Introduction

Depression is a common mental health disorder characterized by persistent sadness, loss of interest, feelings of guilt, and low self-esteem. It ranks among the top three causes of global disease burden and is particularly disabling for women across their life course1,2. One specific form, antenatal depression (AD), occurs during pregnancy and is marked by emotional disturbances such as poor concentration, fatigue, and mood changes lasting more than two weeks. This condition is often underrecognized, despite its significant implications for maternal and child health3.

Pregnancy is often viewed as a time of joy and anticipation. However, many women, especially in resource-limited settings, experience substantial physiological, psychological, and social changes during this period, which may increase their vulnerability to emotional distress4. Global studies report that 15% to 65% of pregnant women may experience symptoms of depression5,6. If left untreated, antenatal depression can affect a woman’s ability to function, strain relationships, disrupt fetal development, reduce antenatal care attendance, and even increase the risk of postnatal depression and infant mortality7,8,9,10,11.

In Ethiopia, maternal mental health remains a neglected area within reproductive health services. Although national and regional efforts have improved maternal care, mental health, particularly during pregnancy, has not been adequately prioritized. Studies conducted across various parts of the country report the prevalence of antenatal depression ranging between 15.2% and 35.4%11,12,13,14,15. However, much of this evidence is concentrated in urban or central settings, leaving many rural areas under-researched.

The Beck Depression Inventory (BDI) captures a broader range of cognitive and somatic symptoms of depression, whereas the Edinburgh Postnatal Depression Scale (EPDS) is designed specifically to emphasize perinatal emotional symptoms. These differences in focus contribute to variations in prevalence estimates reported across studies.

To understand AD in its local context, it is helpful to consider social stress theory and biopsychosocial models, which emphasize the role of socio-demographic, psychological, and environmental stressors in shaping mental health outcomes during pregnancy. These frameworks guide this study in identifying both individual- and community-level risk factors16.

Although several studies in Ethiopia have reported antenatal depression, most are concentrated in urban or central areas. Little is known about rural settings such as East Gojjam Zone, where socio-cultural and economic contexts may uniquely influence maternal mental health.

Methods

Study design, setting, and period

A facility-based cross-sectional study was conducted from February 1 to March 21, 2020, at Shegaw Motta General Hospital, located in Motta town, East Gojjam Zone, Amhara Region, Ethiopia. Motta is situated approximately 371 km northwest of Addis Ababa, the capital city of Ethiopia, and about 120 km southeast of Bahir Dar, the regional capital. The town has one government health center, one general hospital, and five private clinics. According to local estimates, Motta has a total population of 37,504, comprising 18,606 males and 18,898 females. Among the female population, 14,278 are of reproductive age. Shegaw Motta General Hospital provides health services to over 25,000 people and delivers antenatal care (ANC) to more than 400 pregnant women each month.

Population

The source population comprised all pregnant women attending ANC services at Shegaw Motta General Hospital. Women with a self-reported history of depression were included because the study aimed to assess both current and recurrent depressive symptoms. However, pregnant women who were critically ill at the time of data collection were excluded from the study.

Sample size determination

The sample was determined using the first and second objectives.

The required sample size was determined using a single population proportion formula by taking a p-value of 23%17, 95% confidence level, 5% margin of error, and adding a 10% non-response rate.

The formula to determine the sample size was.

Then the sample became 370 after adding a 10% non-response rate.

For the second objective factor17, as shown in the table below, was calculated using Epi Info software version 7 by considering the following assumptions: confidence level of 95%, power of 80%, and exposure to the unexposed ratio of 1 (Table 1).

Table 1 Sample size determination using the second objective for mothers attending ANC follow-up in Shegaw Motta general Hospital, Northwest, Ethiopia, 2020.
Full size table

After adding a 10% non-response rate, the final sample size became 370, and being a housewife was the determining factor of the sample size from the second objective.

Sampling procedure

A systematic random sampling technique was employed to select the study participants. According to the most recent two-month report from the ANC registration book, a total of 830 pregnant women attended ANC services at the facility. The sampling frame included all eligible pregnant women attending ANC during the study period.

To begin, the first participant was selected using the lottery method from among the first three attendees on the initial day of data collection. Subsequently, participants were selected at a regular interval for every third attendee (k = N/n = 830/370 ≈ 2.24, rounded to 3), where N represents the total population and n denotes the required sample size. This process continued in sequence according to the order of arrival for ANC follow-up until the desired sample size was achieved.

Measurements and operational definition

The dependent variable in this study was AD. It was defined as a BDI score ≥ 17, indicating moderate to severe depressive symptoms. Independent variables included socio-demographic characteristics, obstetric factors, substance use, and psychosocial attributes. AD was assessed using the BDI, a widely used standardized screening tool for depressive symptoms. The BDI consists of 21 items, each rated on a 4-point Likert scale, with total scores ranging from 0 to 63. The tool has demonstrated strong internal consistency in various populations, with Cronbach’s alpha ranging from 0.85 to 0.9018. In the current study, the Amharic-translated version of the BDI was pretested, and the reliability yielded a Cronbach’s alpha of 0.87, indicating high internal consistency.

