Migrant women are two to three times at higher risk for postpartum depression than are native women (Dennis et al., 2018; Xiong and Deng, 2020; Zlotnick et al., 2022). They also seem to be at higher risk for postpartum anxiety (Dennis et al., 2017). These important rates of postpartum distress have been associated with various environmental stressors related to the migration experience. The lack of local language proficiency, inappropriate housing, socioeconomic difficulties, cultural differences, separation from family, and social isolation have been identified as major risk factors for postpartum depression and anxiety in migrant women (Fung and Dennis, 2010; Dennis et al., 2017; Xiong and Deng, 2020; Zlotnick et al., 2022). Undocumented (Davoudian, 2012; Jackson et al., 2018) and asylum seeking (Goguikian Ratcliff et al., 2015a; Mestre, 2016) women face cumulative psychosocial difficulties and are at particularly high risk for perinatal depression and anxiety. The absence or uncertainty of a legal status during the perinatal period is a highly significant risk factor.

In Western psychology, parenting and early parental care have been conceptualized in terms of inner parental characteristics such as maternal self-efficacy. Yet, this approach is culturally loaded and embedded in Western individualistic representations of parenting. Parenting practices widely differ from collectivistic to individualistic cultures (Yaman et al., 2010). In collectivistic lifestyle, the presence and help of elder female relatives is of major importance during the postpartum period. In these societies, the mother’s new social status is celebrated and marked with various protective rituals and taboos (Team et al., 2009; Moro et al., 2013). When migrant women coming from non-Western, rather collectivistic societies give birth in Western countries, they are confronted to an individualistic cultural frame, where the baby is often the primary focus of attention, and where maternity is perceived as an individual responsibility of the mother or the couple. The lack of cultural postpartum rituals and family guidance can lead to a sense of maternal incompetence among migrant new mothers (Baubet and Moro, 2013; Goguikian Ratcliff and Diaz-Marchand, 2019). Studies have also shown that lack of social support (Mestre, 2016; Xiong and Deng, 2020) and precarious living conditions (Leahy-Warren et al., 2012) may impinge on maternal self-efficacy.

Previous considerations emphasize the importance of separate screening for depression and anxiety in the postpartum period. Early identification and treatment of these disorders may help to prevent their impact on the mother, the child, and their relationship (Brummelte and Galea, 2016; Dennis et al., 2017). Studies on postpartum anxiety, depression, and maternal self-efficacy among migrant women have been limited. Moreover, existing studies on postpartum distress mostly focused on refugees and forced migration, while undocumented economic migrants have received less attention. Our study focused on voluntary economic migration and did not include asylum seeking women nor refugees. Refugee mothers often report exposure to pre-migration traumatic events, and present high prevalence of Post-Traumatic Stress Disorder (PTSD) (Daou, 2022). In the present study, we chose to focus on legal status and its association with depressive symptoms after childbirth. Moreover, few studies on the subject included a sample of high-income documented migrant women and a control sample of native women.

In Switzerland, nearly two thirds of permanent foreign nationals are documented migrants who come from Western Europe and other Organisation for Economic Co-operation and Development (OECD) countries (Federal Statistical Office, 2017a). Most of them are highly educated, moving to Switzerland for working reasons. They benefit from a long-term residency permit and receive social rights. On the other hand, around 1% of the Swiss population is represented by undocumented migrants, half of whom are women (Morlok et al., 2015; Federal Statistical Office, 2017b). Most of these undocumented migrants come from Latin America, non-EU countries, Asia or North/West Africa (Federal Commission on Migration (FCM), 2024). Most of them work in care services and endure difficult living conditions (Morlok et al., 2015), as they do not have access to basic social services and healthcare.

The present study aimed to compare two groups of economic migrant women of differing legal status (documented vs. undocumented) and socioeconomic conditions with native Swiss women in relation to postpartum depression and anxiety rates, as well as maternal sense of self-efficacy and social support. It further aimed to compare the effects of postpartum depression, anxiety, and social support on maternal self-efficacy among the three groups. In the light of the previous considerations, we hypothesized that, compared with Swiss women, migrant women, in particular undocumented women, would report higher depression and anxiety rates and lower maternal self-efficacy. We also expected that postpartum depression and anxiety would predict low maternal self-efficacy, whereas social support would predict high maternal self-efficacy in all groups.

Simple random sampling was used for the study. Swiss women and documented migrant women were recruited during pregnancy through flyers and in person while attending birth preparation classes at the Maternity Hospital and a collective of independent midwives. Undocumented migrant women were recruited in person during free medical tests performed during pregnancy at the Maternity Hospital. Due to recruitment challenges, particularly in reaching the undocumented group, no formal power analysis was conducted prior to data collection. The sample size was determined by practical constraints, including the limited recruitment timeframe and the availability of participants. Women who agreed to participate were contacted by phone to set up a postpartum interview during the third month after the delivery. Interviews were conducted by the main author between January 2015 and February 2018. This study is part of a larger longitudinal research project on perinatal mental health in migrant women (Sharapova and Goguikian Ratcliff, 2018; Sharapova and Goguikian Ratcliff, 2021).

