Factors associated with stillbirth in four selected hospital maternity units: an unmatched case-control study in Eswatini Authors: Phetsile Nolungelo Thwala, Abednego Ongeso, Rose Maina, Benard D. Mutwiri Published: 2025 ABSTRACT BackgroundStillbirth is a substantially under-recognized adverse pregnancy outcome that predominantly occurs in low-middle-income countries. In 2019, Eswatini’s stillbirth rate was 13.2 per 1,000 births, higher than global targets. However, there is limited research on stillbirths in Eswatini. This study analyzed factors associated with stillbirths among women giving birth in four selected hospitals.MethodsA multi-center unmatched case-control study was conducted using secondary data sources from July 1 to December 31, 2021. Birth records of 268 stillbirths (cases) and 1,151 live births (controls) were selected using consecutive and systematic random sampling, respectively. A piloted data extraction tool was used to extract data. Logistic regression (p < .05, 95% CI) was used to estimate crude and adjusted odds ratios for factors associated with stillbirths, with the final model developed through backward selection.ResultsFactors highly significant and strongly associated with stillbirths (p ˂0.001) included lack of iron and folic acid supplementation (aOR = 2.32; CI = 1.50–3.5), positive rapid plasma regain test (aOR = 7.30; CI = 2.39–22.29), hypertensive disorders of pregnancy (aOR = 3.49; CI = 1.99–6.09), antepartum hemorrhage (aOR = 17.04; CI = 4.53–64.09) birth before arrival (aOR = 1.87; CI = 1.75–2.51), meconium-stained liquor grade II (aOR = 11.42; CI = 4.30-30.35), and fetal complications (aOR = 3.17; CI = 1.99–5.11). Other significant factors include alcohol consumption, having eight or more antenatal care visits, abnormal amniotic fluid volume, use of traditional remedies, gestational diabetes, and anemia.ConclusionStillbirth remains a significant public health burden in Eswatini, with key associated factors that are preventable, highlighting critical gaps in antenatal care. Therefore, strengthening routine screening, and integrated maternal health services offers a cost-effective strategy to reduce stillbirth rates and improve outcomes. Background Perinatal mortality (PM), comprising stillbirths (death of a fetus during pregnancy or labour after viability) and early neonatal deaths (deaths within the first seven days after birth) per 1,000 live births, remains a significant global public burden [1, 2]. In 2019, an estimated 4.3 million perinatal deaths (PD) occurred globally at a rate of 34.7 per 1,000 births [1], of which approximately 2 million were stillbirths, representing a stillbirth rate (SBR) of 13.9 per 1,000 births, and accounting for about 40% of all PD. Similarly, in 2023, stillbirths remained high at approximately 1.9 million with an SBR of 14.3 per 1,000 births [3]. This highlights that stillbirth remains a significant public health issue reflecting persistent gaps in maternal and newborn care and the urgent need for evidence-based interventions. The burden of stillbirth varies significantly between regions and countries. In 2019, the burden of stillbirths was disproportionately higher in low- and middle-income countries (LMICs), contributing about 80% of global cases, with an SBR of 22.7 per 1,000 births, compared to 3 per 1,000 in high-income countries (HICs) [4]. Disparities also exist regionally. For instance, West and Central Africa had a higher SBR of 22.8 per 1000 births, compared to Western Europe’s at 2.9 per 1000 births in 2019 [5]. Furthermore, Sub-Saharan Africa (SSA) had an SBR of 21 per 1000 births [6]. In the same year, Eswatini, an LMIC, reported an SBR of 13.2 per 1,000 births [7], exceeding global targets. The country’s Maternal and Perinatal Death Surveillance and Response (MPDSR) data further indicated that stillbirths accounted for 58% of the 1,146 PD recorded that year [8]. Stillbirth