2014 Volume 56 Issue 6 Pages 485-497
Objectives: As the number of dual-earner couples in Japan has increased, work-life balance has become important. This study aimed to examine the factors that contribute to work-family conflict. Methods: The participants included 3,594 (2,332 men and 1,262 women) civil servants aged 20–59 working for local government on the west coast of Japan. Logistic regression analysis was used to evaluate whether work, family, or lifestyle characteristics were associated with work-family conflict. Results: For men, family-to-work conflict was associated with being elderly, having low-grade employment, working long hours, raising children, and sleeping shorter hours. For women, being married and raising children were strong determinants of family-to-work conflict, and being middle-aged, working long hours, and sleeping shorter hours were also associated with this type of conflict. Regarding work-to-family conflict, working long hours was the strongest determinant of conflict in both sexes. In men, being elderly, living with family, eating dinner late, and sleeping shorter hours were also associated with work-to-family conflict. In women, having high-grade employment, being married, raising children, and eating dinner late were associated with work-to-family conflict. Conclusions: This study showed that working long hours was the primary determinant of work-to-family conflict in both sexes and that being married and raising children were strong factors of family-to-work conflict in women only. Sex differences may reflect divergence of the social and domestic roles of men and women in Japanese society. To improve the work-life balance, general and sex-specific health policies may be required.
(J Occup Health 2014; 56: 485–497)
In recent years, work-life balance (WLB) has been recognized as essential for employees throughout life, and WLB is an important part of socioeconomic policies. Japan, which has adopted a conservative welfare state regime1, 2), has faced many socioeconomic, sociodemographic, and sociocultural changes in recent years3–6) and has been behind Western countries with regard to active participation of women in the labor force and appointment of women as managers; therefore, an incentive to accelerate introduction of the concept of WLB exists. However, small and medium-sized enterprises are reluctant to introduce the concept of WLB due to the cost and conservative organizational culture; there are some differences in the level of concern with regard to WLB. Small and medium-sized enterprises made up 99.7% of enterprises and accounted for approximately 70% of employment in 20067); therefore, for these enterprises, some political inducements and sustainable role models for introduction of the concept of WLB are necessary.
Management of work- and family-related stress8) is now a key issue for employees and their families. Therefore, the Japanese private and public sectors3) have begun to promote WLB policies, based on successful methods and policies used in liberal and social democratic welfare states1, 2), in pursuit of socially worthwhile goals and sustainable role models. Kahn's theory of work-family (W-F) conflict9) has been modified to form two directions of conflict10): family-to-work (F-to-W), in which family interferes with work, and work-to-family (W-to-F), in which work interferes with family. A meta-analysis11, 12) suggested that the factors the contributing to F-to-W and W-to-F conflict differ; work characteristics have a is necessary to investigate the effects of W-F conflict gender roles and faces serious problems including an aging population, declining birthrates and shrinkage of greater effect on W-to-F conflict, and family characteristics have a greater effect on F-to-W conflict. These organizational psychology and organizational behavior theories have gradually been adopted in occupational health and preventive medicine. Previous studies have examined relationships between W-F conflict and physical health, mental health and lifestyle and relationships between W-F conflict and cardiovascular health13), psychiatric disorders14), physical and mental functioning15), mental health16, 17), sickness absence18), fatigue19) and sleep quality15), in addition to lifestyle factors including unhealthy behaviors20, 21). Moreover, an international collaborative study conducted in London, Helsinki, and Japan examined associations between W-to-F and F-to-W conflict levels and domestic roles16) and determined that female Japanese workers exhibited the highest conflict levels. Therefore, it in Japanese society, which has maintained traditional its working population3)
F-to-W and W-to-F conflict may explain work- and family-related stress levels. Therefore, this study sought to answer the following research questions: How strongly does each factor affect W-F conflict? Which groups need to reduce W-F conflict? Will work, family and lifestyle characteristics explain the effects of W-F conflict? Previous studies have not treated F-to-W and W-to-F conflict as outcomes but have used them as independent variables. The working population, which has similar working and residential environments with respect to wage-payment systems, hierarchy systems, personnel management, organizational culture and regions, is helpful in investigating the effects of WFC and feeding the results back to facilitate the development of a sustainable role model. To examine the effects of W-to-F and F-to-W conflict on employees and their families further, a large-scale cross-sectional study examining both types of conflict as health outcomes has been anticipated.
This study included civil servants who participated in the Japanese Civil Servants Study (JACS study)22), an international collaborative study conducted in association with the British Civil Servants Study (the Whitehall II study)15, 16, 22–24). The JACS study was set up in 1998 as a part of the annual health checkups based on the Industrial Safety and Health Law. The personnel department of a local government was responsible for the JACS study, and we collaborated with respect to the surveys. Phase 3 of the JACS study was conducted in 2008 via a postal survey sent to clerical, technical and professional local government employees aged 20–65 who were recruited through the personnel section of a local government agency on the west coast of Japan. The contents and ethical aspects of the study were approved by an ad hoc civil service committee, which comprised an ordinary member of the Industrial Safe and Health committee, members from the personnel section of the local government and representatives from a labor union. Prior to the survey, we formulated a plan with the committee concerning data protection, publication of the results and the duty of confidentiality. In addition, we have supported their industrial health activities by performing a health survey, publishing newsletters on health topics for employees, reporting the findings of the survey to the personnel section and holding a short course and seminar on health and disease prevention for employees in general and administrative positions. In total, 4,429 employees completed the questionnaire (response rate: 91.0%). Participants participated voluntarily, and their informed consent was obtained. We analyzed the data of 3,594 participants (2,332 men and 1,262 women), aged 20–59, who provided complete data regarding age, sex, work characteristics, family characteristics and lifestyle.
