Comparing the prevalence and the risk profile for antenatal depressive symptoms across cultures

Abstract

Background and aims:

Although several studies have analyzed the risk factors of antenatal and post-partum depression, evidence on the prevalence and the risk profile for antenatal depressive symptoms (ADS) between native-born and different groups of non-native born women living in the same country is scant. The aim of this article is to compare the prevalence and the risk profile for ADS across geographical areas in women recruited from two large hospitals of North-western Italy.

Method:

The presence of ADS was defined as an Edinburgh Post-natal Depression Scale (EPDS) score ≥12 or a Beck Depression Inventory, Short Form (BDI-SF) score ≥9 or the presence of suicidal ideation/behavior. Crude and adjusted odds ratios (ORs) of ADS were calculated using logistic regression models.

Results:

The prevalence of ADS was 12.4% among Italian women and ranged from 11.4% in other European to 44.7% in North-African women. Crude ORs of ADS were OR = 3.3 (95% confidence interval (CI), 1.2–8.8) for Asian, 3.3 (95% CI, 1.9–5.6) for South-American and 5.7 (95% CI, 3.4–9.6) for North-African women. Marital problems, at-risk pregnancy, past psychiatric history, pharmacological treatment, psychological treatment, financial problems, change in residence and number of children were significantly associated with ADS in multivariate analyses, regardless of women’s origin. After adjusting for these variables, the OR of ADS remained significant for South-American and North-African women.

Conclusion:

Our results demonstrate that the risk of ADS varies across geographical areas of origin and is highest among North-African women. The risk factors identified should be assessed in routine obstetric care to inform decisions about interventions to prevent post-partum depression and its consequences on the mothers and the newborns.

Keywords

Antenatal depression risk factors immigrant prevalence pregnancy

Introduction

Female migration is a relatively recent phenomenon in Italy, which reached a significant size between late 1960s and early 1970s. The first immigrant women arrived in Italy mostly to work as housemaid through two main routes, that is, the families of former colonists who had come back to Italy and religious groups acting as mediators.

In 1981, the first census of foreigners residing in Italy (Italian Institute of Statistics, https://www.istat.it/it/popolazione-e-famiglie) reported the presence of 321,000 people, one-third ‘permanent’ and the remaining ‘temporary’. One year later, the first program was put forward for the legalization of migrants without papers, and the first law governing the matter was enforced in 1986 (L. 943 of 30 December 1986), to provide non-EU workers with the same rights envisaged for their Italian counterparts. The late 1980s saw the inflow of women for reasons of family reunification. These were mostly women coming from Arabic-speaking countries (Egyptians, Tunisians, Moroccans).

In the 1990s migration continued to rise, reaching 625,000 people in 1991, and since 1993 (the year when the birth rate become negative) it became the sole contributor to the growth of Italian population. Martelli Law, issued in 1990, was the first attempt to introduce measures to regulate inflows, and legalized around 200,000 foreigners, mostly from North Africa. In 1991, Italy had to face a large migration of people coming from Albania (after the demise of the Communist bloc). According to estimated figures from Caritas, a Catholic charity, in 1996 there were 924,500 foreign people in Italy. The Turco-Napolitano Law of 1998 was a further attempt to regulate inflows, by discouraging illegal immigration and establishing – for the first time in Italy – Centri di Permanenza Temporanea (CPT – Temporary Detention Centers) for foreign nationals ‘subject to deportation measures’. In 2002, by the so-called Bossi-Fini Law included in its provisions the possibility of immediate deportation of illegal migrants by law enforcement authorities (Italian Parliament, Law 6 March 1998, no. 40, ‘Regulations of immigration and rules on the status of a foreigner’, published in the Official Gazette no. 59 of 12 March 1998 – Ordinary Supplement no. 40). In 2001 (ISTAT Census), 1,334,889 foreign nationals were living in Italy, and the nationalities mostly represented in the country were Morocco (180,103 people) and Albania (173,064).

As of 1 January 2010 (ISTAT), there were 4,279,000 foreign people living in Italy, and this increased to 5,394,000 people in January 2011, of whom 5,187,000 coming from countries with high migration pressure. As of 1 January 2015, the foreign population in Italy was estimated by ISMU – Foundation for Initiatives and Studies on Multi-Ethnicity – to be 5.8 million (legal and illegal residents), with 150,000 people (+2.7%) more than those reported in 2014 (5.6 million). This increase can be ascribed to legal residents for two-thirds, and illegal residents for the remaining one-third. They account for 9.5% of Italy’s resident population.

As of 1 January 2016, 2,642,899 (52.6%) of foreign women were living in Italy, with peaks >80% in the percentage of females among Russians and other Eastern Europe countries (www.istat.it). Two categories of migrant women can be identified in Italy:

First, women who came on their own, leaving their families behind, with the aim of finding employment (e.g. Latin-American women). The inflow of these women, mostly employed in domestic work or as caregivers, was fostered by the make-up of Italian labor market and by the adoption in the more economically advanced European countries of policies restricting inflows since the 1960s. The migrant workers have become necessary in the sector of low-skilled services, and domestic work and care is still their most common occupation (IDOS Study Center in collaboration and with the support of Institute of Political Studies ‘S. Pio V’, Roman Observatory on Migrations 12° report, 2017).