is a multifactorial outcome influenced by a range of interrelated factors, ranging from socioeconomic, maternal, and fetal to healthcare system factors. Socioeconomic factors positively associated with stillbirth include poverty, low educational levels, and poor access to healthcare services [9]. Maternal factors include smoking, alcohol use, obstetric complications, hypertensive disorders of pregnancy (HDP), and medical complications [10–13]. Additionally, fetal factors such as asphyxia, cord problems, and birth weight also contribute significantly [14–16]. Moreover, documented health system factors include distance to the health facility and health information [17]. In contrast, several protective factors have also been identified such as folic acid supplementation [18], a high number of antenatal care (ANC) visits [19], and timely and appropriate treatment of maternal syphilis [20]. Understanding these diverse factors is essential for designing targeted, evidence-based interventions to prevent stillbirths. Stillbirth, an unfavorable pregnancy outcome has profound implications for the woman, her family, and society, with far-reaching consequences that merit comprehensive attention in maternal and public health discourse. These include a profound loss in the family, psychological outcomes such as low self-esteem [21], postpartum depression [22], higher risk of future pregnancy complications, including increased stillbirth risk, and economic burden [23]. These highlight the multifaceted implications of stillbirth and are a strong indicator of broader systemic, social, and policy gaps, signaling the need for coordinated, and evidence-based interventions to improve maternal and newborn outcomes and prevent and reduce the burden of stillbirth. Stilbirth’s significant implications and persistently high incidence underscore the need for urgent action. Several initiatives have been launched globally to support this effort, including the two key initiatives. First, the Every Newborn Action Plan, endorsed by the World Health Assembly in 2014, aims to reduce stillbirths from 18.4 per 1,000 births in 2015 to 12 or fewer by 2030, with a focus on LMICs [24]. Second, the Global Strategy for Women, Children & Adolescents provides a framework for achieving Sustainable Development Goal (SDG) 3.2, which targets ending preventable deaths of newborns and children under 5 years [25]. Despite the implementation of these initiatives among others, stillbirth remains prevalent in most parts of the world including Eswatini, highlighting the need for reinforced strategies and targeted interventions to address specific associated factors. Addressing underlying factors contributing to stillbirths is crucial for reducing the SBR, as highlighted in various literature [5]. Despite notable advancements in obstetric care, stillbirth continues to persist as an under-recognized public health issue [6] necessitating collective and focused efforts to address and mitigate its occurrence. Despite numerous global studies, a knowledge gap persists in Eswatini, limiting the understanding of contextual factors associated with stillbirth and impeding informed practice, policy, education, and research. Hence, there is a need to conduct research aimed at determining the factors associated with stillbirth in Eswatini, addressing the current knowledge gap, and informing targeted interventions to reduce SBR in the country. Therefore, the purpose of this study was to analyze the factors associated with stillbirth among women giving birth in four hospitals in Eswatini from July to December 2021. Methods Study approach, and design This was a retrospective, unmatched case-control study conducted to analyze factors associated with stillbirth in four selected hospitals of Eswatini. This approach and design were appropriate due to the rarity of stillbirth compared to livebirth and the limited existing