Migrant women were defined as those who were born outside of Switzerland and who migrated during adulthood. Undocumented migrant women were those who had no legal residential permit in Switzerland, while documented migrant women were those who had a long-term residency permit.

The inclusion criteria were as follows: third month postpartum, aged 18 years and older, and able to respond to questions and to complete questionnaires in English, French, Spanish, or Portuguese. The exclusion criterion was receiving follow-up care for a severe psychiatric disorder.

The sample consisted of 108 women: 25 (23%) undocumented migrants, 42 (39%) documented migrants, and 41 (38%) Swiss natives. The study procedure was approved by the Faculty of Psychology and Educational Sciences Ethical Committee at the University of Geneva. All women provided written consent and were informed that clinical information would be codified and treated anonymously. All participants were informed that they could withdraw their participation at any time without any consequences. Women who presented clinical anxiety and depression scores were oriented to specific psychiatric centers, either those working with basic health insurance, or those receiving undocumented migrants with no insurance.

A 21-item questionnaire was used to obtain sociodemographic information (age, nationality, education level, date of arrival in Geneva, legal status, French proficiency) and psychosocial living conditions (housing, work rate, presence of family members and friends in Geneva).

The Edinburgh Postpartum Depression Scale (EPDS) (Cox et al., 1987) was used to screen for postpartum depression. It is a 10-item self-report instrument largely used in international perinatal research (Cox et al., 2014). Responses are scored 0, 1, 2 and 3 based on the seriousness of the symptom (e.g., “I have felt sad or miserable”). The total score is found by adding together the scores for each of the 10 items. Higher score indicates stronger symptom severity. In the present study, a cutoff of 10 was applied, as recommended by the authors when the scale is used as a unique measure of postpartum depressive symptoms (Cox et al., 1987). Furthermore, Levis et al. (2020) suggested using lower cut-off values than frequently applied cutoff of 13 to avoid false negatives and identify most patients who meet diagnostic criteria. Translated and validated versions of the scale in French (Guedeney and Fermanian, 1998), Spanish (Garcia-Esteve et al., 2003), and Portuguese (Tendais et al., 2014) were used for this study. Cronbach’s alpha for this scale was 0.77.

The State–Trait Anxiety Inventory (STAI) (Spielberger, 1983), particularly the state anxiety subscale, was used to assess anxiety as an emotional state. The subscale consists of 20 items (e.g., “I feel nervous”), and each item is scored on a 4-point scale, with overall scores varying from 20 to 80. Anxiety is considered low up to a score of 45, moderate between 46 and 55, and high when the score is 56 or more. For this study, translated and validated versions of the scale in French (Bruchon-Schweitzer and Paulhan, 1993), Spanish (Buela-Casal et al., 2011), and Portuguese (Tendais et al., 2014) were used. The state anxiety subscale has shown very good internal consistency, with a Cronbach’s alpha of 0.91.

The Short-Form Social Support Questionnaire (Sarason et al., 1987) was used to assess social support. The scale allows to measure two dimensions: the extent of social support, i.e., the number of people available for emotional or practical support (e.g., “With whom can you totally be yourself?”), and the degree of satisfaction with the available support (the quality of support), scored on a 6-point scale (score ranging from 6 to 36 points). Higher score indicates higher social support. For this study, translated and validated versions of the scale in French (Bruchon-Schweitzer et al., 2003), Spanish (Friedman et al., 2018), and Portuguese (Matsukura et al., 2007) were used. Cronbach’s alpha was 0.91 for the extent of social support and 0.89 for the quality of support.

The Maternal Self-Efficacy Scale (MSES) (Teti and Gelfand, 1991) is a frequently used self-report instrument that consists of nine items that assess women’s feelings regarding efficacy in relation to specific domains of infant care, such as interpreting baby’s demands, performing routine tasks, and disengaging from the baby (e.g., “How good do you feel you are at feeding, changing, and bathing your baby?”). The 10th item assesses mothers’ global sense of efficacy in mothering. Ratings on each item ranged from 1 (not good at all) to 4 (very good). Higher score indicates higher sense of maternal self-efficacy. Internal consistency of the scale was 0.71 for this sample.

The scores of all scales were used in their continuous form for the analyses.

The mentioned above study languages (English, French, Spanish, Portuguese) were those spoken by the researchers. Validated versions of all standardized questionnaires (EPDS, STAI, SSQ-SF) were used in each language. As for the MSES, it was translated and adapted into French, Spanish, and Portuguese by bilingual psychologists, and then back translated into English to ensure the accuracy of translations. Items requiring construct equivalence brainstorming were discussed within the research team.

Chi-square tests and t-tests were performed to compare rates for categorical variables and means for continuous variables. We then used ANOVAs to test mean-level differences among the three groups for the EPDS, STAI, and MSES. Correlation analysis was conducted to examine the relationship between postpartum anxiety and depression. Multiple regression analyses were performed for each group to test the impact of postpartum depression, anxiety, social support, and multiparity on maternal self-efficacy. All analyses were conducted with a significance threshold of α = 0.05, two-tailed. In the exploratory data analyses, no problems were noted for the assumptions of normality and homogeneity of variance of the variables of interest. The EPDS, the STAI, and the MSES scores were normally distributed (skewness <1, kurtosis <1). There was no missing data as all questionnaires have been completed in the presence of a researcher. There were two outliers who were included in the analyses as they did not affect the results and assumptions.