Second, women who have come to Italy for family reunification (mostly from Egypt, Morocco, Tunisia, Algeria) and are often discouraged to look for employment for several reasons (large families with small children, their husbands’ opposition, low educational level, language and cultural obstacles). While in their home countries these women replaced their husbands at the head of the family, being responsible for the running of the household, also from a financial point of view, and their children’s education, once they joined their husbands, they found themselves pushed back to the role of housewives, deprived of their power within the family and marginalized, as they cannot speak the language, are not familiar with the local context, and feel unable to cope with their new life circumstances. Over 70% of inactive foreign people are women. They are excluded from the labor market mainly for their difficulty in reconciling employment with the need to care for their children or other dependents. A rather common occurrence within the population of migrant women is that of young women who do not study, work or are in training. In some communities, this occurrence is extremely high: almost 7 out of 10 for Moroccans and Egyptians (ISMU – Foundation for Initiatives and Studies on Multi-Ethnicity, Report 2016, Foreign Immigration in Lombardy, Regional Observatory for Integration and Multi-Ethnicity; CNEL – Consiglio Nazionale dell’Economia e del Lavoro, ‘La famiglia nell’immigrazione: condizioni di vita e culture a confronto’, 2004).

The majority of women migrating on their own are Catholic, while those coming to Italy for family reunification are mostly Muslim (Provincia di Reggio Emilia, Migrant women and family reunions, 2007). The husband’s role as a mediator for his wife in dealing with the outside world, in his establishing a space of intimacy where a home can be built, might seem at first a sheltering factor. The fact that the women are economically dependent on the head of the household allows them to live in a world ‘apart’, even for a long period of time. However, these same factors, which facilitate adaptation and ease the initial feeling of uprooting, may turn into elements of significant vulnerability later on. The women have then very few opportunities to learn, and communicate in, the language of the country of destination, making the outside world incomprehensible and hostile, and community services inaccessible. Differently from other employed migrant women who access the services on their own, the wives of immigrants continue to come to the services always with their husbands. According to many providers working in healthcare and social services, it is precisely the husband’s presence that turns out to be a hindrance, as it increases mistrust, diffidence and the feeling of being intimidated. Communication taking place through the husband proceeds haltingly and women’s answers, usually short, reach providers through several back-and-forth exchanges and result to be even more concise and stark. The husband speaking in his wife’s stead about strictly female matters, concerning women’s bodies, birth, delivery and pregnancy, is a consequence of migration (Douki, Zineb, Nacef, & Halbreich, 2007; European Network Against Racism (ENAR), 2016).

The migratory experience, with its changes, disruptions and re-balancing, is marked by a state of affective loneliness for all the women. This is triggered by a feeling of not belonging, of instability, all attributed to the disintegration of the family of origin, oftentimes an ‘enlarged’ household with a prevailing female presence. Living far away from parents and sisters and having no relations in Italy are the causes of this feeling of ‘emptiness’ (Dennis, Merry, & Gagnon, 2017).

Many studies reported that the lack of social support, that is, the lack of emotional and practical help from the partner and family members, is strongly associated with depressive symptoms both during pregnancy or/and in post-partum period and with negative maternal emotional well-being (Bowen, Baetz, Schwartz, & Balbuena, 2014; Elsenbruch et al., 2007; Koleva, Stuart, O’Hara, & Bowman-Reif, 2011). Stressful events may contribute to higher vulnerability to depression, particularly if they occur in the context of poor social support (Westdahl et al., 2007). These life events include financial difficulties, unemployment, miscarriage and stillbirth (Rahman, Iqbal, & Harrington, 2003; Rubertsson, Wickberg, Gustavsson, & Rådestad, 2005). Agostini et al. (2014), in a recent Italian study, underscored the role of specific life stressors such as death or serious problems with a close friend or relative, unemployment, financial problems and moving or housing difficulties.

Research studies carried out in Canada and Norway (Miszkurka, Goulet, & Zunzunegui, 2012; Shakeel et al., 2015) indicate that immigrant women have significantly higher rates of depressive symptoms during pregnancy compared with native-born women and these symptoms are associated with poorer functional status and more somatic symptoms, lack of social support, more stressful life events, poverty and crowding, marital strain or poorer marital adjustment. A small study carried out in a multiethnic community in Canada found that high levels of depressive symptoms were associated with significantly higher levels of perceived stress, lower levels of perceived social support, higher levels of somatic symptoms and a lower likelihood to be working or going to school (Peer, Soares, Levitan, Streiner, & Steiner, 2013). Giardinelli et al. (2012) reported that the immigrant status confers a higher risk of depression during pregnancy, and Stewart, Gagnon, Wahoush, and Dougherty (2008) reported that ‘newcomers’, including immigrants, asylum seekers and refugees, were significantly more likely than Canadian-born women to score ≥10 on the Edinburgh Post-natal Depression Scale (EPDS) in the post-partum.

A recent study carried out in the United Kingdom showed that, compared with White British women, ethnic minority women were twice as likely to have potentially missed common mental disorders (Prady et al., 2016). Although several studies have analyzed the risk factors of antenatal and post-partum depression, evidence on the prevalence and the risk profile for antenatal depressive symptoms (ADS) between native-born and different groups of non-native born women living in the same country is scant. Non-native people experience higher rates of unemployment and lower income than the native population, in addition to stressors unique to the resettlement process, such as adjustment to the host country’s culture, displacement from social resources and other forms of social and economic strain. In general, time since migration plays an important role in the integration in the host country for all non-native born women and increases the chance of building a supportive social network and using effectively social and healthcare services (Almeida, Caldas, Ayres-De-Campos, Salcedo-Barrientos, & Dias, 2013).

A better understanding of the factors associated with antenatal depression in immigrant women is needed given the growing size of this potentially vulnerable population worldwide and in Europe in particular. As to Italy, the recent report of the Italian National Institute of Statistics (www.istat.it) indicates that currently the immigrant population comprises 8.3% of the total resident population. In 2014, 19% of births in Italy were from non-native women and the mean number of children was 1.97 in non-native women versus 1.39 in native women.

In this study, we carried out a broad assessment of the risk factors of depression in the framework of a larger Italian project aimed to prevent depression in immigrant women during pregnancy and in the post-partum. The specific aim of this study is to compare the prevalence and the risk profile for ADS by geographical area of origin.