research on stillbirth in Eswatini. Study setting The study was conducted in four hospitals, one in each of Eswatini’s regions: Mbabane Government Hospital (MGH) in the Hhohho region, Raleigh Fitkin Memorial Hospital (RFM) in the Manzini region, Hlathikhulu Government Hospital (HGH) in the Shiselweni region, and Good Shepherd Mission Hospital (GSMH) in the Lubombo region. All these are public hospitals, with RFM and GSH being mission hospitals- established and run by faith-based organizations, while MGH and HGH are government hospitals- owned, funded, and operated directly by the Eswatini Ministry of Health (MoH). The inclusion of both mission and government hospitals allowed for the representation of a heterogeneous population and diverse health services, care practices, and resource allocation, ensuring the external validity of the findings. On average, MGH, RFM, HGH, and GSMH record 5,000, 6,000, 3,000, and 2,000 births annually, respectively, nearly twice the total births recorded from 1 July to 31 December 2021. Population The target population was all women of reproductive age in Eswatini while the target population comprised all women who gave birth in the four hospitals from the 1 st of July to the 31 st of December 2021. Inclusion and exclusion criteria Women who gave birth at the selected facilities between 1 July and 31 December 2021 with a gestational age (GA) of 28 weeks or more regardless of the birth weight were eligible for inclusion, stillbirth, and live births as cases and controls, respectively. For either, only records with complete information on key variables (GA, birth outcome, maternal age, gravidity, parity, ANC visits, place of birth, and mode of birth) were included. Records were excluded if they had missing key variables or were multiple gestations. Excluded records were replaced immediately with other eligible records from the sampling frame using the same sampling approach. Data sources and study period Secondary data were extracted from birth records and maternity birth registers of women who gave birth between 1 July and 31 December 2021 and analyzed. The period was selected for its recency and to ensure sufficient time for file organization and accessibility at the facility level. Data extraction was conducted from August to September 2022. Sample size and sampling The sample size was estimated using Fleiss’ formula for unmatched case-control studies, assuming a 1:4 case-control ratio, 95% confidence level (CI), 80% power, and an expected odds ratio of 2. A total of 1,400 files were required, with an additional 10%, yielding a final sample size of 1,540. During the study period, 10,025 births occurred (289 stillbirths and 9,736 live births). Consecutive sampling was used to select cases to maximize sample size as stillbirths are rarer than live births, while systematic sampling with intervals of 8, 9, 6, and 4 was used to select controls at MGH, RFM, HGH, and GSMH, respectively. A total sample of 1,419 files (268 stillbirths and 1,151 live births) was included in the study. Data extraction process and tool The researcher, along with two trained research assistants, served as data extractors in all four settings. Maternity registers were used in each setting to compile a sampling frame of women who gave birth between July 1 st and December 31 st, 2021. Upon physically retrieving the files, they were categorized into cases and controls. Cases were consecutively sampled, while controls were systematically sampled to achieve the calculated sample sizes. Subsequently, the sampled files were reviewed, and data were extracted using the predetermined data extraction tool. Data collection occurred from August 4th to September 3rd, 2022. Relevant data were collected from maternity files using a purposively designed piloted data extraction tool consisting of two sections. The first section focused on maternal characteristics, covering