All participants ranged in age from 21 to 42 years (M = 32.7, SD = 4.3). The majority of Swiss women were married or in a couple relationship (n = 40, 98%) and had given birth to their first child (n = 27, 66%). For 83% of them (n = 34), this pregnancy had been planned. All of them had completed secondary or tertiary education and the majority were employed (n = 36, 88%) (Table 1).

The profiles of documented migrant women were not significantly different from those of Swiss women in terms of age, education level, social support, or employment rates (Table 1). They were more often married (χ² = 13.42, p < 0.01) and primiparas (χ² = 6.84, p < 0.05). They were less likely than Swiss women to have family members in Geneva (χ² = 55.13, p < 0.001). They had been living in Switzerland for an average of 5 years (SD = 3.9). All of them had a long-term residence permit or a work permit.

Among undocumented migrant women half of them (n = 13, 52%) were multiparous. For most of them (n = 18, 72%), this pregnancy was unintended. Compared to Swiss women, they had a lower educational level (χ² = 57.85, p < 0.001), more often lived in collective housing (χ² = 21.9, p < 0.001), and reported less social support (t = −5.96, p < 0.001) (Table 1). They had been living in Switzerland for an average of 3.8 years (SD = 3.2). All of them worked part-time or full-time in domestic or care services. Although undocumented migrant women reported poorer social support than documented migrant women did, they had more family members in Geneva (t = −3.56, p < 0.001).

Postpartum anxiety significantly correlated with postpartum depression (r = 0.70; p < 0.001) in all groups. Regarding depression rates, 10% (n = 4) of Swiss women and 12% (n = 5) of documented migrant women presented scores equal to or higher than the cutoff point of 10 on the EPDS. Concerning anxiety rates, 12% (n = 5) of Swiss women and 10% (n = 4) of documented migrant women showed scores higher than 45 on the STAI state anxiety subscale. Therefore, contrary to our expectations, documented migrant women did not significantly differ from Swiss women in terms of the mean scores obtained on the EPDS (p = 0.127) and STAI scales (p = 0.923).

Undocumented migrant women presented more postpartum depression (t = 3.95, p < 0.001) and anxiety symptoms (t = 3.39, p < 0.01) than did Swiss women (Table 2). In fact, 44% (n = 11) of them showed clinical scores on the EPDS, and 40% (n = 10) reported high scores on the STAI state anxiety subscale.

Documented migrant women reported high maternal self-efficacy mean scores, similar to those of Swiss women. However, undocumented migrant women reported a significantly higher sense of maternal self-efficacy than did documented migrant women (t = − 2.67, p < 0.01) and Swiss women (t = 2.67, p < 0.01) (Table 2).

In Swiss women, maternal self-efficacy was predicted by postpartum depression (b = −0.44, p < 0.01), but not postpartum anxiety, social support, nor multiparity. A similar pattern was observed in documented migrant women, as maternal self-efficacy was predicted only by postpartum depression (b = −0.27, p < 0.05). In undocumented migrant women, maternal self-efficacy was not predicted by any of the variables (Table 3).

Several limitations of the study need to be considered. First, documented migrant women in our sample were mostly highly educated women with a high socioeconomic status, which limits the generalizability of the results to all documented migrant women in Switzerland. The recruitment of migrant participants in the English-speaking antenatal classes lead to the overrepresentation of women working in international organizations. Furthermore, the choice of language groups (French, Spanish, Portuguese, or English) limited the inclusion of migrant women speaking other languages, such as documented women coming from non-Western societies. However, the languages chosen for this study are those spoken by most undocumented migrants in Switzerland, as the majority of them come from Latin American countries including Brazil. Furthermore, the relatively small sample size was due to the difficulty to recruit and keep migrant women who had no legal status. Factors hindering their participation included the fear of being denounced as undocumented, missing time due to high workload, or language problems.

Second, we used self-report questionnaires (EPDS, STAI) to measure depressive and anxiety symptoms. These are convenient and easy-to-use methods, but they are not appropriate for establishing a diagnosis of depression and anxiety. A psychiatric classification based, for example, on DSM-5 criteria would give a more accurate estimate of the rates of anxiety and depression disorders, but the establishment of diagnoses was beyond the aims of the present study. It is also worth noting that the antecedents of anxiety or depression disorders for the present sample were unknown.

It may be also argued that the self-report questionnaires used in the present study have been conceived in the Western scientific world and may not capture the non-Western patterns of expression of postpartum anxiety and depression symptoms. To reduce measurement errors due to cultural variables, validated versions of the questionnaires were used. In addition, all questionnaires were completed during oral interviews, therefore, in case of misunderstanding or ambiguity, the researcher was able to reformulate problematic items. The researcher’s physical presence could have affected the results by creating response bias, but the preceding clinical interview helped to establish a climate of trust and address anxiety and depressive symptoms before completing the questionnaires.