Background Few studies have examined the health effects of employment situation among women, taking social and economic conditions into consideration. The objective of this research was to investigate the association of employment situation (full-time or part-time employee and self-employed) with mortality risk in women over a 20-year follow-up period. Additionally, we examined whether the association between employment situation and mortality in women differed by education level and marital status.
Methods We investigated the association of employment situation with mortality among 16 692 women aged 40–59 years enrolled in the Japan Collaborative Cohort Study. Multivariate HRs and 95% CIs for total deaths by employment situation were calculated after adjustment for age, disease history, residential area, education level, marital status and number of children. We also conducted subgroup analysis by education level and marital status.
Results Multivariate HRs for mortality of part-time employees and self-employed workers were 1.48 (95% CI, 1.25 to 1.75) and 1.44 (95% CI, 1.21 to 1.72), respectively, with reference to women working full-time. Subgroup analysis by education level indicated that health effects in women according to employment situation were likely to be more evident in the low education-level group. Subgroup analysis by marital status indicated that this factor also affected the association between employment situation and risk of death.
Conclusions Among middle-aged Japanese women, employment situation was associated with mortality risk. Health effects were likely to differ by household structure and socioeconomic conditions.
- SOCIAL EPIDEMIOLOGY
Statistics from Altmetric.com
Studies show that people with better jobs enjoy better health.1 The inverse association between occupational position and various measures of chronic disease incidence, morbidity and mortality has been well documented in developed countries; people with higher occupational status are likely to be healthier.1–3 In addition, recent studies have shown that employment situation (ie, part-time, dispatch, contract or non-regular) could be a factor that influences health.4–10
Inconsistent results for the association between employment situation and health by gender have been reported. A cross-sectional study in Sweden showed that men with temporary employment reported less job control and lower health status compared with those who were permanently employed and self-employed; however, this was not true for women.11 Another previous study in Korea found that changing from permanent to irregular employment was associated with increased risk of depression, especially among women.12 Gender differences on the health effects by employment situation seem to be evident, although the magnitude of the effects depend on the population and society, and the use of different assessments of employment situation.
In most societies, the household responsibilities of women persist despite more women entering and playing a greater role in the workforce.13 The demands of household duties and responsibilities often necessitate women to work part-time or intermittently,14 ,15 thereby increasing work-family conflicts;16 ,17 this may create gender differences in the association between employment situation and health. However, few longitudinal studies have examined the health effects of employment situation among women, while considering their household environment and socioeconomic conditions.13 Moreover, to the best of our knowledge, no such studies exist in Japan.
Among Japanese women, data of workforce participation rate by age group show a bimodal pattern. This is in contrast to many Western societies, which show a convex shape similar to that among Japanese men.18 In this bimodal pattern, women tend to take a career break in their 30s, for such reasons as child-rearing, and return to the workforce in their 40s.19 Under the Japanese lifetime employment system, most women who re-enter the workforce have no choice other than part-time work.20 Consequently, there is a high proportion of irregularly employed workers (ie, non-regular and part-time workers) among women in Japan. The proportion of irregularly employed workers among women was 57.5% in 2012, in contrast with 22.1% among men.21 In recent decades, women's participation in the workforce has gradually increased in Japan.22 The reason for the increase may be that availability of the greater flexibility with respect to the frequencies and hours of working for part-time employment, which may allow more women to participate in the workforce.
It is important to note that the definition of part-time work in Japan differs from that in Western societies.19 ,23 Part-time employees are defined as individuals who work shorter hours per day or fewer working days per week than full-time workers. However, some part-time employees in Japan actually work nearly as much as full-time employees.24 ,25 The average wage for part-time jobs is the lowest among all employment situations and barely changes regardless of years of work experience.19 Such workers often receive less training, less bonus and no retirement benefits.25 ,26 Moreover, they usually have short, fixed-term contracts (eg, 1 month) or no employment contract, while full-time employees have open-ended employment contracts.25 ,27
Promoting female participation in the workforce has not improved the imbalance with respect to household duties in Japan, where women assume the greatest responsibility. 24 Thus, severe physical and psychological burdens due to their dual or more social roles among working women could affect their physical and mental health;28 in particular, part-time workers whose employment conditions were disadvantaged, compared to full-time workers, may result in more deteriorated health conditions.
In Japanese society, a limited number of women are with managerial or professional jobs, and women are likely to take subordinate jobs regardless of occupational class and education level.20 ,22 Therefore, employment situation may well identify social inequalities among women in Japan, because the differences in employment conditions, such as employment contract, wage, welfare (ie, social safety net), job training opportunities and job security, as stated above, have been identified.