MeasurementQuestionnaire items regarding work-family conflict were taken from the Whitehall II Study. The English questionnaire was translated into Japanese and retranslated into English by a person who was unfamiliar with the original. The translated items were compared with the original version by the Whitehall II Study researchers. Eight work-family conflict items (four F-to-W and four W-to-F) were adopted from the National Study of Midlife Development in the US (MIDUS)25). The F-to-W conflict items concerned family-to-work interference, and the W-to-F conflict items concerned work-to-family interference in the Whitehall II Study (see Appendix)20). Each item had three response options, ranging from 1 (never) to 3 (often). Scores from each scale, ranging from 4 to 12, were calculated separately and grouped into tertiles, with the first tertile considered to represent high conflict levels.
F-to-W conflict and W-to-F conflict were adopted as binary health outcomes used to examine levels of interactive interference between work and family life. We selected four factors, age group, work characteristics, family characteristics and lifestyle, which are relevant to balancing work and family life. Work characteristics included two work condition variables, occupational grade and working hours, considered to affect mental health and/or work-family conflict15). Occupational grades were categorized into two groups according to administrative position because the percentage of female (relative to male) middle and senior managers was extremely small. Working hours were categorized into three groups, <9, 9–11 hours and ≥11 hours according to the normal overtime standards. In order to examine the distinctive factors of Japanese family structures, we included marital status, living with family, living with parent/parents and having a child or children as family characteristic variables. We also included lifestyle factors to investigate employees' actual time management and health behavior conditions. We chose the following timing and health behavior variables: dinner time, bedtime, waking time, sleep duration, smoking status and body mass index (BMI), which were categorized into two groups (below and above the median) except in the case of smoking status and BMI. Smoking status was categorized into two groups, current or never/former. BMI was categorized into three groups based on the Japanese standards for health checkups26): <18.5, 18.5–24.9, and ≥25.0 kg/m2.
Statistical analysisLogistic regression analysis was performed to examine the relationship between F-to-W and W-to-F conflict and the independent variables according to sex. Odds ratios (OR) and 95% confidence intervals were calculated. The final multivariate models were examined using the Hosmer-Lemeshow Goodness-of-Fit Test27). All statistical analyses were performed using the SPSS statistical software package (version 20.0.J, released 2011, IBM Corp., Armonk, NY, USA), and a p value of <0.05 (two tailed) was considered statistically significant.
The mean age of the participants was 45.0 years (SD=9.35) in men and 39.2 years (SD=10.75) in women (p<0.001). More women than men were single, childless and in low-grade employment. More than 75% of all participants lived with family, and almost 50% lived with a parent(s) or parent(s)-in-law. More women than men worked longer hours, slept shorter hours, woke earlier and had higher work-family conflict levels. The percentage of overtime work was highest for men in their 30s and women in their 20s. The percentage of subjects with a high risk of F-to-W conflict was highest in those in their 40s in both sexes, followed by those in their 30s, 50s and 20s. In addition, the percentage of subjects with a high risk of W-to-F conflict was highest in those in their 30s in both sexes, followed by those in their 40s, 50s and 20s. The mean F-to-W conflict score was 6.0 (SD=1.87) in men and 6.4 (SD=2.13) in women (p<0.001). In each age group, the mean score for F-to-W conflict was higher in women than in men, and the score was highest for subjects in their 40s. The mean W-to-F conflict score was 6.5 (SD=1.87) in men and 7.1 (SD=1.98) in women (p<0.001). In almost every age group, the mean score for W-to-F conflict was higher in women than in men. The men and women in their 30s and 40s showed similar high scores. Cronbach's alpha coefficient28) was 0.73 for F-to-W conflict and 0.72 for W-to-F conflict (Table 1).
| Men | ||||||
|---|---|---|---|---|---|---|
| Total N=2,332 | 20–29 | 30–39 | 40–49 | 50–59 | p value | |
| N=129 | N=587 | N=751 | N=865 | |||
| Occupational grade (%) | ||||||
| Nonadministrative | 65.1 | 100.0 | 96.8 | 70.0 | 34.1 | |
| Administrative | 34.9 | 0.0 | 3.2 | 30.0 | 65.9 | p<0.001b |
| Working hours (%) | ||||||
| <9 h a day | 66.0 | 56.6 | 51.1 | 65.9 | 77.7 | |
| 9–11 h a day | 25.7 | 34.1 | 33.6 | 26.1 | 18.7 | |
| ≥11 h a day | 8.3 | 9.3 | 15.3 | 8.0 | 3.6 | p<0.001b |
| Marital status (%) | ||||||
| No | 16.2 | 72.1 | 23.5 | 13.3 | 5.3 | |
| Yes | 83.8 | 27.9 | 76.5 | 86.7 | 94.7 | p<0.001b |
| Living with family (%) | ||||||
| No | 6.1 | 19.4 | 8.9 | 4.8 | 3.4 | |