Methods

The study population consisted of pregnant women recruited from two large hospitals of North-western Italy from January 2010 to February 2012 during the scheduled routine follow-up visits or when they attended the antenatal classes. In the first case, women were invited to an interview, and in the second case women filled out the study instruments after a group intervention in which they received information about post-partum depression. The assessments included an ad hoc form to collect socio-demographic information, distressing life events, past psychiatric history and past psychological and/or pharmacological treatment and the history of migration (for women of non-Italian origin) and the family history for psychological problems and psychiatric disorders and pregnancy-related variables. In addition, the EPDS, the Beck Depression Inventory, Short Form (BDI-SF) and the Social Provisions Scale (SPS) were administered in women’s mother tongue and/or in the presence of a cultural mediator. All data were collected by trained psychologists.

The EPDS (Cox, Holden, & Sagovsky, 1987) is the most widely used measure of post-partum depression symptoms and is commonly used as a screening tool for prenatal depression symptoms as well (Gaynes et al., 2005). The 10 EPDS items do not directly correspond to Diagnostic and Statistical Manual of Mental Disorders (DSM) criteria. They do not include somatic depressive symptoms (appetite change and fatigue), as well as psychomotor agitation/retardation and reduced concentration. Participants based their ratings on their experiences and feelings over the previous week. Each item is scored on a 4-point Likert scale from 0 to 3 with possible total scores ranging from 0 to 30. A higher score indicates higher reported frequency or severity of symptoms. A systematic review (Kozinszky & Dudas, 2015) has confirmed that the screening accuracy of the EPDS in diagnosing depression during pregnancy is satisfactory and that the EPDS can be recommended for use for this purpose.

The EPDS was validated in Italian and proved to have a good internal consistency (Cronbach α = .747); a sensitivity of .556 and a specificity of .989 were associated with the cut-off score of 11/12 (Benvenuti, Ferrara, Niccolai, Valoriani, & Cox, 1999). The EPDS scale includes some items that assess anxiety symptoms. Thus, high scores may not reflect exclusively depression. To address this limitation, we introduced the BDI in the protocol. The BDI (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) is one of the most widely used self-rating scales for measuring depression. Beck and Steer proposed that this scale could be divided in two subscales: the cognitive-affective (items 1–13) and the somatic-performance (items 14–21). In this study, we used the cognitive-affective subscale alone (the so-called BDI short-form (BDI-SF)) to assess depression. Each answer is scored from 0 to 3. Higher total scores indicate more severe depressive symptoms.

The presence of ADS was defined as a score of 12 or higher on the EPDS and/or a BDI score ≥9, and/or a score >0 on EPDS item 10 or BDI-SF item 7, that assess suicidality. BDI suicidality item 7 has four response options (0 = I don’t have any thoughts of hurting myself, 1 = I think I’d be better off dead, 3 = I have specific suicide plans, 4 = I would kill myself if I had the chance) and in the EPDS suicidality (item 10) is conceptualized as follows: ‘The thought of harming myself has occurred to me’ 3 = yes, quite often, 2 = sometimes, 1 = hardly ever and 0 = never.

Women at risk of depression were referred to the Psychological Unit of the Department of Mental Health of the hospital for counseling and appropriate treatment when needed. In the presence of suicidality, they were referred to a psychiatrist.

The study protocol was approved by the Local Ethics Committee of the Hospital. Data were collected and analyzed in compliance with the Italian ‘Code of conduct and professional practice applying to processing of personal data for statistical and scientific purposes’ enforced by this Authority (http://www.garanteprivacy.it/web/guest/home/docweb/-/docweb-display/docweb/1115480; published in the Official Journal no. 190 of 14 August 2004). All participants gave their consent to participate in this study. Said consent was documented in compliance with Article 9, section 4b of the code mentioned above.

Statistical analysis

Chi-square test was used to compare categorical variables among groups. Parametric one-way analysis of variance (ANOVA) was used to compare continuous variables with a normal distribution (such as age) among groups and non-parametric Kruskal–Wallis ANOVA was used to compare continuous or count variables with a non-normal distribution (years since migration, number of children) among study groups. Following a significant ANOVA test, post hoc pairwise comparisons were conducted using Tamhane test or Mann–Whitney test to take into account the non-homogeneous variance in the study groups.

The correlation between EPDS and BDI-SF was analyzed using Spearman’s correlation coefficient to take into account the skewed distribution of scale scores. Crude odds ratios (ORs) of ADS, defined as exceeding the thresholds of the EPDS or the BDI-SF scales or endorsing the suicidality items of EPDS and BDI-SF, were calculated for each category of non-native born women, using Italian women as reference category. Adjusted ORs were obtained using logistic regression models. Variables examined as potential predictors of depressive symptoms included age, marital status (married/cohabiting vs other), education, paid maternity leave, unemployment, change in work, bereavement, change in residence, marital problems, financial problems, trimester of pregnancy, at-risk pregnancy, planned/unplanned pregnancy, number of children, number of previous pregnancies, a history of psychological problems, a family history of psychiatric disorders, pharmacological treatment of depression, psychotherapy for depression and type of administration of study instruments (interview vs self-report). Variables significantly associated with the risk of ADS at p < .05 in univariate analyses and differentially distributed across the citizenship categories were included in the final multivariate logistic regression model. Logistic regression analysis, including as predictors the same variables listed above and years since migration, was also used to obtain the adjusted ORs of ADS in the migrant groups, using other European countries as the reference group.