Data collection tools and procedures

Data were collected using a face-to-face interviewer-administered questionnaire that was pretested for clarity and reliability. The BDI was translated into Amharic using a forward and backward translation process to ensure linguistic and conceptual equivalence. The Amharic version was then pilot-tested in a similar population, demonstrating acceptable validity and reliability for this study. The questionnaire also included items on socio-demographic characteristics, obstetric history, substance use, and psychosocial factors. Two diploma-level midwives served as data collectors, supervised by two BSc-level midwives who ensured data quality.

Data quality control

Data quality assurance measures were implemented before, during, and after data collection. The questionnaire was initially developed in English, translated into Amharic, and subsequently back-translated into English to ensure semantic consistency. A pretest was conducted on 19 participants at Bichena Hospital, a facility not included in the main study, to evaluate the questionnaire’s clarity, appropriateness, and reliability. Based on the pretest findings, revisions and adjustments were made to enhance the tool’s validity. Additionally, a one-day training session was provided to data collectors and supervisors, emphasizing a clear understanding of the research objectives, sampling procedures, and the ethical responsibilities of the research team, as guided by the principal investigator.

Data processing and analysis

After coding and verifying for completeness, the data were entered into EpiData version 4.2 and subsequently exported to SPSS version 25 for statistical analysis. Descriptive statistics were used to summarize the independent variables. To examine the association between AD and the explanatory variables, binary logistic regression analysis was conducted. Variables with a p-value ≤ 0.25 in the bivariate analysis were considered for inclusion in the multivariable logistic regression model. We controlled for potential confounders such as socioeconomic status and obstetric history in the multivariable logistic regression model. While this threshold is more liberal than the conventional p < 0.05, it was chosen to reduce the risk of omitting potentially important variables at the exploratory stage, as recommended in certain epidemiological studies. Logistic regression assumptions were carefully checked before analysis. Model fit was evaluated using the Hosmer-Lemeshow goodness-of-fit test (p = 0.17), suggesting an acceptable fit. However, we acknowledge that this test only assesses overall calibration. Future studies may consider additional model evaluation tools such as the Receiver Operating Characteristic curve and Area Under the Curve to better assess model discrimination. Statistical significance in the final model was declared at p < 0.05, and adjusted odds ratios (AORs) with 95% confidence intervals (CIs) were reported.

Ethical consideration

Ethical approval

for this study was obtained from the Institutional Review Board (IRB) of Bahir Dar University, College of Medicine and Health Sciences, per the Declaration of Helsinki. Before data collection, formal permission letters were secured from the relevant health authorities. All participants were thoroughly informed about the purpose, objectives, and potential benefits of the study. Written informed consent was obtained from each participant, and confidentiality of the information provided was strictly maintained throughout the study process.

Results

Socio-demographic characteristics of respondents

Out of the 370 eligible respondents, 363 participated, yielding a response rate of 98.1%. The mean age of the participants was 25 years (SD ± 6.25). The majority, 254 (70%), resided in urban areas, and 109 (30%) were housewives. Nearly all participants, 355 (97.8%), identified as Amhara by ethnicity. Regarding educational status, 77 (21.2%) had attained only primary education. In terms of economic status, 130 participants (35.8%) reported an average monthly income of less than 2,000 Ethiopian Birr (Table 2).

Table 2 Socio-demographic characteristics of the respondents in Shegaw Motta Hospital, North West, Ethiopia, 2020(n = 363).
Full size table

Obstetric characteristics of respondents

Among the respondents, 106 (29.2%) were primigravida. Nearly one-third, 32 (12.5%), reported a history of preterm birth. Additionally, 63 (17.4%) of the women experienced complications during their most recent pregnancy (Table 3).

Table 3 Obstetric characteristics of respondents in Shegaw Motta Hospital, northwest, Ethiopia, 2020(n = 363).
Full size table

Psychosocial and substance use characteristics of respondents

A total of 41 (11.3%) respondents reported a previous history of depression, while 31 (8.5%) indicated a family history of depression. Approximately half of the respondents, 188 (50.8%), reported having social support. Additionally, 69 (19%) of the participants had experienced stressful life events. Substance use was reported by only 12 (3.3%) of the respondents (Table 4).

Table 4 Psycho-social and substance use characteristics of mothers attending ANC follow-up in Shegaw Motta Hospital, northwest, Ethiopia, 2020(n = 363).
Full size table

Prevalence of antenatal depression

The overall prevalence of AD was 22% with 95% CI (17.75–26.32). Out of the total, tiredness 289(79.6%) and loss of energy 268(73.8%) were the most reported symptoms in the antenatal period (Table 5).

Table 5 Prevalence of antenatal depression of mothers attending ANC follow-up in Shegaw Motta Hospital, northwest, Ethiopia, 2020(n = 363).
Full size table

Factors associated with antenatal depression

In the multivariable analysis, marital status [AOR = 3.87, 95% CI: 1.87–7.99], educational status [AOR = 4.44, 95% CI: 1.79–11.00], history of medical problems [AOR = 3.35, 95% CI: 1.44–7.83], previous history of depression [AOR = 2.60, 95% CI: 1.11–6.10], lack of social support [AOR = 3.82, 95% CI: 1.96–7.45], and unplanned pregnancy [AOR = 3.79, 95% CI: 1.91–7.56] were significantly associated with antenatal depression (Table 6).