five subsections: demographic and socioeconomic factors, behavioral and anthropometric factors, past and present medical factors, past obstetric factors, and present obstetric factors. The second section captured fetal factors, including the number of fetuses, gestational age, fetal state at the onset of labor, fetal presentation, stillbirth type (if applicable), birth weight, sex and congenital malformation, fetal infection, and fetal complications. The data extraction tool was uploaded onto the ODK Collect app, which was employed as a storage platform upon data extraction. Quality assurance measures Two data assistants were trained by the principal investigator over two days. The first day involved physical training on the study protocol, including compiling the sampling frame using birth registers, stratifying records according to birth outcome, applying consecutive sampling for cases and systematic sampling for controls, retrieving physical files, and checking both physical and content completeness. Assistants were also trained to identify key variables such as gestational age, birth outcome, maternal age, gravidity, parity, ANC visits, place of birth, and mode of birth. Dummy maternity files were used for demonstration. The second day involved practical, hands-on training at the pilot facility, where assistants practiced the full data collection process under supervision, two days before the actual pilot study. All necessary modifications were made following the pilot before the tool was used to collect the final data. Throughout the data extraction process, the research assistants were closely supervised by the principal investigator. Study variables The dependent variable (DV) or outcome variable, was the birth outcome, dichotomously categorized as either live birth or stillbirth. The following variables were defined as follows in the study: Live birth: the outcome of a pregnancy where the newborn shows signs of life, such as breathing, heartbeat, voluntary movement, or pulsation of the umbilical cord, regardless of the duration of the pregnancy or the newborn’s weight.Stillbirth: the death of a fetus in utero at any moment during pregnancy at or after 28 weeks gestation, regardless of the birth weight and body length, was adopted. Live birth: the outcome of a pregnancy where the newborn shows signs of life, such as breathing, heartbeat, voluntary movement, or pulsation of the umbilical cord, regardless of the duration of the pregnancy or the newborn’s weight. Live birth: the outcome of a pregnancy where the newborn shows signs of life, such as breathing, heartbeat, voluntary movement, or pulsation of the umbilical cord, regardless of the duration of the pregnancy or the newborn’s weight. Stillbirth: the death of a fetus in utero at any moment during pregnancy at or after 28 weeks gestation, regardless of the birth weight and body length, was adopted. Stillbirth: the death of a fetus in utero at any moment during pregnancy at or after 28 weeks gestation, regardless of the birth weight and body length, was adopted. The independent variables (IV) were broadly classified into maternal and fetal factors. The maternal factors were further divided into five sections as described under the study instrument section. Data processing and management The data extracted were downloaded from the ODK website, pooled into an Excel file, and exported to SPSS version 25. Data completeness was assessed, and missing data were handled using multiple imputations in SPSS where five imputed datasets were created including all analysis and auxiliary variables, as missingness at random (MAR) was assumed. The following variables had more than 50% missing data and were excluded in the final analysis: nutritional status, BMI, blood group, rhesus factor, interpregnancy interval, STI status, and STI/RPR treatment status. Data analysis Descriptive analyses included