This was a cross-sectional prevalence study. Data were collected between July and December of 2013. The research participants were married women who were interviewed 4 weeks to 6 months after giving birth.

The sample size was calculated using the formula “estimating a proportion of the population” (10): n=Z2(1−α/2)p(1−p)d2=355, in which –

Z = value from standard normal distribution corresponding to desired confidence level [Z(1 − α/2) = 1.96 for 95% CI].

A two-stage cluster sampling method was applied [probability proportional to size (PPS)]. The sample size was multiplied by the design effect (DEFF). We chose a DEFF = 1.5, which indicated that n would be 355 × 1.5 = 532. Consequently, 600 research participants were selected.

The PPS sampling and simple random sampling (SRS) were conducted as follows:

Step 1: All wards in Hai Chau District were numbered. A cumulative frequency table was generated, which produced the total number of married women giving birth from 4 weeks to 6 months in Hai Chau District, identified as m (m = 1,075).

The sampling procedure was summarized in Figure 1 as follows.

A questionnaire was used to directly interview women from the 10 selected wards of Hai Chau District from July to December 2013. The questionnaire included questions about the respondents’ demographics as follows: –

Age, employment, working conditions, education, and whether their family experienced financial difficulties (defined as earning <VND 9,600,000/person/year);

The questionnaire was developed by the research team and was validated by a pilot survey. PPD screening was conducted using the Edinburgh Postpartum Depression Scale (EPDS) translated into Vietnamese. The EPDS has been widely used to identify symptoms of PPD and has been verified with a sensitivity of 88% and specificity of 92.5% (11). Most of the studies about PPD conducted in Vietnam apply the EPDS (4). The EPDS questionnaire has 10 items. The mother chooses one of four possible answers to describe what they experienced in the previous week. Each answer is converted to a score on a scale of 0–3. The total possible score is 30. A cutoff point of 12/13 was used to define the presence of PPD symptoms (without PPD symptoms: EPDS ≤12 points/with PPD symptoms: EPDS ≥13) (11).

Data were analyzed using the software of SPSS 16.0 Version for Windows. Chi-square tests and multivariate logistic regression using a backward conditional method were undertaken. All variables associated with increased risk of PPD in unadjusted tests were used.

Three principles of research ethics were observed throughout this study: respect for privacy, fairness, and benefit to the participants. Before the interview, research purposes were clearly explained. It was also explained that participation was voluntary. The participants were assured that they could stop at any point, if they so desired, and reassured that all responses would be kept confidential.

The mean age of respondents was 28.76 (±4.99), of whom the youngest was 17 and the oldest was 44. The 20–34 age group made up the highest proportion of respondents (84.8%). In terms of employment, the largest single category was public officials, accounting for 36.7%, while the smallest category was manual workers, which accounted for 13.8% of respondents. Overall, 75% of the mothers reported having stable jobs. The research sample consisted of women of all educational levels. Mothers who had less than a secondary level of education accounted for 17.3% of the respondents, including those who had not attended school and those who attended primary or secondary school only. Those with tertiary education and above made up the highest proportion of respondents, at 82.7%, of which 32.7% were graduates or postgraduates.

The proportion of women with financial difficulties within the family was 13.8%. The percentage of women who were economically dependent on their husband or family was 36.5%. Most women lived together with their husband (98.2%), among which the majority lived with their own parents or parents in law (58.7%).

Women who had undertaken a mental health examination during their previous pregnancies or postnatal period had a higher rate of PPD than women who had not had a mental health examination, but this difference was not statistically significant (p > 0.05).

The prevalence of PPD symptoms based on an EPDS score ≥13 was 19.3% (95% CI 16.16–22.50) (n = 116). Among the 116 women with PPD symptoms, 37.9% reported having suicidal thoughts in the previous 7 days (95% CI: 28.96–46.89).

Table 1 includes all sociodemographic and personal variables of respondents.

Table 2 showed the distribution of factors that were associated with PPD symptoms among the respondents.

Women’s factors associated with PPD: women’s working conditions, their economic dependence on husbands, current place of residence, pregnancy expectation, expected baby gender, their anxiety about other matters in addition to giving birth, reply on their husbands for help, sharing all problems with their husbands, feeling scared of their husbands during 12 months, violence against women by husbands, exercise after giving birth, and infant illness in postnatal period (p < 0.05).

