

Mortensen

E. L.

Lassen

A. J.

Gow

A. J.

Harridge

S. D. R.

Hellsten

Kjaer

Kujala

U. M.

Rhodes

R. E.

Pike

E. C. J.

Vina

(2019). Copenhagen consensus statement 2019: Physical activity and ageing. British Journal of Sports Medicine, 53(14), 856–858, https://doi.org/10.1136/bjsports-2018-100451

Bassuk

S. S.

Berkman

L. F.

Amick

B. C.

3rd . (2002). Socioeconomic status and mortality among the elderly: findings from four US communities. American Journal of Epidemiology, 155(6), 520–533, https://doi.org/10.1093/aje/155.6.520

Beenackers

A. M.

Kamphuis

B. M. C.

Giskes

Brug

Kunst

E. A.

Burdorf

van Lenthe

F. J.

(2012). Socioeconomic inequalities in occupational, leisure-time, and transport related physical activity among European adults: A systematic review. The International Journal of Behavioral Nutrition and Physical Activity, 9(1), 116, https://doi.org/10.1186/1479-5868-9-116

Cho

Chen

T. Y.

Janke

M. C.

(2021). A 2-year longitudinal relationship between work-family conflict and health among older workers: Can gardening help? Journal of Applied Gerontology, 40(10), 1330–1341, https://doi.org/10.1177/0733464820934678

Choi

K. S. C.

Robert

Michael

(2013). Participation in productive activities and depression among older Europeans: Survey of health, ageing and retirement in Europe (SHARE). International journal of geriatric psychiatry, 28(11), 1157, https://doi.org/10.1002/gps.3936

Choi

Kim

J. H.

Park

E. C.

(2015). The effect of subjective and objective social class on health-related quality of life: New paradigm using longitudinal analysis. Health and Quality of Life Outcomes, 13(1), 121, https://doi.org/10.1186/s12955-015-0319-0

Damman

Henkens

(2020). Gender differences in perceived workplace flexibility among older workers in the Netherlands: A brief report. Journal of Applied Gerontology, 39(8), 915–921, https://doi.org/10.1177/0733464818800651

Dorcas

Marteen

George

A. W.

(2002). Many faces of the minimal clinically important difference (MCID): A literature review and directions for future research. Current Opinion in Rheumatology, 14(2), 109–114, https://doi.org/10.1097/00002281-200203000-00006

10.

Fillenbaum

G. G.

(1985). Screening the elderly. A brief instrumental activities of daily living measure. Journal of the American Geriatrics Society, 33(10), 698–706, https://doi.org/10.1111/j.1532-5415.1985.tb01779.x

11.

Fukuhara

Suzukamo

(2005). Health related QOL scale—SF-8 & SF-36. Igakuno Ayumi, 213(2), 133–136.

12.

Ganzeboom

H. B. G.

Graaf

P. M. D.

Treiman

D. J.

(1992). A standard international socio-economic index of occupational status. Social Science Research, 21(1), 1–56, https://doi.org/10.1016/0049-089x(92)90017-b

13.

Grønning

Rødevand

Steinsbekk

(2010). Paid work is associated with improved health-related quality of life in patients with rheumatoid arthritis. Clinical Rheumatology, 29(11), 1317–1322, https://doi.org/10.1007/s10067-010-1534-z

14.

Haruyama

Yokomichi

Yamagata

(2020). Farm working experience could reduce late-life dependency duration among Japanese older adults: The Yamanashi healthy-active life expectancy cohort study based on the STROBE guidelines. Medicine (Baltimore), 99(38), Article e22248, https://doi.org/10.1097/md.0000000000022248

15.

Hayley

Liana

(2018). Volunteering in the community: Potential benefits for cognitive aging. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 73(3), 399–408, https://doi.org/10.1093/geronb/gbx134

16.

Hays

J. C.

(2002). Living arrangements and health status in later life: A review of recent literature. Public Health Nursing (Boston, Mass.), 19(2), 136–151, https://doi.org/10.1046/j.1525-1446.2002.00209.x

17.

(2007). Does social participation by the elderly reduce mortality and cognitive impairment? Aging and Mental Health, 11(6), 699-707, https://doi.org/10.1080/13607860701366335

18.