calculations of frequency and percentages and mean and standard deviation for categorical and continuous variables, respectively. Chi-square or Fisher’s exact test was used to compare differences between cases and controls. Subsequently, a binary logistic regression model was used to assess the factors associated with stillbirth, reported in terms of adjusted odds ratio (aOR) with a 95% significance level. Initially, bivariable analysis was conducted and variables with p-values < 0.025 were pulled into a multivariable analysis to estimate the relative odds of stillbirth while controlling for covariates using backward stepwise selection. A confidence level of 95%, and a p-value less than 0.05 (p <.05) was used to determine statistical significance in the final model. The assumption of standard logistic regression, such as independence of observation, adequate sample size, linearity of the logit, and absence of multicollinearity were assessed and met during analysis. A total weighted sample of 1,419 (cases 268; controls 1151) was included in the study. Factors were categorized as (1) highly significant and strongly associated (p ˂ 0.001), or (2) moderately or weekly significant and associated (0.001 ˂ p ≤.05). Results A total of 10,025 births occurred in the four hospitals from 1 July to 31 December 2021 and of these, 1,419 files were sampled comprising 268 stillbirths and 1,151 live births. Maternal characteristics Table 1 shows the socio-demographic and economic characteristics of the women whose birth files were analyzed. About 65% of the women with live births and 60% with stillbirths were aged 21 to 34. All the women were of African ethnicity. The Manzini region had the highest representation in both groups, with about 33% of the women with live births and 47% with stillbirths. For both outcomes, most of the women were single- about 63% with live births and 55% with stillbirths. Christianity was the predominant religion in both outcomes, with over 95%. Similarly, the employment status was comparable at about 63% for both groups. Table 1Maternal sociodemographic and economic characteristicsExplanatory variablesBirth outcome (n = 1,419)Live birth (n = 1,151)Stillbirth (n = 268)X2/F; p-valueAge group ≤ 20 years230 (19.98)46 (17.16)9.69; 2 ** 21–34 years752 (65.34)162 (60.45) ≥ 35 years169 (14.68)60 (22.39)Ethnicity African1151 (100.00)268 (100.00)Region of residence Hhohho248 (21.55)62 (23.13)29.41; 3 *** Manzini381 (33.10)126 (47.01) Shiselweni278 (24.15)55 (20.52) Lubombo244 (21.20)25 (9.32)Marital status Single722 (62.73)147 (54.85)6.18; 1 * Married/widowed/separated429 (37.27)121 (45.12)Religion Christianity1104 (95.92)257 (95.90)0.50; 1 Other47 (4.08)11 (4.10)Level of education None/illiterate62 (5.39)20 (7.47)4.91; 3 Primary213 (18.51)38 (14.18) Secondary649 (56.39)161 (60.07) Tertiary227 (19.71)49 (18.28)Employment status Unemployed748 (64.99)169 (63.06)0.10; 1 Employed403 (35.01)99 (36.94)* p ˂0.05, ** p ˂0.01, *** p˂0.001, X2/F− chi−square or Fisher’s exact test Table 1Maternal sociodemographic and economic characteristicsExplanatory variablesBirth outcome (n = 1,419)Live birth (n = 1,151)Stillbirth (n = 268)X2/F; p-valueAge group ≤ 20 years230 (19.98)46 (17.16)9.69; 2 ** 21–34 years752 (65.34)162 (60.45) ≥ 35 years169 (14.68)60 (22.39)Ethnicity African1151 (100.00)268 (100.00)Region of residence Hhohho248 (21.55)62 (23.13)29.41; 3 *** Manzini381 (33.10)126 (47.01) Shiselweni278 (24.15)55 (20.52) Lubombo244 (21.20)25 (9.32)Marital status Single722 (62.73)147 (54.85)6.18; 1 * Married/widowed/separated429 (37.27)121 (45.12)Religion Christianity1104 (95.92)257 (95.90)0.50; 1 Other47 (4.08)11 (4.10)Level of education None/illiterate62 (5.39)20 (7.47)4.91; 3 Primary213 (18.51)38 (14.18) Secondary649 (56.39)161 (60.07) Tertiary227 (19.71)49 (18.28)Employment status Unemployed748 (64.99)169 (63.06)0.10; 1 Employed403 (35.01)99 (36.94)* p ˂0.05, ** p ˂0.01, *** p˂0.001, X2/F− chi−square or Fisher’s exact test Maternal sociodemographic and economic characteristics * p ˂0.05, ** p ˂0.01, *** p˂0.001, X2/F− chi−square or Fisher’s exact test Fetal characteristics Table 2 shows the fetal characteristics of the birth records analyzed in the study. A majority 83% of the live births were term while 63% of the stillbirths were preterm. At the onset of labour, about 94% of live births were known to be alive, while around 61% of stillbirths were known to be dead. Most fetuses had a cephalic presentation, with 93% in live births compared to 86% in stillbirths. Regarding the birth weight, 88% of live births ranged from 2500 to 4000 g, whereas 67% of the stillbirths were under 2500 g. Although female and male fetuses were nearly equal in both groups, females slightly predominated at 50% in live births and 47% in stillbirths. Less than 1% of live births and about 1% of stillbirths had congenital malformations. Regarding fetal complications such as fetal distress and nuchal cord, 14% of live births and 46% of stillbirths were affected. Table 2Fetal characteristics at birthExplanatory variablesBirth outcome (n = 1,419)Live birth (n = 1,151)Stillbirth (n = 268)X2/FGestational age Preterm95 (8.25)169 (63.06)322.36, 2 *** Term952 (82.71)92 (34.33) Postterm104 (9.04)7 (2.61)State of the fetus onset of labour Unknown63 (5.47)34 (12.69)346.04, 2 *** Alive1087 (94.44)70 (26.12) Death1 (0.087)164 (61.19)Fetal presentation Cephalic1078 (93.66)231 (86.19)16.94, 1 *** Other73 (6.34)37 (13.81) Birth weight ‹2500 g103 (8.95)180 (67.16)379.75, 2 *** 2500–4000 g1013 (88.01)82 (30.60) ›4000 g35 (3.04)6 (2.24)Sex Female583 (50.65)128 (47.76)0.64, 1 Male568 (49.35)140 (52.24)Congenital malformation None1146 (99.57)265 (98.88)1.82, 1 Yes5 (0.43)3 (1.12)Fetal complications None982 (85.32)144 (53.73)132.36, 1 *** Present169 (14.68)124 (46.27)* p ˂0.05; ** p ˂0.01; *** p˂0.001 Table 2Fetal characteristics at birthExplanatory variablesBirth outcome (n = 1,419)Live birth (n = 1,151)Stillbirth (n = 268)X2/FGestational age Preterm95 (8.25)169 (63.06)322.36, 2 *** Term952 (82.71)92 (34.33) Postterm104 (9.04)7 (2.61)State of the fetus onset of labour Unknown63 (5.47)34 (12.69)346.04, 2 *** Alive1087 (94.44)70 (26.12) Death1 (0.087)164 (61.19)Fetal presentation Cephalic1078 (93.66)231 (86.19)16.94, 1 *** Other73 (6.34)37 (13.81) Birth weight ‹2500 g103 (8.95)180 (67.16)379.75, 2 *** 2500–4000 g1013 (88.01)82 (30.60) ›4000 g35 (3.04)6 (2.24)Sex Female583 (50.65)128 (47.76)0.64, 1 Male568 (49.35)140 (52.24)Congenital malformation None1146 (99.57)265 (98.88)1.82, 1 Yes5 (0.43)3 (1.12)Fetal complications None982 (85.32)144 (53.73)132.36, 1 *** Present169 (14.68)124 (46.27)* p ˂0.05; ** p ˂0.01; *** p˂0.001 Fetal characteristics at birth * p ˂0.05; ** p ˂0.01; *** p˂0.001 Factors associated with stillbirth Maternal and fetal factors were analyzed using logistic regression to assess stillbirth, first with an unadjusted model, and then with an adjusted model using backward selection. To ensure data quality, factors with over 50% missing data were excluded from the analysis, and only significant ones in the final model are presented; others are in the supplemental materials. For clarity, the significant factors are grouped into (1) maternal health and behavior (Table 3), (2) pregnancy and antenatal care (Table 4), and (3) labour and fetal factors (Table 5). Maternal health and behavior factors Table 3 below shows that 4% of women with live births and 20% with stillbirths reported alcohol consumption. The results show that women who consumed alcohol had significantly higher odds of having stillbirth compared to those who did not (aOR = 4.05; 95% CI = 2.04–8.03; p ˂0.01). Regarding iron and folic acid (IFAS) supplementation, 32% of women with live births and 56% with stillbirth had not taken it, and they had significantly higher odds of stillbirth compared to