Results

Characteristics of participants

Study participants were 830 women, with a mean age of 32.2 years (standard deviation (SD) = 5.1, range = 16–45), recruited at the first (8.4%), second (28.7%) and third (62.9%) trimester of pregnancy. The large majority (71%) were Italian, 70 (8.4%) were from other European countries (mainly Eastern countries), 76 (9.2%) from North Africa, 76 (9.2%) from South America and 19 from Asia or Middle-East. Socio-demographic characteristics, pregnancy-related variables, and past psychiatric history and distressing life events by area of origin are provided in Table 1. Post hoc pairwise comparisons indicated non-native-born women had several differences from Italian women.

Table 1.

Socio-demographic characteristics, life events, history of migration, pregnancy-related variables, personal and family history for psychological problems and psychiatric disorders.

	Italian(n = 589)a	Other European countries(n = 70)b	Asian(n = 19)c	African(n = 76)d	South-American(n = 76)e	Test and significance of difference
Age, mean (SD)	32.9 (4.8)	30.2 (4.2)	30.0 (6.2)	29.3 (5.7)	31.8 (6.0)	F = 12.8, p < .001 d < a
Marital status
Married, n (%)	557 (94.7)	67 (95.7)	17 (89.5)	74 (97.4)	63 (84.0)	χ² = 16.3, p = .003 e < a, d
Educational level, n (%)						χ² = 76.0, p < .001
None	0	0	0	3 (4.0)	0	d > a, b
Primary school	2 (0.3)	0	0	3 (4.0)	3 (4.0)	d, e > a
Middle school	88 (15.0)	13 (18.6)	8 (42.1)	13 (17.3)	17 (22.7)	c > a
High school	281 (47.9)	27 (38.6)	9 (47.4)	21 (28.0)	39 (52.0)	d > a, e
University degree	216 (36.8)	30 (42.9)	2 (10.5)	35 (46.7)	16 (21.3)	d > c
Living condition, n (%)						χ² = 29.2, p = .004
Alone	10 (1.7)	1 (1.4)	0	0	3 (4.0)
With partner and/or children	542 (92.5)	65 (92.9)	14 (73.7)	75 (98.7)	64 (85.3)	d > c
With original family	19 (3.2)	0	2 (10.5)	1 (1.3)	5 (6.7)
Other	15 (2.6)	4 (5.7)	3 (15.8)	0	3 (4.0)	c > a, d
Work, n (%)
Paid maternity leave	488 (83.1)	31 (44.3)	8 (42.1)	14 (18.4)	53 (70.7)	χ² = 178.8, p < .001a > b > d
Unemployed	30 (5.1)	11 (15.7)	2 (10.5)	5 (6.6)	14 (18.7)	χ² = 25.2, p < .001e, b > a
Change in work	35 (6)	6 (8.6)	0	4 (5.3)	10 (13.3)	χ² = 7.7, p = .102
Distressing life events, n (%)
Marital problems	41 (7.0)	7 (10.0)	3 (15.8)	5 (6.6)	17 (22.7)	χ² = 21.8, p < .001e > a
Financial problems	55 (9.4)	12 (17.1)	7 (36.8)	26 (34.2)	24 (32.0)	χ² = 60.3, p < .001c, d, e > a
Bereavement	39 (6.7)	5 (7.1)	2 (10.5)	3 (3.9)	5 (6.7)	χ² = 1.4, p = .846
Change of residence	43 (7.4)	5 (7.1)	2 (10.5)	15 (19.7)	8 (10.7)	χ² = 13.4, p = .01d > a
Migration historyYears since migration, median (range)		7 (1–18)	10 (1–26)	5.5 (0.5–24)	9 (1–19)	K-W = 18.0, p < .001c > b, d; e > d, b
Reasons for migration, n (%)						χ² = 58.3, p < .001
Financial		31 (46.0)	6 (33.3)	11 (14.5)	41 (54.7)	b > a, d
Family reunion		31 (31.3)	10 (55.6)	61 (80.3)	23 (30.7)	c > a, d
Political		0	0	2 (2.6)	0
Other		15 (22.4)	2 (11.1)	2 (2.6)	11 (14.7)	a > c
Pregnancy-related variables, n (%)						χ² = 26.2, p < .01
Planned	400 (69.2)	48 (68.6)	9 (47.4)	49 (64.5)	33 (44.0)	a, b > e
Unplanned	155 (26.8)	20 (28.6)	10 (52.6)	26 (34.2)	37 (49.3)	e > a
Assisted reproduction	23 (4.0)	2 (2.8)	0	1 (1.3)	5 (6.7)
Trimester
First	39 (6.7)	12 (17.1)	3 (15.8)	11 (14.7)	4 (5.3)	χ² = 80.1, p < .001
Second	126 (21.6)	34 (48.6)	10 (52.6)	27 (36.0)	39 (52.0)	a < b, c, e
Third	419 (71.7)	24 (34.3)	6 (31.6)	37 (49.3)	32 (42.7)	a > b, c, d, e
At-risk pregnancy, n (%)	82 (14.0)	7 (10.0)	1 (5.3)	12 (15.8)	11 (14.7)	χ² = 2.3, p = .672
Number of children, median (range)	0 (0–5)	0 (0–2)	1 (0–5)	1 (0–4)	0 (0–4)	K-S = 95.1, p < .001d > a, b; b > a
Number of pregnancies, median (range)	1 (0–6)	1 (0–5)	2 (0–6)	2 (0–5)	2 (0–10)	K-S = 102.8, p < .001b > a
Psychiatric history, n (%)	151 (25.9)	10 (14.3)	5 (26.3)	12 (15.8)	16 (21.3)	χ² = 7.9, p = .093
Past treatment, n (%)						χ² = 29.7, p = .003
Psychological treatment	69 (11.8)	2 (2.9)	0	0	8 (10.7)	a > b, d
Pharmacological treatment	12 (2.1)	0	0	2 (2.6)	1 (1.3)	–
Both	39 (6.7)	1 (1.4)	1 (5.3)	3 (3.9)	0	–
None	464 (79.5)	67 (95.7)	18 (94.7)	71 (93.4)	66 (88.0)	–
Family history of psychiatric disorders, n (%)	149 (25.5)	5 (7.1)	1 (5.3)	3 (3.9)	11 (14.7)	χ² = 33.7, p < .001b, d > a

Specifically, North-African women had more children and previous pregnancies, were more likely to have financial problems and change address, but were less likely to have a family history of psychiatric disorders and to have been treated for psychological problems. South-American women were younger, less likely to be married, more likely to have financial problems and problems with the partner, to have an unplanned pregnancy, to have changed job or to be unemployed.

