The present study found that, compared with unemployment, employment was independently related to a lower prevalence of depressive symptoms during pregnancy, regardless of whether the employment was full- or part-time. With respect to job type, holding a professional or technical job and holding a clerical or related occupation were independently inversely associated with the prevalence of depressive symptoms during pregnancy. There were no relationships between household income and education and the prevalence of depressive symptoms during pregnancy after mutual adjustment for employment, household income and education.
In a study of 5404 pregnant US women, employment, income, and education were significantly inversely related to the depression severity score based on the Beck Depression Inventory after mutual adjustment [4
]. A study in Jamaica showed that employment, but not education, was significantly inversely related to the Edinburgh Postnatal Depression Scale score during pregnancy [5
]. Employment and education were significantly inversely associated with depression during pregnancy in a UK study [6
]. The current results regarding the relationship between employment status and depressive symptoms are in agreement with those findings. Our results regarding the null association with household income are consistent with those in a US study which showed no association between household income and depressive symptoms during pregnancy [7
]. In another study in South Africa, however, lower income was significantly positively associated with the prevalence of depressed mood during pregnancy, while no relationships were observed between employment or education and depression [8
]. Similarly, a study in Australia found that income was significantly inversely related to antenatal depression while education was not associated with antenatal depression [9
]. The present results in relation to the absence of an association between education and depressive symptoms are in agreement with those findings. To date, the findings on this topic have clearly been contradictory. A US cross-sectional study of Hispanic women of Caribbean Island heritage showed that household income and education were significantly inversely related to the prevalence of depression in early pregnancy [10
]. In a cross-sectional study in Brazil, higher income and education levels of the expectant couple were independently associated with a lower prevalence of depression during pregnancy [11
]. A significant inverse association was found between income and depression during pregnancy in a study conducted in the US [12
]. In a study in Lithuania, lower education level was significantly related to a higher prevalence of depression at 12 to 16 weeks of pregnancy [13
]. Our results concerning the relationship between income and education and depressive symptoms are at variance with the findings described here. In particular, the current findings are not consistent with those of the above-mentioned systematic review that showed small associations between lower income and lower education and depressive symptoms during pregnancy, although unemployment was not related to depressive symptoms during pregnancy [3
]. The discrepancies among studies may be explained, at least in part, by differences in the study populations and designs, in the socioeconomic assessment methods used, in the definitions of depressive symptoms, and in the confounders considered.
Because we found a non-significant inverse relationship between household income and depressive symptoms during pregnancy, the observed significant positive association between unemployment and depressive symptoms during pregnancy is likely to be attributable to some extent to financial stress. In a meta-analysis of 46 studies, employees with lower levels of job satisfaction tended to have higher levels of depression [17
]. Given the positive associations between holding a professional or technical job or a clerical or related job and job satisfaction, the observed inverse associations of those job types with depressive symptoms might be attributable in part to higher job satisfaction. In this study, however, information on job satisfaction was not available.
Several methodological limitations of our study warrant mention. First, we assessed depressive symptoms using the CES-D scale rather than structured diagnostic interviews. The CES-D includes questions on physical symptoms such as fatigue and physical discomfort, which are also typical complaints of pregnancy; the consequence of this symptom overlap could have been an overestimation of depression. The prevalence of depressive symptoms in this population was, however, lower than that in a representative sample of the Japanese general population: the prevalence of depressive symptoms (CES-D score of ≥16) was 30.7% in 2315 women aged 30–39 years [18
]. Moreover, our study subjects participated in the baseline survey at various points between the 5th and 39th week of pregnancy. Therefore, it is difficult to accurately estimate the incidence and prevalence of depressive symptoms during pregnancy. The possibility of non-differential outcome misclassification might have biased the magnitude of the observed associations toward the null.
Second, the participation rate could not be calculated because the exact number of eligible pregnant women who were provided with the KOMCHS documents and application form by the 423 collaborating obstetric hospitals is not available. Nevertheless, the participation rate must have been fairly low, given that the present study used data from only 970 pregnant women who lived in Fukuoka Prefecture, while, according to the government of Fukuoka Prefecture, the number of childbirths was 46,393 in 2007 and 46,695 in 2008. We were not able to assess the differences between participants and non-participants because information on personal characteristics such as age, socioeconomic status, and history of depressive symptoms was not available for non-participants. Nevertheless, we can presume that our subjects were probably not a representative sample of Japanese women in the general population. For example, educational levels were higher in the current study population than in the general population. According to the 2000 population census of Japan, the proportions of women aged 30 to 34 years in Fukuoka Prefecture with <13 years of education, 13–14 years of education, ≥15 years of education, and unknown were 52.0%, 31.5%, 11.8%, and 4.8%, respectively [19
]. The corresponding figures for the current study were 24.6%, 33.1%, 42.3%, and 0.0%, respectively. The present population therefore might have had a greater awareness regarding health issues than the general population.
Third, although adjustment was made for a variety of potential confounders, residual confounding could not be ruled out. In particular, we could not control for sociocultural factors and personal and family relations.
Finally, the nature of cross-sectional studies prevents conclusions from being drawn about causality.