| Yes | 93.9 | 80.6 | 91.1 | 95.2 | 96.6 | p<0.001b |
| Living with a parent/parents (%) | ||||||
| No | 57.2 | 43.4 | 65.1 | 59.4 | 51.9 | |
| Yes | 42.8 | 56.6 | 34.9 | 40.6 | 48.1 | p<0.001b |
| Have a child/children (%) | ||||||
| No | 37.1 | 86.8 | 38.5 | 24.5 | 39.7 | |
| Yes | 62.9 | 13.2 | 61.5 | 75.5 | 60.3 | p<0.001b |
| Dinner time (%) | ||||||
| Before 21:00 | 78.0 | 69.8 | 67.3 | 73.2 | 90.6 | |
| 21:00 and after | 22.0 | 30.2 | 32.7 | 26.8 | 9.4 | p<0.001b |
| Bedtime (%) | ||||||
| Before 23:30 | 49.3 | 22.5 | 41.9 | 43.5 | 63.4 | |
| 23:30 and after | 50.7 | 77.5 | 58.1 | 56.5 | 36.6 | p<0.001b |
| Waking time (%) | ||||||
| Before 6:30 | 55.6 | 27.1 | 39.4 | 55.7 | 70.9 | |
| 6:30 and after | 44.4 | 72.9 | 60.6 | 44.3 | 29.1 | p<0.001b |
| Sleep duration (%) | ||||||
| ≥7 h a day | 45.6 | 40.3 | 44.3 | 40.1 | 52.0 | |
| <7 h a day | 54.4 | 59.7 | 55.7 | 59.9 | 48.0 | p<0.001b |
| Smoking status (%) | ||||||
| Never/former | 72.3 | 77.5 | 77.3 | 72.0 | 68.4 | |
| Current | 27.7 | 22.5 | 22.7 | 28.0 | 31.6 | p<0.001b |
| Body mass index (kg/m2) (%) | ||||||
| <18.5 | 2.9 | 4.7 | 4.3 | 2.1 | 2.1 | |
| 18.5–24.9 | 70.8 | 85.3 | 72.2 | 70.2 | 67.5 | |
| ≥25.0 | 26.2 | 10.1 | 23.5 | 27.7 | 30.4 | p<0.001b |
| Work-to-family conflict (%) | ||||||
| 1st–2nd tertile | 59.5 | 84.5 | 55.0 | 53.0 | 64.5 | |
| 3rd tertile (high risk) | 40.5 | 15.5 | 45.0 | 47.0 | 35.5 | p<0.001b |
| Family-to-work conflict (%) | ||||||
| 1st–2nd tertile | 70.5 | 84.5 | 64.4 | 65.0 | 77.3 | |
| 3rd tertile (high risk) | 29.5 | 15.5 | 35.6 | 35.0 | 22.7 | p<0.001b |
| Mean family-to-work conflicta | 6.0 (1.87) | 4.96 (1.47) | 6.25 (2.03) | 6.29 (1.87) | 5.82 (1.72) | p<0.001c |
| Mean work-to-family conflicta | 6.5 (1.87) | 5.74 (1.63) | 6.75 (1.96) | 6.77 (1.90) | 6.18 (1.75) | p<0.001c |
| Women | ||||||
| Total N=1,262 | 20–29 | 30–39 | 40–49 | 50–59 | p value | |
| N=309 | N=368 | N=291 | N=294 | |||
| Occupational grade (%) | ||||||
| Nonadministrative | 89.1 | 100.0 | 97.3 | 86.3 | 70.1 | |
| Administrative | 10.9 | 0.0 | 2.7 | 13.7 | 29.9 | p<0.001b |
| Working hours (%) | ||||||
| <9 h a day | 53.2 | 41.4 | 55.2 | 59.1 | 57.5 | |
| 9–11 h a day | 34.2 | 37.5 | 34.0 | 32.0 | 33.3 | |
| ≥11 h a day | 12.5 | 21.0 | 10.9 | 8.9 | 9.2 | p<0.001b |
| Marital status (%) | ||||||
| No | 38.1 | 84.8 | 38.9 | 18.6 | 7.5 | |
| Yes | 61.9 | 15.2 | 61.1 | 81.4 | 92.5 | p<0.001b |
| Living with family (%) | ||||||
| No | 12.0 | 24.6 | 13.3 | 5.8 | 3.1 | |
| Yes | 88.0 | 75.4 | 86.7 | 94.2 | 96.9 | p<0.001b |
| Living with a parent/parents (%) | ||||||
| No | 49.8 | 38.2 | 63.9 | 49.5 | 44.9 | |
| Yes | 50.2 | 61.8 | 36.1 | 50.5 | 55.1 | p<0.001b |
| Have a child/children (%) | ||||||
| No | 52.5 | 91.6 | 52.4 | 24.4 | 39.1 | |
| Yes | 47.5 | 8.4 | 47.6 | 75.6 | 60.9 | p<0.001b |
| Dinner time (%) | ||||||
| Before 21:00 | 76.3 | 68.3 | 76.4 | 81.1 | 79.9 | |
| 21:00 and after | 23.7 | 31.7 | 23.6 | 18.9 | 20.1 | p<0.001b |
| Bedtime (%) | ||||||
| Before 23:30 | 43.3 | 33.3 | 47.0 | 41.2 | 51.4 | |
| 23:30 and after | 56.7 | 66.7 | 53.0 | 58.8 | 48.6 | p<0.001b |
| Waking time (%) | ||||||
| Before 6:30 | 65.8 | 42.1 | 60.3 | 75.9 | 87.1 | |
| 6:30 and after | 34.3 | 57.9 | 39.7 | 24.1 | 12.9 | p<0.001b |
| Sleep duration (%) | ||||||
| ≥7 h a day | 34.2 | 38.5 | 41.8 | 28.2 | 25.9 | |
| <7 h a day | 65.8 | 61.5 | 58.2 | 71.8 | 74.1 | p<0.001b |
| Smoking status (%) | ||||||
| Never/former | 96.4 | 96.1 | 97.3 | 97.6 | 94.6 | |
| Current | 3.6 | 3.9 | 2.7 | 2.4 | 5.4 | p=0.171b |
| Body mass index (kg/m2) (%) | ||||||
| <18.5 | 14.4 | 23.6 | 14.1 | 9.6 | 7.8 | |
| 18.5–24.9 | 76.4 | 73.1 | 78.3 | 78.4 | 76.5 | |
| ≥25.0 | 9.2 | 3.2 | 7.6 | 12.0 | 15.6 | p<0.001b |
| Work-to-family conflict (%) | ||||||
| 1st–2nd tertile | 52.5 | 79.3 | 46.2 | 35.1 | 49.3 | |
| 3rd tertile (high risk) | 47.5 | 20.7 | 53.8 | 64.9 | 50.7 | p<0.001b |
| Family-to-work conflict (%) | ||||||
| 1st–2nd tertile | 59.4 | 69.6 | 54.3 | 54.6 | 59.9 | |
| 3rd tertile (high risk) | 40.6 | 30.4 | 45.7 | 45.4 | 40.1 | p<0.001b |
| Mean family-to-work conflicta | 6.4 (2.13) | 5.20 (1.64) | 6.76 (2.25) | 7.27 (2.05) | 6.53 (1.94) | p<0.001c |
| Mean work-to-family conflicta | 7.1 (1.98) | 6.69 (2.00) | 7.21 (2.06) | 7.33 (1.89) | 7.01 (1.90) | p<0.001c |
High F-to-W conflict levels in men were observed in all age groups after adjusting for age and work characteristics (Model 2). ORs showed a negative impact for the administrative position group and a positive impact for the overtime work group. However, ORs for all age groups were visibly reduced in the age- and family characteristics-adjusted model (Model 3). In the final model (Model 5), in which the variables were adjusted for lifestyle and Model 4, age group, occupational grade, working hours, having a child and sleep duration were found to be correlated with F-to-W conflict in men in the following groups: 30s (OR=2.94 [1.73–4.98]), 40s (OR=3.24 [1.90–5.52]), 50s (OR=2.51 [1.44–4.38]), administrative position (OR=0.75 [0.61–0.93]), overtime (9–11 hours daily; OR=1.32 [1.07–1.64]), parental (OR=1.65 [1.33–2.05]) and short sleep duration (OR=1.58 [1.28–1.96]) (Table 2).