Higgins

P. J.

Green

(2011). Cochrane handbook for systematic reviews of interventions. In Higgins

P. J.

Green

(Eds.), Identifying and measuring heterogeneity (Version 5.1.0 ed., pp. 9.5.2): The Cochrane Collaboration.

19.

Huang

Ghose

Tang

(2020). Effect of financial stress on self-rereported health and quality of life among older adults in five developing countries: A cross sectional analysis of WHO-SAGE survey. BMC Geriatrics, 20(1).

20.

ICSE . (1993). Resolution concerning the International Classification of Status in Employment (ICSE), adopted by the Fifteenth International Conference of Labour Statisticians https://www.ilo.org/wcmsp5/groups/public/–-dgreports/–-stat/documents/normativeinstrument/wcms_087562.pdf

21.

James

(2017). Class and health inequality in later life: Patterns, mechanisms and implications for policy. International Journal of Environmental Research and Public Health, 14(12), 1533, https://doi.org/10.3390/ijerph14121533

22.

Johnston

M. C.

Crilly

Black

Prescott

G. J.

Mercer

S. W.

(2019). Defining and measuring multimorbidity: A systematic review of systematic reviews. European Journal of Public Health, 29(1), 182–189, https://doi.org/10.1093/eurpub/cky098

23.

Kempen

G. I. J. M.

Brilman

E. I.

Ranchor

A. V.

Ormel

(1999). Morbidity and quality of life and the moderating effects of level of education in the elderly. Social Science and Medicine (1982), 49(1), 143–149, https://doi.org/10.1016/s0277-9536(99)00129-x

24.

Koyano

Shibata

Nakazato

Haga

Suyama

(1991). Measurement of competence: Reliability and validity of the TMIG Index of Competence. Archives of Gerontology and Geriatrics, 13(2), 103–116, https://doi.org/10.1016/0167-4943(91)90053-s

25.

Liang

K. Y.

Zeger Scott

(1986). Longitudinal data analysis using generalized linear models. Biometrika, 73(1), 13–22, https://doi.org/10.1093/biomet/73.1.13

26.

Lindsay

R. P.

Jennifer

Robert

S. W.

Ute

Denis

A. E.

Carlos

F. M. D. L.

(2016). Occupational cognitive requirements and late-life cognitive aging. Neurology, 86(15), 1386–1392, https://doi.org/10.1212/wnl.0000000000002569

27.

Labor and Welfare Ministry of Health . (1989). Change of Labor Force Population and Labor Force Participation Rate.Ministry of Health, Labour and Welfare https://www.mhlw.go.jp/english/policy/affairs/dl/01.pdf

28.

Labor and Welfare Ministry of Health . (2016). Long-term care insurance system of Japan https://www.mhlw.go.jp/english/policy/care-welfare/care-welfare-elderly/dl/ltcisj_e.pdf

29.

Ministry of Internal Affairs and Communications . (2014). Population estimation in 1st of October in 2013 https://www.e-stat.go.jp/dbview?sid=0003094401

30.

Nelis

S. M.

Y. T.

Matthews

F. E.

Martyr

Quinn

Rippon

Rusted

Thom

J. M.

Kopelman

M. D.

Hindle

J. V.

Jones

R. W.

Clare

(2019). The impact of co-morbidity on the quality of life of people with dementia: Findings from the IDEAL study. Age and Ageing, 48(3), 361–367, https://doi.org/10.1093/ageing/afy155

31.

OECD . (2006). Live longer, work longer. Paris: OECD Publishing.

32.

Orfila

Ferrer

Lamarca

Tebe

Domingo-Salvany

Alonso

(2006). Gender differences in health-related quality of life among the elderly: The role of objective functional capacity and chronic conditions. Social Science and Medicine, 63(9), 2367–2380, https://doi.org/10.1016/j.socscimed.2006.06.017

33.

Otsuki

Ando

(2021). Amendment of the Act on Stabilization of Employment of Elderly Persons.

34.