Self-employed individuals were found to have greater control over their work and better health.11 Such work-related factors, like job autonomy or control, working hours and workload, were thought to provide greater flexibility for self-employed individuals than for employees; however, the flexibility may vary by gender. For example, the situation of self-employed women often differs from that of men in Japan (ie, self-employed men often own their own business and have great job control and autonomy, while self-employed women work as a worker in a family business and may not have the job control or autonomy that their male counterparts have).25
The health effects of employment situation among women could be influenced by their socioeconomic conditions.1 ,13 ,29 Women may self-select homemaking or being employed outside the home based on their available financial and material resources; some women need to work hard to support their household economy, while others work only to gain extra income. One could also argue that health effects of employment situation could depend on whether they have another breadwinner in their household. Thus, different employment situations could have different meanings depending on their socioeconomic circumstances; consequently, impacts on health could also vary.13
The objective of this study is to investigate the associations of employment situation (full-time employee, part-time employee or self-employed) with mortality over a 20-year period. We aimed to answer the following specific research questions:
Are there any differences in mortality, according to employment situations, among Japanese working women?
Are the associations noted above modified by socioeconomic conditions (ie, education level) and household structure (ie, marital status)?
This study used a cohort of 110 585 individuals aged 40–79 years who participated in the Japan Collaborative Cohort Study for Evaluation of Cancer Risk. Details of the study procedure are described elsewhere.30–32 Briefly, during 1988–1990, 46 395 men and 64 190 women living in 45 municipalities across Japan participated in the study. Participants completed self-administered questionnaires that included sociodemographic and lifestyle-related factors and medical history at baseline.
Of the total 64 190 women in the 45 municipalities, we excluded those in 12 municipalities from which information about occupation or education level was not obtained. Of the remaining 48 160 women, we excluded the following: those aged 60 years or older (n=21 869). To make sure that the effects of employment situation on mortality are not confounded by initial health status, we excluded women with a history of cancer or cardiovascular disease at baseline (n=697); women with missing information about employment situation (n=746); those who chose ‘other’ in the question asking about employment situation (n=1179); unemployed women (n=5052); and those with missing information about education level (n=1160) and marital status (n=817). We included the remaining 16 692 women as the final study population.
Employment situation was assessed using a self-reported baseline questionnaire. Participants were asked to choose one of the following: (1) full-time employee; (2) part-time employee or (3) self-employed.
Age (year), history of disease, residential area, education level, marital status and number of children at baseline were treated as confounding factors. Participants reported any history of the following at baseline: hypertension, diabetes mellitus, kidney disease, liver disease, gallstones or gallbladder disease, gastric or duodenal ulcer, and tuberculosis or pleuritis. We categorised women who reported having had at least one of those conditions as the ‘reported disease’ group. Education level was categorised into three groups, using the self-reported age of completed formal education in the baseline questionnaire: 15 years or younger; 16–18 years; and 19 years or older. For the subgroup analysis, we collapsed the latter two groups into one category because of the small numbers. Participants were categorised into three groups based on marital status: (1) married; (2) widowed/separated/divorced; (3) never married. For the subgroup analysis, we collapsed the widowed/separated/divorced and never married groups into one ‘unmarried’ category because of the small numbers. The number of children was determined on the basis of the reported number of children at baseline.
The end point of this study was all-cause mortality. Participants were followed up for 20 years (mean follow-up period, 17.7 years), from the date of consenting to the baseline survey until 31 December 2009. Exceptions to the follow-up period were: women from four municipalities who were followed until the end of 1999, those from three municipalities followed until the end of 2003, and women from two municipalities followed until the end of 2008. Investigators confirmed residence status and survival annually, using registers kept by public health centres in each municipality.30 The ethics committees of Nagoya University School of Medicine, University of Tsukuba and Osaka University approved the present study.
Proportions of sociodemographic factors were compared by employment situation and deaths using the χ2 test. Age-adjusted and area-adjusted HRs and 95% CIs for total deaths by employment situation were calculated (model 1). Multivariate HRs (95% CI) for risk of death by employment situation were calculated, adjusting for age, residential area, education level, history of disease, marital status and number of children at baseline, which were considered potential confounding variables (model 2). To further address the possibility that healthier women are more likely to have a full-time job, a propensity score was used to model the probability of having a full-time job, and inverse probability weighting was used to correct for possible selection bias in the analyses. This full-time job propensity score was calculated from a logistic regression as the probability of having a full-time job, given all relevant factors at baseline including age, history of disease, area, education level, marital status and number of children at baseline. Thus, the primary analysis was repeated by weighting the findings with the inverse probability of having a full-time job (model 3).33
We conducted subgroup analysis by education level and marital status to examine the modifying effect of the identified association by those factors.
Table 1 shows the basic characteristics of the study participants, and the distributions of those factors according to employment situation and mortality. During the mean follow-up period of 17.7 years, 1019 deaths (6%) were documented.
