The aim of this study was to investigate the prospective influence of work-related factors on symptoms of PPD, i.e., whether precarious working conditions, WPC, and ERI ratio increase the risk for PPD. In this prospective cohort study precarious working conditions, WPC, and ERI ratio were individually significantly positively associated with symptoms of PPD within regression models, when controlling for age, professional education, parity, and anxiety during pregnancy. Within a regression including all three predictors, i.e., precarious working conditions, WPC, and the ERI reward scale), all predictors remained significantly associated with PPD. Therefore, precarious working conditions and WPC might act as prospective risk factors for PPD, whereas reward at work might act as a protective factor. These findings should be discussed in the light of previously conducted research. Studies, which found a decreased risk for symptoms of depression and employment have often included the general employment status (e.g. employed vs. unemployed) (66,67) supporting the general benefits of employment for peripartum mental health. However, the present study could differentiate between different work factors that might cause depression.
Precarious working conditions
Precarious working conditions, especiallylow or insufficient wages might act as prospective risk factors of PPD. The effects seem to spill over into the postpartum period of maternity leave. In the present study, the subscale wages showed the strongest association with symptoms of PPD within participants with higher education. The association of wages with symptoms of PPD was surprising as a recent meta-analysis found no association between education or income as a risk factor for PPD (68). On the other hand, some research has indicated that low socioeconomic status is associated with depressive symptoms (69,70), whereas others found no association between income and PPD but could differentiate between the type of job held and the risk for PPD (71). A higher association with vulnerability was expected since its previously found strong relationship with mental health (20).
Psychosocial work stress- work-privacy conflict and effort-reward imbalance
A perceived interference from work into private life might act as a prospective risk factor for symptoms of PPD. Moreover, a perceived imbalance between effort put into and reward received from work might act as a prospective risk factor for symptoms of PPD. The effects seem to spill over into the postpartum period. In another a previous study among hospital employees in Switzerland, both concepts were found to be significantly associated with burnout, but WPC was found to be a stronger predictor for burnout than ERI among health professionals (72). Interestingly, when looking at hospital staff with different levels of professional education, ERI has been found to be more relevant for burnout in tertiary-educated staff. This indicates that in higher educated, professional groups such as therapists, physicians, and medical-technical staff effort-reward imbalance seems to be important for mental health (72). This could be due to a different perspective on the job such as a higher personal commitment and will to work after hours. Similarly, higher educated professionals generally show higher scores for the subscale ERI effort (51) and therefore might be particularly susceptible to burnout or other negative health outcomes. Hence, future analysis should classify between different kinds of professional groups within the peripartum population to investigate different influences of WPC and ERI on PPD. Moreover, among physicians in a hospital setting, Hämmig (35) could show that exposure to stress and the outcomes burnout and intention to leave the profession were partly or largely mediated by WPC and ERI. Whereas WPC predicted burnout symptoms, the ERI ratio showed the strongest effect for predicting thoughts of leaving the profession, indicating that WPC and ERI measure different aspects of psychosocial work stress and therefore might predict different outcomes (35). More research is required to draw better conclusions and to develop a more integrated model of psychosocial work stress, its biological foundations, and their interaction with the individual.
Whereas the ERI ratio did not remain a significant predictor of PPD symptoms in a regression with WPC and precarious working conditions (see Table 4), the ERI subscale reward did proof to be a significant predictor of PPD symptoms. The reward subscale had a stronger association with the outcome in the peripartum period. This has also been found in another study, where reward, rather than effort, was found to be positively associated with gestational age (38). A recent review has also claimed that a “crisis of gratification”, where expected and legitimate reward is not being experienced, is very relevant to depressive disorders. Seven epidemiological studies concerning the ERI ratio and depression revealed a two-fold elevated relative risk of incident depressive disorder in Europe with findings pointing towards an altered immune function and inflammatory processes (73).
Concerning the peripartum period. recent findings indicate that ERI scores can fluctuate during the prepartum period (39,40). The factor scores for effort and reward seem to decline during the course of pregnancy, indicating that women put less effort into work and simultaneously receive less reward from it. Reward might be especially relevant for expectant mothers, as Siegrist & Li (37) found that aspects of low reward at work seem to be particularly important for biomarkers and therefore physical correlates of health. Reward at work should thus be investigated to a greater extent in pregnant populations. It needs to be noted that a steady ERI ratio was not shown for all participants in those previous studies (39,40). Therefore, it seems important to consider individual or systematic differences in the trajectory of the ERI.