| High (%) | Model 1 OR (95% CI) | Model 2 OR (95% CI) | Model 3 OR (95% CI) | Model 4 OR (95% CI) | Model 5 OR (95% CI) | |
|---|---|---|---|---|---|---|
| Age | ||||||
| 20–29 | 15.5 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| 30–39 | 45.0 | 4.45 (2.69–7.37)*** | 4.42 (2.67–7.33)*** | 3.04 (1.80–5.12)*** | 2.98 (1.77–5.04)*** | 2.94 (1.73–4.98)*** |
| 40–49 | 47.0 | 4.83 (2.94–7.95)*** | 5.41 (3.27–8.96)*** | 3.02 (1.79–5.10)*** | 3.38 (2.00–5.73)*** | 3.24 (1.90–5.52)*** |
| 50–59 | 35.5 | 3.00 (1.82–4.93)*** | 3.81 (2.27–6.41)*** | 1.97 (1.17–3.34)* | 2.56 (1.48–4.41)*** | 2.51 (1.44–4.38)*** |
| Occupational grade | ||||||
| Nonadministrative | 42.5 | 1.00 | 1.00 | 1.00 | ||
| Administrative | 36.7 | 0.78 (0.63–0.97)* | 0.75 (0.60–0.92)** | 0.75 (0.61–0.93)** | ||
| Working hours | ||||||
| <9 h a day | 37.7 | 1.00 | 1.00 | 1.00 | ||
| 9–11 h a day | 45.1 | 1.37 (1.13–1.68)** | 1.40 (1.15–1.71)*** | 1.32 (1.07–1.64)** | ||
| ≥11 h a day | 48.7 | 1.53 (1.12–2.08)** | 1.54 (1.12–2.12)** | 1.32 (0.93–1.89) | ||
| Marital status | ||||||
| No | 26.3 | 1.00 | 1.00 | 1.00 | ||
| Yes | 43.2 | 1.31 (0.93–1.84) | 1.33 (0.94–1.87) | 1.39 (0.98–1.97) | ||
| Living with family | ||||||
| No | 26.8 | 1.00 | 1.00 | 1.00 | ||
| Yes | 41.4 | 1.21 (0.77–1.93) | 1.23 (0.77–1.96) | 1.22 (0.77–1.95) | ||
| Living with parents | ||||||
| No | 43.1 | 1.00 | 1.00 | 1.00 | ||
| Yes | 36.9 | 0.87 (0.72–1.04) | 0.87 (0.72–1.05) | 0.86 (0.71–1.04) | ||
| Have children | ||||||
| No | 29.8 | 1.00 | 1.00 | 1.00 | ||
| Yes | 46.8 | 1.58 (1.28–1.96)*** | 1.61 (1.29–1.99)*** | 1.65 (1.33–2.05)*** | ||
| Dinner time | ||||||
| Before 21:00 | 39.0 | 1.00 | ||||
| 21:00 and after | 45.8 | 1.10 (0.86–1.41) | ||||
| Bedtime | ||||||
| Before 23:30 | 39.5 | 1.00 | ||||
| 23:30 and after | 41.5 | 0.85 (0.68–1.07) | ||||
| Waking time | ||||||
| Before 6:30 | 41.0 | 1.00 | ||||
| 6:30 and after | 39.8 | 1.02 (0.84–1.24) | ||||
| Sleeping duration | ||||||
| ≥7 h a day | 35.7 | 1.00 | ||||
| <7 h a day | 44.4 | 1.58 (1.28–1.96)*** | ||||
| Smoking status | ||||||
| Never/former | 40.3 | 1.00 | ||||
| Current | 40.9 | 1.07 (0.88–1.30) | ||||
| Body mass index | ||||||
| <18.5 | 41.1 | 0.87 (0.72–1.07) | ||||
| 18.5–24.9 | 47.1 | 1.00 | ||||
| ≥25.0 | 38.1 | 1.35 (0.81–2.24) |
Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; high, prevalence of high conflict. Model 1 is adjusted for age. Model 2 is adjusted for age and work characteristics. Model 3 is adjusted for age and family characteristics. Model 4 is adjusted for age, work characteristics and family characteristics. Model 5 is adjusted for age, work characteristics, family characteristics and lifestyle factors. Hosmer-Lemeshow Goodness-of-Fit Test: 0.410.