Women from other European countries were significantly younger, more frequently unemployed, had financial problems, more children and more previous pregnancies. However, they had less frequently a past psychiatric history, previous treatment for psychological problems and a family history of psychiatric disorders. Asian women had a lower educational level, were more likely to live with the original family or with others, had more financial problems and were more likely to have an unplanned pregnancy.

Concerning migration, North-Africans had the most recent migration history (median, 5.5 years), followed by other European, South-American and Asian women. The most common reasons for migration were family reunion (48.7%) and financial reasons (37.7%). These reasons were differentially distributed among the study groups, in particular family reunion was the primary reason for the majority of North-African women (80.3%) while the search of a job was the main reason for South-American women (54.7%).

The mean EPDS score was 6.1 (SD = 4.7, median = 5, interquartile range (IQR) = 2–9) in the overall sample. The boxplots of Figure 1, showing the distribution of EPDS total score across women subgroups, indicated higher median scores in women from Asia, North Africa and South America compared with Italian women. In women from other European countries, the median EPDS was the same as in Italian women. More variability was found among the other migrant groups because of their smaller sample size compared with Italians. Some outliers with very high EPDS scores were found.

Figure 1.

Boxplots showing the distribution of EPDS scores across women’s area of origin. The segment inside the box represents the median, and the bottom and top of the box are the first and third quartiles. Outliers are represented as small circles.

BDI scores were available for a subset of women with complete data (N = 691). Spearman’s correlation coefficient between BDI and EPDS total score was .54, p < .001, suggesting that the two instruments have a moderate correlation and a partially overlapping content. The percentage agreement between the two scales was 88.9% and Cohen’s kappa was .451, denoting moderate agreement. Specifically, of the 691 women assessed with both scales, 574 did not exceed the threshold for depressive symptoms on EPDS or BDI, 40 exceeded it on both BDI and EPDS, 18 on BDI alone and 59 on EPDS alone.

Prevalence and correlates of ADS

The percentage of women with ADS was 12.4% in Italy and ranged from 11.4% in other European to 44.7% in North-African women (Figure 2).

Figure 2.

Prevalence of depressive symptoms according to women’s origin.

Compared with Italian women, the risk of depression was at least threefold in women from geographical areas other than Europe. Specifically, crude ORs (Table 2) of ADS were OR = 3.3 (95% confidence interval (CI) = 1.2–8.8) for Asians, 3.3 (95% CI = 1.9–5.6) for South-Americans and 5.7 (95% CI = 3.4–9.6) for North-Africans, while the OR for other European women did not differ significantly from that of Italian women, OR = 0.91 (95% CI = 0.42–1.98).

Table 2.

Crude and adjusted risk of depressive symptoms by women’s area of origin.

	Crude OR	95% CI		Adjusted OR^a	95% CI
	Crude OR	Lower limit	Upper limit	Adjusted OR^a	Lower limit	Upper limit
Italy^b	1			1
Other European versus Italy	0.912	0.420	1.982	0.787	0.335	1.850
Asia versus Italy	3.262	1.203	8.849	2.169	0.695	6.771
North Africa versus Italy	5.722	3.421	9.570	5.238	2.736	10.026
South America versus Italy	3.262	1.897	5.611	1.935	1.001	3.740
Constant	0.141			0.021

OR: odds ratio; CI: confidence interval.

Significant OR is reported in bold.

Adjusted for change in residence, financial problems, marital problems, number of children, past psychiatric history, past psychological treatment, past pharmacological treatment, at-risk pregnancy and type of interview.

Reference group.

Socio-demographic variables that proved to be significantly associated with ADS and area of origin in univariate analyses were being married, living arrangement, education, change in work, unemployment, change in residence, at-risk pregnancy, trimester of pregnancy (lower risk with increasing trimester), unplanned pregnancy, family history of psychiatric disorders, past psychiatric history, pharmacological treatment, psychological treatment, marital problems, financial problems, number of children (each child conferring a 47.6% increase in risk), number of pregnancies (each pregnancy conferring a 29.2% increase in risk) and type of administration. These variables were included in a multivariate logistic regression model. After a backward stepwise procedure, only change in residence, at-risk pregnancy, past psychiatric history, pharmacological treatment, psychological treatment, financial problems, marital problems and number of children were retained as independent predictors of risk of depressive symptoms. The final risk adjustment model examining the association between citizenship and risk of depressive symptoms included these variables as covariates and in addition age and type of assessment. Adjusted ORs of ADS (Table 3) were no longer significant for Asian (OR = 2.2, 95% CI = 0.7–6.8), but remained significant for North-African (OR = 5.2, 95% CI = 2.7–10.0) and South-American women (OR = 1.9, 95% CI = 1.0–3.7).

Table 3.

Adjusted risk of depressive symptoms by migrant groups.

	Adjusted OR^a	95% CI
	Adjusted OR^a	Lower limit	Upper limit
Other European^b	1
Asia versus other European	2.44	0.61	9.77
North Africa versus other European	6.98	2.64	18.46
South America versus other European	2.82	1.08	7.34
Constant	–2.184

OR: odds ratio; CI: confidence interval.