Table 2 shows the multivariate adjusted HRs for all-cause mortality according to employment situation. The multivariate adjusted HRs for death for part-time employees and self-employed workers were 1.48 (95% CI 1.25 to 1.75) and 1.44 (95% CI 1.21 to 1.672), respectively, with reference to women who worked full-time (model 2). The weighted model analysis results, using the inverse probability of having a full-time job, identified attenuated HRs but did not differ significantly from the results presented above (model 3).
The association between employment situation and mortality risk was significant, particularly in the low education-level group (table 3) In particular, the impact of part-time employment situation on mortality was modified by education level (p value for interaction =0.04). Subgroup analysis by marital status showed that unmarried, self-employed individuals had the higher HRs compared with the married group, although it was not statistically significant.
In this study of middle-aged Japanese women, those with part-time jobs or who were self-employed had a higher mortality risk than women who worked full-time. These associations seemed to be more evident in the low education-level group. Subgroup analysis by marital status also indicated that this factor also affected the association between employment situation and risk of death; unmarried self-employed women had a higher risk of death compared with full-time employees.
Longitudinal studies on employment situation and mortality are limited. One study in Finland reported that temporary employment was associated with increased risk of death from alcohol-related causes among women over a 10-year follow-up.7 Another study found a possible lower mortality among women working part-time in a 24-year follow-up study in Sweden.8 These studies appear to show conflicting results, but they may reflect different aspects related to employment situation, employment contract and working hours. Insecure conditions relating to employment contract seem to be detrimental to health, but the associated shorter working hours may have some protective effect. As noted in the introduction, having a part-time job in Japan does not necessarily mean reduced working hours. Unfortunately, we could not examine the effect of working hours and employment contract owing to a lack of information, but we speculate that the identified deteriorating health effect of working part-time may have been influenced by oppressive working conditions, including insecure employment.
Although only a limited number of similar studies have been conducted in Japan, our results are in line with the findings of those studies.26 ,34 A cross-sectional study with a national representative sample of employees in Japan found an association with poor mental health among women working as temporary contract workers.34 Another study reported that unstable employment with short working hours was associated with a lower proportion of self-rated poor health compared with the reports of permanently employed women; however, no such association was identified among women working more than 40 h/week.26 These studies suggest a deteriorating health effect of an insecure employment contract, and a protective health effect of shorter working hours. We could not differentiate between the aspects of employment contract and working hours according to employment situation, so we cannot make a direct comparison with these published results. However, we speculated that the instability of an insecure employment contract could be one of the factors that explained the increased mortality among part-time workers.
Self-employed women had a higher risk of death than those with full-time employment. The hypothesised possible benefits of being self-employed as job autonomy or control, working hours,11 may not be enjoyed by female self-employees in Japan. Previously conducted studies indicated lowered mortality from circulatory disorders among Japanese self-employed men with reference to employed counterparts, but no association was identified among women.35 Although we could not compare these results with the present study since the reference groups were different, both studies indicated that self-employed women were not likely to have the health-related advantages that their male counterparts enjoyed.
Selection bias must be considered as another possible explanation for the association between employment situation and mortality risk. For example, women with health problems may choose to work part-time or as homemakers. To reduce this possibility, we excluded women with a history of major illness and adjusted for several diseases at baseline in the statistical analysis. We also conducted complementary weighted analysis using the inverse probability of having a full-time job, though the analysis did not produce markedly different results.
The results of our subgroup analysis indicated that the health impacts of employment situation depend on education level and marital status. Women employed part-time with a lower level of education had the highest mortality risk among the groups that combined employment situation and education level. These findings are similar to those of studies that identified education level as an effect modifier in the relationship between employment situation and health.36–38 Given that education level often indicates socioeconomic conditions, the stronger association among women with lower education levels may be due to several factors, such as fewer opportunities and socioeconomic resources, which could have increased the deteriorating health impacts of insecure conditions.19
We also found that marital status influenced the association between employment situation and mortality risk. The mortality of part-time employees was different by marital status. Socioeconomic disadvantages among unmarried women without full-time work could elevate their mortality risk, although it was not statistically significant in a multiplicative sense of interaction. A previously conducted study in Japan identified that female workers living precariously in single-parent households suffer from poorest self-rated health,39 which suggested possible explanations for our identified high mortality among unmarried women without a full-time job.