High levels of F-to-W conflict were observed in women in all age groups in Model 2. In Model 3, ORs for all age groups decreased significantly. The married women and parental groups, in particular, showed higher F-to-W conflict levels than similar groups of men. In Model 5, age group, working hours, marital status, having a child and sleep duration were found to be correlated with F-to-W conflict in women in the following groups: 30s (OR=1.94 [1.30–2.91]), 40s (OR=1.78 [1.11–2.87]), overtime (9–11 hours daily; OR=1.50 [1.12–2.03]), overtime (≥11 hours daily; OR=1.94 [1.22–3.10]), married (OR=4.81 [3.12–7.42]), parental (OR=2.80 [2.06–3.83]) and short sleep duration (OR=1.41 [1.01–1.97]) (Table 3).
| High (%) | Model 1 OR (95% CI) | Model 2 OR (95% CI) | Model 3 OR (95% CI) | Model 4 OR (95% CI) | Model 5 OR (95% CI) | |
|---|---|---|---|---|---|---|
| Age | ||||||
| 20–29 | 20.7 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| 30–39 | 53.8 | 4.46 (3.16–6.28)*** | 4.82 (3.39–6.84)*** | 1.78 (1.19–2.63)** | 1.92 (1.29–2.87)*** | 1.94 (1.30–2.91)*** |
| 40–49 | 64.9 | 7.09 (4.92–10.22)*** | 7.91 (5.41–11.57)*** | 1.58 (1.01–2.48)* | 1.79 (1.13–2.85)* | 1.78 (1.11–2.87)* |
| 50–59 | 50.7 | 3.93 (2.75–5.63)*** | 4.44 (3.02–6.53)*** | 0.80 (0.51–1.26) | 0.90 (0.56–1.46) | 0.90 (0.55–1.48) |
| Occupational grade | ||||||
| Nonadministrative | 46.7 | 1.00 | 1.00 | 1.00 | ||
| Administrative | 54.3 | 0.88 (0.60–1.31) | 0.81 (0.53–1.25) | 0.79 (0.51–1.22) | ||
| Working hours | ||||||
| <9 h a day | 46.4 | 1.00 | 1.00 | 1.00 | ||
| 9–11 h a day | 48.8 | 1.29 (0.99–1.67) | 1.57 (1.17–2.10)*** | 1.50 (1.12–2.03)** | ||
| ≥11 h a day | 48.7 | 1.67 (1.14–2.46)** | 2.12 (1.39–3.24)*** | 1.94 (1.22–3.10)** | ||
| Marital status | ||||||
| No | 20.2 | 1.00 | 1.00 | 1.00 | ||
| Yes | 64.4 | 4.45 (2.92–6.79)*** | 4.77 (3.11–7.34)*** | 4.81 (3.12–7.42)*** | ||
| Living with family | ||||||
| No | 18.5 | 1.00 | 1.00 | 1.00 | ||
| Yes | 51.5 | 0.91 (0.51–1.60) | 0.99 (0.55–1.76) | 1.02 (0.57–1.82) | ||
| Living with parents | ||||||
| No | 52.0 | 1.00 | 1.00 | 1.00 | ||
| Yes | 43.1 | 1.14 (0.83–1.56) | 1.13 (0.83–1.55) | 1.13 (0.82–1.55) | ||
| Have children | ||||||
| No | 27.5 | 1.00 | 1.00 | 1.00 | ||
| Yes | 69.7 | 2.79 (2.06–3.77)*** | 2.75 (2.03–3.73)*** | 2.80 (2.06–3.83)*** | ||
| Dinner time | ||||||
| Before 21:00 | 47.9 | 1.00 | ||||
| 21:00 and after | 46.5 | 1.12 (0.79–1.58) | ||||
| Bedtime | ||||||
| Before 23:30 | 52.1 | 1.00 | ||||
| 23:30 and after | 44.1 | 0.87 (0.63–1.20) | ||||
| Waking time | ||||||
| Before 6:30 | 52.7 | 1.00 | ||||
| 6:30 and after | 37.6 | 1.15 (0.84–1.58) | ||||
| Sleeping duration | ||||||
| ≥7 h a day | 44.5 | 1.00 | ||||
| <7 h a day | 49.1 | 1.41 (1.01–1.97)* | ||||
| Smoking status | ||||||
| Never/former | 48.2 | 1.00 | ||||
| Current | 31.1 | 0.67 (0.33–1.38) | ||||
| Body mass index | ||||||
| <18.5 | 47.8 | 1.08 (0.74–1.59) | ||||
| 18.5–24.9 | 41.8 | 1.00 | ||||
| ≥25.0 | 54.3 | 1.11 (0.71–1.73) |
Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; high, prevalence of high conflict. Model 1 is adjusted for age. Model 2 is adjusted for age and work characteristics. Model 3 is adjusted for age and family characteristics. Model 4 is adjusted for age, work characteristics and family characteristics. Model 5 is adjusted for age, work characteristics, family characteristics and lifestyle factors. Hosmer-Lemeshow Goodness-of-Fit Test: 0.087.