Table 3 continued to show results of multivariate logistic regression analysis to examine factors, after confounding variables were adjusted, associated with PPD based on EPDS.

The prevalence of PPD symptoms within the research population was 19.3%. Other researchers who also used the EPDS and a cutoff point of 13, and conducted their studies on women in various periods within the first year after giving birth, have found differing rates of prevalence. For example, Aydin et al. conducted a study in a Turkish community of women after giving birth from 1 to 12 months and found the PPD prevalence to be 40.1% (4), while research by Pocan et al. (12), also in Turkey, found that the prevalence of PPD in women after giving birth from 4 to 6 weeks was 28.9% (12). In 2010–2011, Murray et al. studied mothers who had given birth from 4 weeks to 6 months in Hue, Vietnam, and found a prevalence of 18.1% (8).

The prevalence found in the study reported here was similar to that found in Murray et al.’s study, and also the findings of a large systematic review of common perinatal mental disorders in low-and-middle-income countries where context-specific as well as gender-sensitive factors considerably contributed to PPD (3). However, it was lower than that found by Fisher et al.’s studies, conducted with mothers after giving birth from 4 to 6 weeks in Ho Chi Minh City in 2004 and in the northern area of Vietnam in 2010, which were 33 and 29.9%, respectively (7, 13). The differences within PPD prevalence may be attributed to many factors. PPD is an illness associated with economic, cultural, and social issues; thus the studies conducted in places with various cultural and social backgrounds will have different rates of PPD. The findings could be influenced by the research design, and the period during which the research was conducted within the first year after giving birth. Moreover, the EPDS as a tool reports prevalence at a point in time rather than incidence of ongoing depression.

Our findings indicated that the number of mothers with unstable jobs with symptoms of PPD was higher than that of those who had stable jobs (26.7 vs 16.9%). In addition, the women who were financially dependent on their husbands or families had a higher rate of PPD (24.2%) than those who were financially independent (16.5%). Vietnam has strong policies promoting gender equality, which means both men and women have opportunities for studying and working. However, if a mother did not have stable employment outside of the home, or unemployed or freelances, she could not make financial contributions to the family and thus becomes economically dependent on her husband. Child rearing is a major task, and mothers must spend a number of months caring for their newborn child. Not having independent earnings or paid maternity leave can therefore increase the financial burden on the family. The results of the study here were similar to the study by Trinh (14) and Fisher et al. (7).

Regarding family living arrangements, mothers who lived in rented houses or who were living with their parents had higher rates of symptoms of PPD (22.0%) than those living in their own houses (13.8%). The study by Ekuklu et al. also concluded that living in a rented house was a factor related to PPD [c.f. (4)]. The rapid urbanization of Danang city in recent years has included the building of many industrial parks and has attracted a large number of laborers from rural areas and other provinces. This mass immigration to Danang for living and working has increased the demand for housing and health services among the local people. As a result, mothers living in rented houses will be subjected to the financial pressures of paying rent and/or be concerned that the house lacks basic amenities, hygiene, and security. These factors may affect the family’s living standards, which may partially explain this association with increased PPD symptoms in mothers who rent.

The results also found that mothers who had an unexpected pregnancy were more likely to have symptoms of PPD than that of those who had an expected pregnancy (27.0% compared with 17.8%). This result was consistent with the findings of Klainin and Arthur (4), which reported that unexpected pregnancy was meaningfully related to PPD in 22 countries from Asia (4). With an unexpected pregnancy, the mother may not have time, support, or resources to prepare spiritually, physically, and economically to have a new baby.