This study is one of the few longitudinal investigations to observe the association between employment situation and risk of death, and it is the first to do so in Japan. However, there are several limitations. First, misclassification bias; the direction of misclassification bias is not certain because we do not have any information on the change of employment situation during the follow-up period in this study. The baseline period of this study is 1988–1990, which was just before the economic recession hit Japan. Thus, the recession may have impacted on the employment scenario of our study population. The unemployment rate in Japan has increased from 2.1% in 1990 to 4.7% in 2000.40 However, the female employment rate has increased in all age groups during the economic recession; homemakers started to work outside the home in order to support their household economy.19 However, considering the few opportunities for women who are part-time workers, or self-employed, to become full-time employees in Japan, we speculated that it is unlikely that this misclassification inflated our estimates. Second, although our study population consisted of people living in various areas throughout Japan, it may not be nationally representative. Notably, we did not include metropolitan areas. Thus, the generalisability of our results requires caution. Third, the definition of job type deserves consideration. Full-time work may have included both regular (permanent) and non-regular (temporary) employment; we lacked information about employment contracts and could not differentiate between these categories. However, since most non-regular employees were part-time workers at the baseline of this study (approximately 90% in 1990), and few non-regular full-time workers existed at that time,40 we believe the effect of misclassification to be small. Fourth, residual confounding could have occurred by measurement errors in our variables and unmeasured confounding variables. In particular, we included only education level as an indicator of socioeconomic conditions. It is clearly necessary to consider including comprehensive measures of socioeconomic conditions (ie, household income, husband's indicator of socioeconomic conditions) in future studies. Fifth, selection bias could not be ruled out in the observational study design.
Working part-time and being self-employed showed harmful health effects when compared with full-time employment among middle-aged Japanese women. This seemed to be more evident among less-educated and unmarried women. Employment situation was associated with risk of death, and health effects were likely to differ by social and economic conditions.
What is already known on this subject?
Inconsistent results have been reported for the association between employment situation and health by gender possibly due to the influence of gender role in some societies.
Few longitudinal studies—no studies in Japan—have examined the health effects of employment situation among women, taking social and economic conditions into consideration.
What this study adds?
Working part-time and being self-employed had a higher mortality than working full-time among middle-aged Japanese women.
Social and economic conditions affected the association between employment situation and mortality.
The authors thank all staff members who were involved in this study for their valuable help in conducting the baseline survey and follow-up. The present members of the JACC Study Group who co-authored this paper are: AT (present chairperson of the study group), Hokkaido University Graduate School of Medicine; Dr Mitsuru Mori and Dr Fumio Sakauchi, Sapporo Medical University School of Medicine; Dr Yutaka Motohashi, Akita University School of Medicine; Dr Ichiro Tsuji, Tohoku University Graduate School of Medicine; Dr Yosikazu Nakamura, Jichi Medical School; HI, Osaka University School of Medicine; Dr Haruo Mikami, Chiba Cancer Center; Dr Michiko Kurosawa, Juntendo University School of Medicine; YH, Yokohama Soei University; Dr Naohito Tanabe, University of Niigata, Prefecture; Dr Koji Tamakoshi, Nagoya University Graduate School of Health Science; Dr Kenji Wakai, Nagoya University Graduate School of Medicine; Dr Shinkan Tokudome, National Institute of Health and Nutrition; Dr Koji Suzuki, Fujita Health University School of Health Sciences; Dr Shuji Hashimoto, Fujita Health University School of Medicine; Dr Shogo Kikuchi, Aichi Medical University School of Medicine; Dr Yasuhiko Wada, Faculty of Nutrition, University of Kochi; Dr Takashi Kawamura, Kyoto University Center for Student Health; Dr Yoshiyuki Watanabe, Kyoto Prefectural University of Medicine Graduate School of Medical Science; Dr Kotaro Ozasa, Radiation Effects Research Foundation; Dr Tsuneharu Miki, Kyoto Prefectural University of Medicine Graduate School of Medical Science; Dr Chigusa Date, School of Human Science and Environment, University of Hyogo; Dr Kiyomi Sakata, Iwate Medical University; Dr Yoichi Kuroza wa, Tottori University Faculty of Medicine; Dr Takesumi Yoshimura and Dr Yoshihisa Fujino, University of Occupational and Environmental Health; Dr Akira Shibata, Kurume University; Dr Naoyuki Okamoto, Kanagawa Cancer Center; and Dr Hideo Shio, Moriyama Municipal Hospital.
Contributors KH contributed to the conception of the work, analysis or interpretation of data for the work, drafting or revising it critically for important intellectual content. AI contributed to the analysis and interpretation of data for the work, and revising it critically for important intellectual content. HI, YF and AT contributed to the conception of the work, interpretation of data for the work, revising it critically for important intellectual content. All authors approved the version to be published.
Funding This study was supported by grants-in-aid for scientific research from the Ministry of Education, Science, Sports and Culture of Japan (Monbusho); 61010076, 62010074, 63010074, 1010068, 2151065, 3151064, 4151063, 5151069, 6279102, 11181101, 17015022, 18014011 and 20014026.
Competing interests None declared.
Patient consent Obtained.
Ethics approval The ethics committees of Nagoya University School of Medicine, University of Tsukuba and Osaka University.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.