High W-to-F conflict levels in men were observed in all age groups, and working overtime had stronger effects on W-to-F than F-to-W conflict (in Model 2). ORs for every age group were attenuated mildly in Model 3. In Model 5, age group, working hours, living with family, dinner time and sleep duration were found to be correlated with W-to-F conflict in men in the following groups: 30s (OR=2.86 [1.62–5.04]), 40s (OR=3.58 [2.00–6.40]), 50s (OR=2.41 [1.31–4.46]), overtime (9–11 hours daily; OR=3.10 [2.48–3.88]), overtime (≥11 hours daily; OR=8.87 [5.99–13.14]), living with family (OR=1.75 [1.04–2.95]), late dinner time (OR=1.58 [1.22–2.05]) and short sleep duration (OR=1.34 [1.05–1.71]) (Table 4).
| High (%) | Model 1 OR (95% CI) | Model 2 OR (95% CI) | Model 3 OR (95% CI) | Model 4 OR (95% CI) | Model 5 OR (95% CI) | |
|---|---|---|---|---|---|---|
| Age | ||||||
| 20–29 | 15.5 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| 30–39 | 35.6 | 3.01 (1.82–5.00)*** | 3.01 (1.75–5.16)*** | 2.78 (1.65–4.70)*** | 2.89 (1.65–5.06)*** | 2.86 (1.62–5.04)*** |
| 40–49 | 35.0 | 2.94 (1.78–4.84)*** | 3.82 (2.22–6.56)*** | 2.72 (1.61–4.60)*** | 3.65 (2.07–6.46)*** | 3.58 (2.00–6.40)*** |
| 50–59 | 22.7 | 1.60 (0.97–2.64) | 2.28 (1.29–4.02)** | 1.47 (0.86–2.50) | 2.22 (1.22–4.03)** | 2.41 (1.31–4.46)** |
| Occupational grade | ||||||
| Nonadministrative | 29.4 | 1.00 | 1.00 | 1.00 | ||
| Administrative | 29.7 | 1.20 (0.94–1.54) | 1.20 (0.93–1.54) | 1.18 (0.92–1.51) | ||
| Working hours | ||||||
| <9 h a day | 18.2 | 1.00 | 1.00 | 1.00 | ||
| 9–11 h a day | 44.4 | 3.58 (2.90–4.44)*** | 3.60 (2.91–4.46)*** | 3.10 (2.48–3.88)*** | ||
| ≥11 h a day | 73.1 | 12.27 (8.59–17.52)*** | 12.53 (8.75–17.95)*** | 8.87 (5.99–13.14)*** | ||
| Marital status | ||||||
| No | 26.3 | 1.00 | 1.00 | 1.00 | ||
| Yes | 30.1 | 1.08 (0.76–1.53) | 0.89 (0.61–1.30) | 0.95 (0.64–1.39) | ||
| Living with family | ||||||
| No | 25.4 | 1.00 | 1.00 | 1.00 | ||
| Yes | 29.8 | 1.36 (0.85–2.18) | 1.79 (1.06–3.01)* | 1.75 (1.04–2.95)* | ||
| Living with parents | ||||||
| No | 31.8 | 1.00 | 1.00 | 1.00 | ||
| Yes | 26.4 | 0.81 (0.66–0.99)* | 0.84 (0.68–1.05) | 0.84 (0.67–1.04) | ||
| Have children | ||||||
| No | 27.4 | 1.00 | 1.00 | 1.00 | ||
| Yes | 30.7 | 0.94 (0.74–1.18) | 1.01 (0.79–1.30) | 1.06 (0.82–1.35) | ||
| Dinner time | ||||||
| Before 21:00 | 23.1 | 1.00 | ||||
| 21:00 and after | 52.0 | 1.58 (1.22–2.05)*** | ||||
| Bedtime | ||||||
| Before 23:30 | 23.8 | 1.00 | ||||
| 23:30 and after | 35.0 | 0.92 (0.71–1.19) | ||||
| Waking time | ||||||
| Before 6:30 | 27.5 | 1.00 | ||||
| 6:30 and after | 32.0 | 1.16 (0.93–1.46) | ||||
| Sleeping duration | ||||||
| ≥7 h a day | 22.9 | 1.00 | ||||
| <7 h a day | 35.1 | 1.34 (1.05–1.71)* | ||||
| Smoking status | ||||||
| Never/former | 30.2 | 1.00 | ||||
| Current | 27.8 | 1.01 (0.80–1.25) | ||||
| Body mass index | ||||||
| <18.5 | 29.5 | 0.79 (0.42–1.49) | ||||
| 18.5–24.9 | 22.1 | 1.00 | ||||
| ≥25.0 | 30.2 | 1.04 (0.83–1.30) |
Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; high, prevalence of high conflict. Model 1 is adjusted for age. Model 2 is adjusted for age and work characteristics. Model 3 is adjusted for age and family characteristics. Model 4 is adjusted for age, work characteristics and family characteristics. Model 5 is adjusted for age, work characteristics, family characteristics and lifestyle factors. Hosmer-Lemeshow Goodness-of-Fit Test: 0.641.
High W-to-F conflict levels were observed in women in all age groups (in Model 2). Although the ORs for all age groups decreased after adjusting for age and family characteristics, the statistical significance in the 40s and 50s age groups disappeared. In Model 5, age group, occupational grade, working hours, marital status, having a child and dinner time were found to be correlated with W-to-F conflict in women in the following groups: 30s (OR=1.68 [1.01–2.40]), overtime (9–11 hours daily; OR=3.19 [2.42–4.22]), overtime (≥11 hours daily; OR=5.65 [3.64–8.77]), married (OR=1.79 [1.16–2.75]), parental (OR=1.48 [1.07–2.04]) and late dinner time (OR=1.75 [1.14–2.46]), administrative position (OR=1.56 [1.27–2.42]) (Table 5).