Strengths and limitations
This study is the first to include multiple theoretical approaches of work stress within a representative sample of women in the peripartum period. Moreover, no studies were found that previously applied the WPC and the EPRES to the peripartum period. The present investigation is part of a large prospective-longitudinal cohort study covering many fields of interest regarding employment, mental health, and associated factors (42). Therefore, it will be possible to incorporate further theoretical concepts and factors in future analyses. Examples include paternal mental health as well as hair cortisol levels of fathers, mothers, and their offspring as biological indicators of stress within the DREAM-study. DREAM aims to further investigate the impact of work-related factors on the well-being of all family members.
However, some possible limitations need to be noted. The applications of findings in this study to the general population might be limited as the sample is highly educated in comparison to the general female population of Dresden. However, previous research has shown that self-selection according to sociodemographic variables such as education had little impact on prevalence estimates (74). Concerning the predictors, the ERI ratio itself might be modified by the peripartum period due to changing perspectives on work (40). Changes of the ERI score could not be observed in this study, as there was no measurement point before the pregnancy. Within the EPRES, the subscale rights, concerning workplace rights, showed a low internal consistency contrary to the findings of the authors (20,21). This could be due to the answer categories: “yes”, “no”, “don’t know”. Most participants of the present study were in an employment situation with the workplace right “paid maternity leave”. While most participants were aware of their right to receive paid maternity leave, most were unaware whether they would receive a dismissal wage. When recoding the answer “don’t know” into “missing”, the internal consistency of the scale rose to Cronbach’s alpha = .67. A further limitation is that dropout analyses showed a slightly higher EPRES score for completers vs. non-completers. Women with more precarious working conditions might have been particularly motivated to take part in the study to promote changes within the workplace for pregnant women. However when comparing predictor scores to other healthy samples, participants in this study might experience less precarious working conditions. This could be due to the highly educated sample making it even more important to screen for precarious employment in a less educated sample.
Implications
Outcomes and consequences of PPD on the affected women, their families, and especially their children are considered to be severe. Longitutinal research including risk and protective factors is necessary to provide the best basis for effective treatment approaches. The present research indicated that work-related factors need to be considered when screening for PPD or during treatment of PPD. Future analyses should include more confounding variables, such as social support, recurrent depression and paternal mental health, especially within the context of a biopsychosocial model of PPD (3).The DREAM study will be able to include more concepts and possible covariates, as well as hair cortisol concentrations, in future analysis to present a more comprehensive view on peripartum health.
WPC should be lowered, especially considering the changing private situation of women during the perinatal period. The finding that WPC remained significantly associated with PPD when controlling for other psychosocial stress factors at work highlights the importance of introducing strict work guidelines to prevent employees from working at home after working hours. Previous research suggests that various family-specific support systems, such as family friendly organizational policies and climate can reduce WPC (75). A growing body of research has investigated family-supportive supervisors, i.e., supervisors at work who promote the management of work and non-work responsibilities and acknowledge their employees private life (76,77). Facilitating family-supportive supervisors through training might be an effective approach to improve employee work, family, and health outcomes (76). Additionally, mindfulness and self-monitoring trainings (78) as well as cognitive-behavioural interventions such as coaching sessions (79) have recently been investigated to successfully reduce WPC.
Further, reward at work seems to be a protective factor against symptoms of PPD for expecting mothers. Women might experience less appreciation and reward at work in their peripartum period due to changing physical capacity or the upcoming period of maternal and shifting priorities away from the job. Valuing the efforts spent at work during late pregnancy might protect mothers from PPD symptoms and additionally prepare them for a better re-entry into employment after maternity leave.
Some factors of precarious employment, especially defenselessness to authoritarian treatment and low or insufficient wages, seem to increase the risk for PPD. Workplace policies could be implemented to reduce this risk. Julià, Vives, Tarafa, & Benach (23) suggested a surveillance system to monitor different precarious employment dimensions and identify populations at risk to reduce mental health impact. Moreover, the high association of wages with PPD in comparison to the other predictors in this study raises the discussion of equal payment for women in relation to men