Table 6 Bivariable and multivariable logistic regression analysis of factors associated with antenatal depression in Motta Hospital, Northwest, Ethiopia, 2020.
Full size table

Discussion

The prevalence of depression in our study was 22%, compared to 24.5% in Northeast Ethiopia12 and 23% in Gondar17, Hawassa (21.5%)19, North India (21%)20, and Brazil (21.1%) [21. However, the observed prevalence was lower than that reported in Maichew (31.1%)21, Adama Hospital (31.2%)22, Saudi Arabia (31.9%)23, and Bangalore (35.7%)5. These discrepancies may be attributed to several contextual factors, including variations in cultural beliefs, socioeconomic status, and healthcare accessibility. For example, in settings like Saudi Arabia and Bangalore, different cultural expectations and limited mental health integration into antenatal services may contribute to higher AD rates. Methodological differences also play a role; unlike our study, which used the BDI, other studies employed tools such as the EPDS, which differ in sensitivity and symptom threshold. Such methodological variations can influence the detection rate and severity classification of depression symptoms.

Conversely, our prevalence was higher than findings from Debre Tabor (11.8%)10, Brazil (14.2%)24, and Washington (9.9%)25. These differences may be influenced by differences in the sociodemographic makeup of the population, awareness and stigma surrounding mental health, or sample size and timing of data collection during pregnancy.

This study identified several significant predictors of antenatal depression, each of which was explained as follows:

Marital status

Single women were more likely to experience antenatal depression than their married counterparts. This finding is consistent with studies from Northeast Ethiopia12, Arbaminch26, and Brazil27. The increased vulnerability may be attributed to stigma associated with premarital pregnancy, leading to social isolation, lack of partner and family support, and economic challenges. In the Ethiopian context, cultural expectations around marriage and motherhood can intensify emotional stress for unmarried pregnant women, further compounding mental health risks.

Chronic medical illness

Women with a history of chronic illness showed significantly higher odds of experiencing depression, showing findings from Nigeria28, Arbaminch26, and Washington25. Chronic illnesses may reduce a woman’s ability to carry out daily activities, lower her self-confidence, and foster a sense of hopelessness, factors that can aggravate depressive symptoms. In rural Ethiopian communities, limited access to ongoing treatment or specialist care may worsen this burden.

Unplanned pregnancy

Consistent with literature from Dubti1, Oman29, and Turkey2, distress, uncertainty, and financial insecurity often accompany unplanned pregnancies, especially in low-income settings where reproductive autonomy may be limited. In traditional Ethiopian society, cultural expectations around marriage, childbearing, and family roles can result in negative community attitudes toward unplanned pregnancies, increasing social pressure and stigma.

Lack of family support

Women reporting low social support were at increased risk of depression, as seen in studies from Hawassa19, Rwanda30, and Bangladesh31. In Ethiopian culture, family is often the primary support system during pregnancy. When this is absent due to migration, marital conflict, or stigma, the resulting emotional and material insecurity can predispose women to mental distress. Support systems provide not only emotional comfort but also practical help in navigating pregnancy-related challenges.

Previous history of depression

Women with a prior diagnosis of depression were more likely to report antenatal depression. Similar associations were found in Debre Tabor10, Tanzania9, and Brazil24. This suggests that biological vulnerability, along with unresolved psychological stressors, can recur during pregnancy. Given the cyclical nature of depression, pregnancy may act as a trigger for relapse.

Educational status

Lack of formal education was significantly associated with antenatal depression, supporting evidence from Nigeria28, Arbaminch32, and Bangladesh31. Lower educational attainment may limit women’s health literacy, economic independence, and coping strategies. In the Ethiopian setting, uneducated women may have reduced access to accurate health information, leading to delayed help-seeking and greater susceptibility to emotional stressors.

Implications and limitations

This study was done by using a cross-sectional design, which prevents the determination of cause-and-effect relationships. As this was a hospital-based study, the findings may not fully represent the prevalence of antenatal depression in the wider community. Moreover, relying on self-reported data may introduce reporting bias, as participants might underreport symptoms due to stigma or cultural perceptions of mental illness. Methodologically, the use of different screening tools like EPDS vs. BDI can impact prevalence estimates. Each tool has unique cutoff thresholds and symptom sensitivity, which may partially explain cross-study differences.

These findings suggest integrating depression screening into routine ANC visits, strengthening community-based psychosocial support networks, and prioritizing vulnerable groups such as women with unplanned pregnancies, single mothers, and those with chronic illness. Training health workers in early identification and referral of antenatal depression should be a key policy priority.

Conclusion

Antenatal depression affected 22% of pregnant women in this study. Factors such as being single, uneducated, having a chronic illness, lacking family support, and experiencing an unplanned pregnancy increased the risk. Integration of depression screening into ANC, coupled with targeted psychosocial interventions, is urgently needed to improve maternal and child health outcomes in Ethiopia.