Expectations of infant gender were also associated with symptoms of PPD. In all, 27.8% of the mothers with PPD symptoms did not expect the gender of their baby (primarily, those who hoped for baby boy but gave birth to baby girl). The prevalence of PPD symptoms in these mothers was higher than those with an infant whose gender was expected (17.5%). These results were similar to the study by Klainin and Arthur (4), which synthesized seven studies in Asia and recognized the connection between unexpected gender and PPD (4). In Vietnam, the preference for sons over daughters originates from Confucianism and patriarchal societal norms, where property and family names were inherited through males (8). Similarly, a study by Thanh et al. (15) revealed that 6.2% of women were forced to divorce by their husband’s family because they were unable to deliver a baby boy (15). Murray et al. (8) found that while having an infant that was not the gender that was expected affected maternal wellbeing, it was not significantly associated with postpartum depressive symptoms. The inconsistent findings about this association from a number of studies indicate that more research into the effect of son preference and family planning policies on perinatal mental health is needed in Vietnam.

Our results indicate that women who were anxious about matters other than giving birth had a higher rate of PPD (34.2%) than those without anxiety (12.0%). Siu et al. (16) claimed that stresses in life were meaningfully connected to PPD (RR = 2.56; 95% CI: 1.84–3.57; p < 0.001) (16).

Regarding women’s relationships with their husbands, the analysis showed that women who did not get support from their husband had a higher rate of PPD (44.9%) than those who had support from their husband (16.0%). As shown in the multivariate logistic regression analysis in Table 3, women who were not helped by their husband were twice as likely to have symptoms of PPD compared to those who relied on their husband (OR = 2.26; 95% CI: 1.36–4.81). This result was similar to those found in studies by Fisher et al. (7, 13). After giving birth, a woman needs a lot support from her husband and family. Not having time to share their problems with their husband was a factor related to PPD (OR = 2.45; 95% CI: 1.35–4.45). This finding was consistent with other studies by Trinh (14) who reported that pregnant women who can confide in their husbands will have a reduced risk of being affected by PPD (14).

Thirty-two percent of women who felt afraid of their husband in the first year post-childbirth had symptoms of PPD, which was associated with higher EPDS scores. In addition, there was a difference in the prevalence PPD symptoms among women who were hit by their husbands (38.9%) compared to those who did not experience being hit (18.1%). This result was similar to that of the study by Murray et al. (8) and reveals the close connection between women who were afraid of their husband in the previous year and the average EPDS (p < 0.001) (8). Similarly, the study by Fisher et al. (3) concludes that domestic violence was a social factor that was strongly associated with PPD in low-and-middle-income countries, including Vietnam (3). In Danang, according to the statistics of the Department of Culture, Sports and Tourism in 2014, the victims of domestic violence were mostly women, with the main causes being attitudes related to feudal ideology, family’s financial problems, alcohol abuse, controlling behavior, and a lack of legal knowledge (17). The results of our study were also consistent with those found by Woolhouse et al. (18) and Pocan et al. (12): there was a strong relation between domestic violence and PPD (12, 18).

Physical health during the postpartum period was also relevant to the findings. We found that women who exercised were less likely to have PPD symptoms than those who did not (20.9 vs 5.0%, respectively). Our study concluded that not exercising after giving birth was a factor related to PPD symptoms (OR = 4.86; 95% CI: 1.44–16.39). There was evidence that postnatal exercise can help women recover postpartum, regain strength, and reduce stress. Diem and Nguyen (19) suggest that exercise helps to produce and increase serotonin in the brain, which regulates mood, and will make one feel more relaxed and comfortable (19).

Another important association with symptoms of PPD was having an infant who was sick in the period between giving birth and the time of the interview. Mothers with sick infants were more likely to have PPD symptoms than mothers whose infants were healthy (35.2 vs 13.6%). The multivariate regression analysis also indicates that having a sick newborn was a factor related to PPD (OR = 0.29; 95% CI: 0.18–0.46). Clearly, the health of the infant can have a considerable effect on their mother’s mood. With respect to this factor, the findings were similar to studies by Xuan Đen (Ho Chi Minh City), Glasser (Israel), Manl, KD (USA) (4, 15, 20).

We finally only five factors were significantly associated with PPD symptoms after multivariate logistic regression analysis (p < 0.05). These five factors included: women can rely on their husband for help, women who can share all problems with their husbands, anxious about other matters in addition to giving birth, doing exercise after giving birth and having a sick infant in the postnatal period. In Vietnam, especially in coastal provinces in the central region as Danang city, unequal treatment of individuals based on their gender is a deeply rooted problem. Such attitudes, together with the impacts of modern lifestyles may influence the prevalence of PPD.