Significant OR is reported in bold.

Reference group.

In a secondary analysis, we compared the risk of elevated depressive symptoms among migrant women with different citizenship, excluding Italian women. In this model, we included the number of years since migration, in addition to the factors included in the previous model. Results indicated that years since migration and age were unrelated to antenatal risk of depression. Compared with Europeans, North-African women had about a sevenfold risk and South-American women about a threefold risk. In addition, a past history of psychological problems and marital problems conferred each a fourfold risk of antenatal depression (Table 3).

Suicidality

In total, 1.3% of women reported suicidality on EPDS, 2.2% among Italians, 1.4% (1/69) among other European countries, 15.8% (3/19) among Asians, 9.2% among North-Africans and 6.6% (5/76) among South-Americans. Percentages of suicidality on BDI in these groups were, respectively, 0.6% (3/517), 0%, 0%, 5% (3/60) and 5% (3/60). The agreement between the two scales was mild, Cohen’s kappa = .253.

Discussion

The results of this study indicate a very high prevalence of ADS among African and South-American women (44.7% and 31.6%), that is, significantly larger compared with that of Italian women (12.4%). The prevalence of clinically detected depression during pregnancy in the literature ranges from 3.1% to 4.9% for major depressive disorder and from 8.5% to 12.4% if minor depression is included (Banti et al., 2011; Gaynes et al., 2005). A review of 11 validation studies of the Edinburgh Post-partum Depression Scale (EPDS) carried out during pregnancy reported an even larger variability in the point prevalence of depression, depending on the EPDS cut-off, the research criteria used to define depression, the trimester of assessment and women’s culture (Kozinszky & Dudas, 2015; Leigh & Milgrom, 2008). Studies carried out among immigrant women indicate that the prevalence of ADS ranges from 23% to 42% (Miszkurka et al., 2012; Zelkowitz, 2004; Zelkowitz et al., 2008) and the prevalence of high-level depressive symptoms is 17% (Peer et al., 2013). A meta-analysis of 12 studies (Anderson, Hatch, Comacchio, & Howard, 2017) found no evidence for an elevated risk of depression in the antenatal period among migrant women compared with non-migrant women, OR = 0.91 (95% CI = 0.62–1.33). Still, the results of the meta-analysis are strongly affected by the inclusion of seven US studies, which were conducted in minority low-income populations. In this subset of studies, migrant women had a decreased risk of antenatal elevated depression symptoms compared with US-born women, and this was explained by a lower exposure to the stresses of being a minority among foreign-born women.

Our results also indicate an increased likelihood of depressive symptoms in North-African and South-American compared with native-born women even after controlling for the risk factors. The comparison of baseline characteristics of women according to their geographical area of origin reveals in fact diverse socio-demographic characteristics but confirms the presence of risk factors that can be ascribed to the migrant condition in general (Gavin et al., 2011; Ratcliff, Sharapova, Suardi, & Borel, 2015). However, the heterogeneity of our groups and the paucity of studies conducted in this field in specific Arabian and Latin-American ethnicities did not allow us to determine whether the prevalence found in this study reflects the prevalence of perinatal depression in their country of origin.

Our findings confirm available evidence on the high prevalence of ADS among immigrant women compared with native-born women (Miszkurka et al., 2012; Peer et al., 2013; Zelkowitz, 2004; Zelkowitz et al., 2008). Suicidality proved to be uncommon in the sample. We found a mild agreement between the BDI and the EPDS on the presence/absence of suicidality. However, the two instruments conceptualize suicidality in a different way. EPDS has a focus on the frequency on suicidal thoughts, while the BDI makes a distinction between active and passive suicidality.

We identified several risk factors of depressive symptoms during pregnancy. Marital problems proved to be the strongest predictor of depression during pregnancy regardless of cultural origin. Our results are in line with Lancaster review of 159 studies that found an association with medium-large effect size between the lack of support from the partner and ADS (Lancaster et al., 2010). Moreover, our study allowed to identify in multivariate analyses the effect of past psychiatric history and treatment, change in residence, financial problems and number of children.

The demographic and obstetric risk factors of ADS reported in the literature encompass lower education, lower income, lower socioeconomic status, single status and unwanted pregnancy (Gotlib, Whiffen, Mount, Milne, & Cordy, 1989; Lancaster et al., 2010; Leigh & Milgrom, 2008; Rich-Edwards et al., 2006), but evidence regarding age and ethnicity is inconsistent. Some studies highlighted that maternal anxiety (Berle et al., 2005; Lancaster et al., 2010) and a personal history of depression are strongly associated with antenatal depression (Flynn, Walton, Chermack, Cunningham, & Marcus, 2007).

We found that women from other European countries are similar to Italian women in terms of risk factors and prevalence of ADS. On the contrary, South-Americans tend to have less stable relationships and have more unplanned pregnancies. North-African women have a more recent immigration compared to the other groups, move to Italy for a family reunion, but experience social isolation because they are housewives or unemployed and have limited opportunities to learn the Italian language. Asian women are strongly rooted in their ethic community, but have a low socioeconomic level and frequent unplanned pregnancies, conditions that increase their vulnerability to depression. These different profiles suggest that interventions should be targeted to address specific stressors or the context in which women are living.

Prenatal care providers are in a unique position to manage ADS because they have multiple opportunities to assess, treat and follow-up women during a several month span. A Cochrane review identified some psychosocial and psychological interventions that significantly reduce the number of women who develop post-partum depression. Promising interventions include the provision of intensive, professionally-based home visits, telephone-based peer support and interpersonal psychotherapy (Dennis & Dowswell, 2013a, 2013b).