Pindek

Segel-Karpas

(2021). Do satisfied older employees perceive more social support over time? a cross-lagged examination. Journal of Applied Gerontology, 40(6), 670–678, https://doi.org/10.1177/0733464820910043

35.

Robards

Evandrou

Falkingham

Vlachantoni

(2012). Marital status, health and mortality. Maturitas, 73(4), 295–299, https://doi.org/10.1016/j.maturitas.2012.08.007

36.

Rubin

D. B.

(1987). Multiple imputation for nonresponse in surveys. John Wiley and Sons, Ltd.

37.

Russo

Onder

Cesari

Zamboni

Barillaro

Capoluongo

Pahor

Bernabei

Landi

(2006). Lifetime occupation and physical function: A prospective cohort study on persons aged 80 years and older living in a community. Occupational and Environmental Medicine, 63(7), 438–442, https://doi.org/10.1136/oem.2005.023549

38.

Schwingel

Niti

M. M.

Tang

T. P.

(2009). Continued work employment and volunteerism and mental well-being of older adults: Singapore longitudinal ageing studies. Age and Ageing, 38(5), 531–537, https://doi.org/10.1093/ageing/afp089

39.

Sugihara

Sugisawa

Shibata

Harada

(2008). Productive roles, gender, and depressive symptoms: Evidence from a national longitudinal study of late-middle-aged Japanese. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 63(4), P227–P234, https://doi.org/10.1093/geronb/63.4.p227

40.

Thelin

Holmberg

Nettelbladt

Thelin

. (2009). Mortality and morbidity among farmers, nonfarming rural men, and urban referents: A prospective population-based study. International Journal of Occupational and Environmental Health, 15(1), 21–28, https://doi.org/10.1179/oeh.2009.15.1.21

41.

Tomáš

(2008). Does persistent low fertility threaten the future of European populations? In Johan

Patrick

(Eds.), Demographic challenges for the 21st century: A state of the art in demography (pp. 27–90). Brussels: VUBPRESS Brussels University Press.

42.

Tomioka

Kurumatani

Hosoi

(2017). Age and gender differences in the association between social participation and instrumental activities of daily living among community-dwelling elderly. BMC Geriatrics, 17(1), 99, https://doi.org/10.1186/s12877-017-0491-7

43.

Toru

Sei

Taro

Noriaki

Kenji

Takashi

Ohnishi

Ito

Fukuma

Hamaguchi

Fukuhara

Hasegawa

Yajima

Niihata

Iida

Kobayashi

Sasaki

Nishiwaki

Tominaga

the Sukagawa Study Group . (2021). Cohort profile: A super-elderly population-based cohort in Japan (the Sukagawa Study). International Journal of Epidemiology, 50(3), 727-727, https://doi.org/10.1093/ije/dyaa285

44.

Tuula

Jussi

Hugo

Jaana

Noora

Marianna

Kawachi

Kivimaki

(2011). Is retirement beneficial for mental health?: Antidepressant use before and after retirement. Epidemiology, 22(4), 553-559, https://doi.org/10.1097/ede.0b013e31821c41bd

45.

United Nations . (2019). World Population Ageing 2019 Highlights https://www.un.org/en/development/desa/population/publications/pdf/ageing/WorldPopulationAgeing2019-Highlights.pdf

46.

Feng

. (2022). Social exclusion among older rural residents in China: Does either non-agricultural work or living away from their hometowns over the courses of their lives make any difference? Journal of Applied Gerontology, 41(9), 2074–2083, https://doi.org/10.1177/07334648221097690

47.

Zhenmei

. (2019). Productive activities and cognitive decline among older adults in China: Evidence from the China health and retirement longitudinal study. Social Science and Medicine (1982), 229, 96–105, https://doi.org/10.1016/j.socscimed.2018.09.052

48.

Yoshinori

Shoji

Erika

Ushio

Hiroyuki

Hideyo

Ishizaki

Kumagai

Watanabe

Furuna

Suzuki

(2016). Engagement in paid work as a protective predictor of basic activities of daily living disability in Japanese urban and rural community-dwelling elderly residents: An 8-year prospective study. Geriatrics and Gerontology International, 16(1), 126–134, https://doi.org/10.1111/ggi.12441

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.24 MB