| High (%) | Model 1 OR (95% CI) | Model 2 OR (95% CI) | Model 3 OR (95% CI) | Model 4 OR (95% CI) | Model 5 OR (95% CI) | |
|---|---|---|---|---|---|---|
| Age | ||||||
| 20–29 | 30.4 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| 30–39 | 45.7 | 1.92 (1.40–2.64)*** | 2.63 (1.86–3.72)*** | 1.35 (0.95–1.92) | 1.70 (1.16–2.48)** | 1.68 (1.14–2.46)** |
| 40–49 | 45.4 | 1.90 (1.36–2.65)*** | 2.62 (1.81–3.80)*** | 1.13 (0.75–1.70) | 1.39 (0.90–2.17) | 1.34 (0.85–2.11) |
| 50–59 | 40.1 | 1.53 (1.10–2.15)* | 1.85 (1.25–2.73)** | 0.93 (0.61–1.40) | 0.96 (0.60–1.53) | 0.91 (0.56–1.48) |
| Occupational grade | ||||||
| Nonadministrative | 38.6 | 1.00 | 1.00 | 1.00 | ||
| Administrative | 56.5 | 1.64 (1.09–2.49)* | 1.66 (1.09–2.53)* | 1.56 (1.01–2.40)* | ||
| Working hours | ||||||
| <9 h a day | 27.2 | 1.00 | 1.00 | 1.00 | ||
| 9–11 h a day | 52.1 | 3.13 (2.41–4.08)*** | 3.48 (2.65–4.57)*** | 3.19 (2.42–4.22)*** | ||
| ≥11 h a day | 65.8 | 6.47 (4.35–9.62)*** | 7.36 (4.90–11.05)*** | 5.65 (3.64–8.77)*** | ||
| Marital status | ||||||
| No | 31.2 | 1.00 | 1.00 | 1.00 | ||
| Yes | 46.4 | 1.44 (0.97–2.14) | 1.69 (1.11–2.58)* | 1.79 (1.16–2.75)** | ||
| Living with family | ||||||
| No | 30.5 | 1.00 | 1.00 | 1.00 | ||
| Yes | 41.9 | 1.13 (0.69–1.86) | 1.52 (0.90–2.59) | 1.47 (0.86–2.52) | ||
| Living with parents | ||||||
| No | 43.6 | 1.00 | 1.00 | 1.00 | ||
| Yes | 37.6 | 0.89 (0.67–1.18) | 0.83 (0.61–1.12) | 0.86 (0.63–1.17) | ||
| Have children | ||||||
| No | 33.5 | 1.00 | 1.00 | 1.00 | ||
| Yes | 48.3 | 1.43 (1.10–1.92)* | 1.38 (1.01–1.89)* | 1.48 (1.07–2.04)* | ||
| Dinner time | ||||||
| Before 21:00 | 35.2 | 1.00 | ||||
| 21:00 and after | 57.9 | 1.75 (1.27–2.42)*** | ||||
| Bedtime | ||||||
| Before 23:30 | 38.9 | 1.00 | ||||
| 23:30 and after | 41.8 | 0.95 (0.70–1.29) | ||||
| Waking time | ||||||
| Before 6:30 | 42.1 | 1.00 | ||||
| 6:30 and after | 37.6 | 1.14 (0.85–1.54) | ||||
| Sleeping duration | ||||||
| ≥7 h a day | 35.3 | 1.00 | ||||
| <7 h a day | 43.3 | 1.19 (0.87–1.63) | ||||
| Smoking status | ||||||
| Never/former | 40.5 | 1.00 | ||||
| Current | 42.2 | 1.25 (0.64–2.44) | ||||
| Body mass index | ||||||
| <18.5 | 35.7 | 0.87 (0.60–1.25) | ||||
| 18.5–24.9 | 40.5 | 1.00 | ||||
| ≥25.0 | 49.1 | 1.48 (0.96–2.28) |
Abbreviations: OR, odds ratio; 95% CI, 95% confidence interval; high, prevalence of high conflict. Model 1 is adjusted for age. Model 2 is adjusted for age and work characteristics. Model 3 is adjusted for age and family characteristics. Model 4 is adjusted for age, work characteristics and family characteristics. Model 5 is adjusted for age, work characteristics, family characteristics and lifestyle factors. Hosmer-Lemeshow Goodness-of-Fit Test: 0.545.
Women showed higher percentages of both types of conflict than men did in almost all of the variables.
Concerning F-to-W conflict, common contributing factors of both sexes included participant age in the 30s or 40s, longer working hours, having children and shorter sleep duration. Although people in their 30s and 40s are in the prime of their working lives and typically busy raising children, the WLB of these groups was unsatisfactory. Regarding the ORs for the 30s, 40s and 50s age groups in both sexes, F-to-W conflict significantly decreased after adjusting for family characteristics; however, some sex differences were noted. In particular, the OR for women in their 50s was attenuated, and the statistical significance disappeared. This result suggests that family characteristics, such as having a spouse and/or a child or children, have greater effects on younger generations than older generations. In women, the relationship between age and F-to-W conflict is partly explained by family characteristics. Positive impacts of having children on F-to-W conflict were observed in both sexes. In a previous study16), being married with children was associated with lower mental functioning and higher F-to-W and W-to-F conflict levels than being married without children in both sexes. It is necessary for working parents to use family-friendly programs, which are more readily accessible in their workplaces, according to family structure. In this study, shorter sleep duration negatively affected F-to-W conflict, and it can deteriorate sleep quality and increase fatigue. This finding is consistent with results of a previous study15) showing that long working hours and high F-to-W conflict levels led to sleep quality deterioration.
The correlation between F-to-W conflict and occupational grade in men was distinctive. Working in an administrative position had negative impacts on F-to-W conflict in men. Sekine et al.29) reported that holding a high-ranking position was associated with better physical and mental functioning, and Martikainen et al.24) reported that beneficial effects of higher rank on physical functioning are observed in men. The correlation between F-to-W conflict and marital status in women was distinctive, indicating that roles and responsibilities in married life may impose physical and mental burdens on female workers. Chandola et al.16) reported that relative to men, married women with children in Helsinki, London, and Japan showed higher W-to-F conflict levels.