The strengths of our study are the large number of risk factors considered and the definition of ADS based on the combined use of two scales (Zhao et al., 2015). Limitations are the small sample size of Asian women, which prevents from drawing conclusions about this subgroup. A possible underreporting of depressive symptoms may be present in specific cultures (Leong & Lau, 2001). Some authors suggested that, even after a problem is cognitively defined as psychological, willingness to report problems and express them publicly may be low because of feelings of shame associated with psychological difficulties (Leong & Lau, 2001).

In conclusion, this study underscores the need to assess routinely in pregnant women the risk factors of antenatal depression, with the aim of preventing post-partum depression and its consequences for the mothers and the newborns. This is a healthcare priority given the increasing number of newborns from non-native women in Italy. Thus, efforts in this direction should be made taking into account the cultural specificities that may hinder or facilitate access to healthcare services and women’s willingness to disclose their depressive symptoms and undertake treatment when needed.

Footnotes

Acknowledgements

Dr Mario Meroni, Head of Obstetrics and Gynecology Unit of the Hospital Trust Niguarda Ca’ Granda and Prof Anna Maria Marconi, Head of Obstetrics and Gynecology Unit of the Hospital San Paolo and their equipes are gratefully acknowledged for facilitating the conduction of the study.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

References

Agostini

Neri

Salvatori

Dellabartola

Bozicevic

Monti

(2014). Antenatal depressive symptoms associated with specific life events and sources of social support among Italian women. Maternal and Child Health Journal, 19, 1131–1141. doi:10.1007/s10995-014-1613-x

Almeida

L. M.

Caldas

Ayres-De-Campos

Salcedo-Barrientos

Dias

(2013). Maternal healthcare in migrants: A systematic review. Maternal and Child Health Journal, 17, 1346–1354. doi:10.1007/s10995-012-1149-x

Anderson

F. M.

Hatch

S. L.

Comacchio

Howard

L. M.

(2017). Prevalence and risk of mental disorders in the perinatal period among migrant women: A systematic review and meta-analysis. Archives of Women’s Mental Health, 20, 449–462. doi:10.1007/s00737-017-0723-z

Banti

Mauri

Oppo

Borri

Rambelli

Ramacciotti

… Cassano

G. B.

(2011). From the third month of pregnancy to 1 year postpartum: Prevalence, incidence, recurrence, and new onset of depression. Results from the Perinatal Depression-Research & Screening Unit study. Comprehensive Psychiatry, 52, 343–351.

Beck

Ward

Mendelson

Mock

Erbaugh

(1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561–571.

Benvenuti

Ferrara

Niccolai

Valoriani

Cox

J. L.

(1999). The Edinburgh Postnatal Depression Scale: Validation for an Italian sample. Journal of Affective Disorders, 53, 137–141.

Berle

Mykletun

Daltveit

A. K.

Rasmussen

Holsten

Dahl

A. A.

(2005). Neonatal outcomes in offspring of women with anxiety and depression during pregnancy: A linkage study from the Nord-Trøndelag Health Study (HUNT) and Medical Birth Registry of Norway. Archives of Women’s Mental Health, 8, 181–189. doi:10.1007/s00737-005-0090-z

Bowen

Baetz

Schwartz

Balbuena

(2014). Antenatal group therapy improves worry and depression symptoms. The Israel Journal of Psychiatry and Related Sciences, 51, 226–231.

Cox

J. L.

Holden

J. M.

Sagovsky

(1987). Detection of postnatal depression: Development of the 10-item Edinburgh Postnatal Depression Scale. British Journal of Psychiatry, 150, 782–786. doi:10.1192/bjp.150.6.782

10.

Dennis

C. L.

Dowswell

(2013a). Interventions (other than pharmacological, psychosocial or psychological) for treating antenatal depression (Review). Available from http://www.thecochranelibrary.com

11.

Dennis

C. L.

Dowswell

(2013b). Psychosocial and psychological interventions for preventing postpartum depression. Cochrane Database of Systematic Reviews. Available from http://www.thecochranelibrary.com

12.

Dennis

C. L.

Merry

Gagnon

A. J.

(2017). Postpartum depression risk factors among recent refugee, asylum-seeking, non-refugee immigrant, and Canadian-born women: Results from a prospective cohort study. Social Psychiatry and Psychiatric Epidemiology, 52, 411–422. doi:10.1007/s00127-017-1353-5

13.

Douki

Zineb

S. B.

Nacef

Halbreich

(2007). Women’s mental health in the Muslim world: Cultural, religious, and social issues. Journal of Affective Disorders, 102, 177–189.

14.

Elsenbruch

Benson

Rücke

Rose

Dudenhausen

Pincus-Knackstedt

M. K.

… Arck

P. C.

(2007). Social support during pregnancy: Effects on maternal depressive symptoms, smoking and pregnancy outcome. Human Reproduction, 22, 869–877. doi:10.1093/humrep/del432

15.

European Network Against Racism. (2016). Forgotten women: The impact of Islamophobia on Muslim women in Italy. Retrieved from http://www.enar-eu.org/Forgotten-Women-the-impact-of-Islamophobia-on-Muslim-women

16.

Flynn

H. A.

Walton

M. A.

Chermack

S. T.

Cunningham

R. M.

Marcus

S. M.

(2007). Brief detection and co-occurrence of violence, depression and alcohol risk in prenatal care settings. Archives of Women’s Mental Health, 10, 155–161. doi:10.1007/s00737-007-0188-6

17.

Gavin

A. R.

Melville

J. L.

Rue

Guo

Dina

K. T.

Katon

W. J.

(2011). Racial differences in the prevalence of antenatal depression. General Hospital Psychiatry, 33, 87–93. doi:10.1016/j.genhosppsych.2010.11.012

18.

Gaynes

B. N.

Gavin

Meltzer-Brody

Lohr

K. N.

Swinson

Gartlehner

… Miller

W. C.

(2005). Perinatal depression: Prevalence, screening accuracy, and screening outcomes. Evidence Report/Technology Assessment, 119, 1–8.