Common factors correlating with W-to-F conflict in both sexes were being in the 30s age group, longer working hours and eating dinner late. These workers, who are typically busy balancing childcare and work, exhibited high conflict levels. The relationship between age and W-to-F conflict in women may be explained by family characteristics. Similar to the ORs seen for F-to-W conflict, those for the 30s, 40s and 50s age groups decreased significantly after adjusting for family characteristics. The OR for the 50s age group was lower than for the 30s and 40s age groups because family burdens and responsibilities decrease and family structures change with age. In men, no significant relationship between having children and W-to-F conflict was observed, indicating that their gender and family roles might have been unchanged. Longer working hours had a negative effect on family life in both sexes. Due to working overtime, the available time of employees for domestic duties and family communication is limited, which necessitates leaving domestic duties unfinished or family problems unsolved. If this occurs frequently or daily, F-to-W and W-to-F conflict may trigger a negative chain reaction, leading to the deterioration of physical and mental well-being, which is consistent with previous findings30). Japanese employees generally work long hours, and Japan ranks 20th out of 35 OECD countries in this regard31). In an international comparative study conducted in England, Finland and Japan, the percentage of women working overtime was highest in Japan21, 30). Regarding preventive medicine and mental health32–35), it is necessary for Japanese workers, who have difficulty following their bodies' natural rhythms because of long working hours and domestic roles, to balance their work and private lives. Eating dinner late has negative effects on employees and their families. Gaina et al.36) reported a relationship between parents' employment and children's dietary disorders or BMI. Eating meals together may enhance parent-child relationships, particularly communication and cohesion37). Eating dinner late was associated with high W-to-F conflict levels in both sexes. Eating dinner earlier is necessary for employees to be able to communicate with their families while eating. Regarding family health, the relationship between dinner-to-bed time and BMI38) is notable. The time parents arrive home after work directly affects the family's dinner and bed times39, 40). To decrease the number of working hours, shortening the weekly working hours to <40 is effective when implemented alongside checks undertaken by industrial physicians and/or nurses at annual health checkups, and administration of working hours using computer systems is effective.
In men, the relationships between W-to-F conflict and living with family and sleeping less were distinctive, indicating that living alone is less stressful than living with family, where men are expected to put family first. Longer working hours and eating dinner late may encourage an irregular lifestyle and negatively affect sleep. Relationships between W-to-F conflict and sleep quality deterioration and longer working hours have been reported15). In women, associations between W-to-F conflict and administrative position and being married with children were distinctive, and holding an administrative position had negative effects on W-to-F conflict, in contrast to the effects observed in men. Martikainen et al.24) reported that working in higher-grade occupations positively affected physical functioning in civil servants in Helsinki, London and Japan, except in Japanese women, who did not show a similar tendency. Our findings indicated that Japanese women in administrative positions had higher W-to-F conflict levels. For women, working in higher-grade occupations may impose physical and mental burdens; therefore, further political inducement is required to increase the number of female managers3) and reorganize workplace cultures. Being married and having children negatively affected W-to-F conflict in women. This result may explain the problem of the M-shaped female labor participation rate3) with a dip in their 30s. In this study, the 20s age group was set as a reference for OR comparisons for each age group; therefore, the OR ranges were narrower in women than in men; women in their 20s displayed a higher prevalence of F-to-W and W-to-F conflict than men did. Higher F-to-W and W-to-F conflict levels in their 20s and 30s may lead young women to delay marriage and cease working following marriage. Improvements in M-shaped labor participation ratios will also increase the number of female managers.
There were several methodological limitations to this study. First, the study design was cross-sectional study. Although the results identified factors contributing to W-F conflict, it is difficult to confirm causal relationships between F-W conflict and factors such as age, work characteristics, family characteristics and lifestyle. It is important to expand the WLB concept and plan a longitudinal study to compare conflict levels before and after a WLB system is implemented. Second, this study included civil servants in a local government; therefore, the results cannot be generalized to the entire Japanese working population, and there is a possibility of selection bias with respect to the region and occupational categories. In addition, the ratio of female employees to the total number of employees, 42.7% in 20114), was higher than that in this study (35.1%). However, according to the second national survey of families6), conducted in 1998, a higher percentage (50.4%) of public sector workers continue to work after giving birth to their first child compared with private sector workers. In large enterprises with more than 1,000 employees, just 13.7% of workers continue to work after giving birth to their first child. Based on these data, workers in the public sector from every age group are worthy subjects for investigation of WLB and W-F conflict in Japan. Furthermore, public sector organizations, which typically have conservative organizational cultures, reflect the conservative welfare state. Therefore, our results are useful for policy makers and managers accelerating WLB policies in countries with conservative welfare state regimes.
In conclusion, the findings of the present study indicate that to accelerate WLB policy introduction and reduce W-F conflict, political inducements for reducing long working hours, reinforcement of gender- and age group-specific family-friendly policies, more flexible and diverse work environments, awareness of WLB and establishment of comprehensive support for employees and their families are required. To establish WLB policies and systems matched to the conservative welfare state, sustainable role models for introduction of the concept of WLB, formed through public sector social and political experimentation, may be useful for organizations with conservative cultures and local communities.
Acknowledgments: We are indebted to all the civil servants in the local government who participated in this study. This study was supported by grants from the Ministry of Health, Labour and Welfare, the Japanese Society for the Promotion of Science, the Universe Foundation (98.04.017), the Daiwa Anglo-Japanese Foundation (03/2059) and the Great Britain Sasakawa Foundation (2551). Funding organizations were not involved in the design, conduct, interpretation or analysis of the study and did not review or approve the manuscript.
F-to-W conflict was investigated with respect to the level of interference from family to work with the following four items: (1) family matters reduce the time you can devote to your job, (2) family worries or problems distract you from your work, (3) family activities stop you from getting the amount of sleep you need to do your job well and (4) family obligations reduce the time you need to relax or be yourself. Each item had three response options, ranging from 1 (never) to 3 (often).
W-to-F conflict was investigated with respect to the level of interference from work to family with the following four items: (1) your job reduces the amount of time you can spend with the family, (2) problems at work make you irritable at home, (3) your work involves a lot of travel away from home and (4) your job takes so much energy you do not feel up to doing things that need attention at home. Each item had three response options, ranging from 1 (never) to 3 (often).