19.

Giardinelli

Innocenti

Benni

Stefanini

M. C.

Lino

Lunardi

… Faravelli

(2012). Depression and anxiety in perinatal period: Prevalence and risk factors in an Italian sample. Archives of Women’s Mental Health, 15, 21–30. doi:10.1007/s00737-011-0249-8

20.

Gotlib

I. H.

Whiffen

V. E.

Mount

J. H.

Milne

Cordy

N. I.

(1989). Prevalence rates and demographic characteristics associated with depression in pregnancy and the postpartum. Journal of Consulting and Clinical Psychology, 57, 269–274. doi:10.1037//0022-006X.57.2.269

21.

Koleva

Stuart

O’Hara

M. W.

Bowman-Reif

(2011). Risk factors for depressive symptoms during pregnancy. Archives of Women’s Mental Health, 14, 99–105. doi:10.1007/s00737-010-0184-0

22.

Kozinszky

Dudas

R. B.

(2015). Validation studies of the Edinburgh Postnatal Depression Scale for the antenatal period. Journal of Affective Disorders, 176, 95–105. doi:10.1016/j.jad.2015.01.044

23.

Lancaster

C. A.

Gold

K. J.

Flynn

H. A.

Yoo

Marcus

S. M.

Davis

M. M.

(2010). Risk factors for depressive symptoms during pregnancy: A systematic review. American Journal of Obstetrics and Gynecology, 202(1), 5–14. doi:10.1016/j.ajog.2009.09.007

24.

Leigh

Milgrom

(2008). Risk factors for antenatal depression, postnatal depression and parenting stress. BMC Psychiatry, 8, 24. doi:10.1186/1471-244X-8-24

25.

Leong

F. T. L.

Lau

A. S. L.

(2001). Barriers to providing effective mental health services to Asian Americans. Mental Health Services Research, 3, 201–214.

26.

Miszkurka

Goulet

Zunzunegui

M. V.

(2012). Antenatal depressive symptoms among Canadian-born and immigrant women in Quebec: Differential exposure and vulnerability to contextual risk factors. Social Psychiatry and Psychiatric Epidemiology, 47, 1639–1648. doi:10.1007/s00127-011-0469-2

27.

Peer

Soares

C. N.

Levitan

R. D.

Streiner

D. L.

Steiner

(2013). Antenatal depression in a multi-ethnic, community sample of Canadian immigrants: Psychosocial correlates and hypothalamic-pituitary-adrenal axis function. Canadian Journal of Psychiatry, 58, 579–587.

28.

Prady

S. L.

Pickett

K. E.

Petherick

E. S.

Gilbody

Croudace

Mason

… Wright

(2016). Evaluation of ethnic disparities in detection of depression and anxiety in primary care during the maternal period: Combined analysis of routine and cohort data. British Journal of Psychiatry, 208, 453–461. doi:10.1192/bjp.bp.114.158832

29.

Rahman

Iqbal

Harrington

(2003). Life events, social support and depression in childbirth: Perspectives from a rural community in the developing world. Psychological Medicine, 33, 1161–1167. doi:10.1017/S0033291703008286

30.

Ratcliff

B. G.

Sharapova

Suardi

Borel

(2015). Factors associated with antenatal depression and obstetric complications in immigrant women in Geneva. Midwifery, 31(9), 87–18. doi:10.1016/j.midw.2015.04.010

31.

Rich-Edwards

J. W.

Kleinman

Abrams

Harlow

B. L.

McLaughlin

T. J.

Joffe

Gillman

M. W.

(2006). Sociodemographic predictors of antenatal and postpartum depressive symptoms among women in a medical group practice. Journal of Epidemiology and Community Health, 60, 221–227. doi:10.1136/jech.2005.039370

32.

Rubertsson

Wickberg

Gustavsson

Rådestad

(2005). Depressive symptoms in early pregnancy, two months and one year postpartum – Prevalence and psychosocial risk factors in a national Swedish sample. Archives of Women’s Mental Health, 8, 97–104. doi:10.1007/s00737-005-0078-8

33.

Shakeel

Eberhard-Gran

Sletner

Slinning

Martinsen

E. W.

Holme

Jenum

A. K.

(2015). A prospective cohort study of depression in pregnancy, prevalence and risk factors in a multi-ethnic population. BMC Pregnancy and Childbirth, 15(1), 5. doi:10.1186/s12884-014-0420-0

34.

Stewart

Gagnon

Wahoush

Dougherty

(2008). Postpartum depression in newcomers. Canadian Journal of Psychiatry, 53, 121–124.

35.

Westdahl

Milan

Magriples

Kershaw

T. S.

Rising

S. S.

Ickovics

J. R.

(2007). Social support and social conflict as predictors of prenatal depression. Obstetrics and Gynecology, 110, 134–140. doi:10.1097/01.AOG.0000265352.61822.1b

36.

Zelkowitz

(2004). Factors associated with depression in pregnant immigrant women. Transcultural Psychiatry, 41, 445–464. doi:10.1177/1363461504047929

37.

Zelkowitz

Saucier

J. F.

Wang

Katofsky

Valenzuela

Westreich

(2008). Stability and change in depressive symptoms from pregnancy to two months postpartum in childbearing immigrant women. Archives of Women’s Mental Health, 11, 1–11. doi:10.1007/s00737-008-0219-y

38.

Zhao

Kane

Wang

Shen

Luo

Shi

(2015). Combined use of the postpartum depression screening scale (PDSS) and Edinburgh postnatal depression scale (EPDS) to identify antenatal depression among Chinese pregnant women with obstetric complications. Psychiatry Research, 226, 113–119. doi:10.1016/j.psychres.2014.12.016