those who did (aOR = 2.32; 95% CI = 1.50–3.57; p˂0.001). Also, 5% of women with live births and 19% with stillbirths had consumed traditional remedies, and they had three times higher odds of having stillbirth compared to their counterparts (aOR = 3.83; 95% CI = 1.02–3.87). Table 3Maternal health and behaviour factors associated with stillbirthExplanatory variablesBirth outcome (n = 1,419)Multivariable analysisLive birth (n = 1,151)Stillbirth (n = 268)uORaORAlcohol intakeNo1102 (95.74)213 (79.48)11Yes49 (4.26)55 (20.52)6.06(3.39–10.82) ***4.05(2.04–8.03) **IFASNo379 (32.93)149 (55.59)2.55(1.95–3.34) ***2.32(1.50–3.57) ***Yes772 (67.07)119 (44.40)11Intake of traditional remediesNo1083 (94.09)216 (80.60)11Yes68 (5.91)52 (19.40)3.83(2.60–5.66) ***1.99(1.02–3.87) ** p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, IFAS iron and folic acid Table 3Maternal health and behaviour factors associated with stillbirthExplanatory variablesBirth outcome (n = 1,419)Multivariable analysisLive birth (n = 1,151)Stillbirth (n = 268)uORaORAlcohol intakeNo1102 (95.74)213 (79.48)11Yes49 (4.26)55 (20.52)6.06(3.39–10.82) ***4.05(2.04–8.03) **IFASNo379 (32.93)149 (55.59)2.55(1.95–3.34) ***2.32(1.50–3.57) ***Yes772 (67.07)119 (44.40)11Intake of traditional remediesNo1083 (94.09)216 (80.60)11Yes68 (5.91)52 (19.40)3.83(2.60–5.66) ***1.99(1.02–3.87) ** p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, IFAS iron and folic acid Maternal health and behaviour factors associated with stillbirth * p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, IFAS iron and folic acid Pregnancy and antenatal care factors Only a few women, 3% with live births and less than 1% with stillbirths had eight or more ANC visits, and they had significantly lower odds of having stillbirth compared to their counterparts (aOR = 0.03; 95% CI = 0.02–0.30.02.30; p˂0.01). Regarding syphilis test (RPR), 1.3% and 9.3% of women with live births and stillbirths, respectively, tested positive and they had seven times higher odds of having stillbirth compared to those who tested negative (aOR = 7.30; 95% CI = 2.39–22.29; p˂0.001). Less than 30% of women, 7% with live births and 28% with stillbirths had hypertensive disorders of pregnancy (HDP) and had significantly higher odds of stillbirth compared to those who did not have (aOR = 3.49; 95% CI = 1.99–6.09; p˂0.001). Women who had gestational diabetes, less than 1% with live births and 2% with stillbirth had higher odds of stillbirth compared to their counterparts (aOR = 19.52; 95% CI = 1.96–194.69.96.69; p ˂0.05). Regarding anemia in pregnancy, 5% and 15% of women with live births and stillbirth respectively, had two times higher odds of stillbirth compared to their counterparts (aOR = 2.12; 95% CI = 1.99–6.09; p ˂0.05). 1% (1%) and 8% of women with live births and stillbirths, respectively, had antepartum hemorrhage (APH), and had higher odds of stillbirth compared to their counterparts (aOR = 17.04; 95% CI = 4.53–64.09; p˂0.001). Similarly, less than 1% and 3% had live births and stillbirths had abnormal amounts of amniotic fluid (oligohydramnios or polyhydramnios), respectively, and had higher odds of stillbirth compared to those with normal amounts (aOR = 7.44; 95% CI = 1.73–31.99; p˂0.01). Table 4Pregnancy and antenatal care factors associated with stillbirthExplanatory variablesBirth outcomeMultivariable analysisLive birthStillbirthUORAORANC visits0–3157 (13.64)85 (31.72)114–7963 (83.66)182 (67.91)0.35(0.26–0.48.26.48) ***0.66(0.39–1.13.39.13)≥ 831 (2.69)1 (0.37)0.06(0.01–0.44.01.44) **0.03(0.02–0.30.02.30) **RPR testNot done430 (37.36)122 (45.52)11Negative706 (61.34)121 (45.15)0.60(0.46–0.80.46.80) ***0.766(0.49–1.19.49.19)Positive15 (1.30)25 (9.33)5.87(3.00–11.49.00.49) ***7.30(2.39–22.29) ***HDPNo1075 (93.39)193 (72.01)11Yes76 (6.60)75 (27.99)5.50(3.86–7.83) ***3.49(1.99–6.09) ***GDMNo1149 (99.83)262 (97.76)11Yes2 (0.17)6 (2.24)13.16(2.64–65.55) **19.52(1.96–194.69.96.69) *Anemia in pregnancyNo1089 (94.61)229 (85.45)11Yes62 (5.39)39 (14.55)2.99(1.96–4.57) ***2.12(1.99–6.09) *APHNo1139 (98.96)247 (92.16)11Yes12 (1.04)21 (7.84)8.07(3.92–16.62) ***17.04(4.53–64.09) ***Amniotic fluid volumeNormal1142 (99.21)260 (97.01)11Abnormal9 (0.78)8 (2.99)3.90(1.49–10.22) **7.44(1.73–31.99) *** p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, ANC antenatal care, RPR rapid plasma regain, HDP hypertensive disorders of pregnancy, GDM gestational diabetes mellitus, APH antepartum hemorrhage Table 4Pregnancy and antenatal care factors associated with stillbirthExplanatory variablesBirth outcomeMultivariable analysisLive birthStillbirthUORAORANC visits0–3157 (13.64)85 (31.72)114–7963 (83.66)182 (67.91)0.35(0.26–0.48.26.48) ***0.66(0.39–1.13.39.13)≥ 831 (2.69)1 (0.37)0.06(0.01–0.44.01.44) **0.03(0.02–0.30.02.30) **RPR testNot done430 (37.36)122 (45.52)11Negative706 (61.34)121 (45.15)0.60(0.46–0.80.46.80) ***0.766(0.49–1.19.49.19)Positive15 (1.30)25 (9.33)5.87(3.00–11.49.00.49) ***7.30(2.39–22.29) ***HDPNo1075 (93.39)193 (72.01)11Yes76 (6.60)75 (27.99)5.50(3.86–7.83) ***3.49(1.99–6.09) ***GDMNo1149 (99.83)262 (97.76)11Yes2 (0.17)6 (2.24)13.16(2.64–65.55) **19.52(1.96–194.69.96.69) *Anemia in pregnancyNo1089 (94.61)229 (85.45)11Yes62 (5.39)39 (14.55)2.99(1.96–4.57) ***2.12(1.99–6.09) *APHNo1139 (98.96)247 (92.16)11Yes12 (1.04)21 (7.84)8.07(3.92–16.62) ***17.04(4.53–64.09) ***Amniotic fluid volumeNormal1142 (99.21)260 (97.01)11Abnormal9 (0.78)8 (2.99)3.90(1.49–10.22) **7.44(1.73–31.99) *** p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, ANC antenatal care, RPR rapid plasma regain, HDP hypertensive disorders of pregnancy, GDM gestational diabetes mellitus, APH antepartum hemorrhage Pregnancy and antenatal care factors associated with stillbirth * p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, ANC antenatal care, RPR rapid plasma regain, HDP hypertensive disorders of pregnancy, GDM gestational diabetes mellitus, APH antepartum hemorrhage Labour and fetal factors Regarding place of birth, 6% and 11% of live births and stillbirths were born before arrival, and the women had significantly higher odds of having stillbirths compared to those who gave birth in health facilities (aOR = 1.87; 95% CI = 1.75–2.51; p˂0.001). About 1% of live births and 7% of stillbirths had MSL II, while less than 1% and 3%, respectively had MSL III, and all these had significantly higher odds of stillbirth compared to their counterparts (aOR = 11.42; 95% CI = 4.30–30.35.30.35; p˂0.001 & aOR = 7.11; 95% CI = 1.49–33.91; p ˂0.05). Regarding gestational age, 8% live births and 63% stillbirth were preterm, and had significantly higher odds of being born still compared to those who were term (aOR = 14.27; 95% CI = 4.71–43.19; p˂0.001). About 14% of live births and 46% of stillbirths had fetal complications, which were significantly associated with three times higher odds of being stillbirths (aOR = 3.17; 95% CI = 1.99–5.11; p˂0.001). Table 5Labour and fetal factors associated with stillbirthExplanatory variablesBirth outcome (n = 1,419)Multivariable analysisLive birth (n = 1,151)Stillbirth (n = 268)UORAORPlace of birthBBA70(6.08)32(11.94)2.09(1.35–3.26) ***1.87(1.75–2.51) ***Health facility1081(93.92)236(88.06)11MSLNone1060 (92.09)225 (83.96)1-MSL I71 (6.17)14 (5.22)0.93(0.51–1.68.51.68)1.29(0.57–2.94.57.94)MSL II16 (1.39)19 (7.09)5.59(2.83–11.05) ***11.42(4.30–30.35.30.35) ***MSL III4 (0.35)10 (3.73)11.78(3.66–37.89) ***7.12(1.49–33.91) *GAPreterm95 (8.25)169 (63.06)26.43(11.81–59.15) ***14.27(4.71–43.19) ***Term952 (82.71)92 (34.33)1.44(0.65–3.18.65.18)3.58(0.12–10.02.12.02)Postterm104 (9.04)7 (2.61)11Fetal complicationNone982 (85.32)144 (53.73)11Present169 (14.68)124 (46.27)5.00(3.74–6.69) ***3.17(1.99–5.11) **** p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, BBA birth before arival, MSL meconium−stained liquor, GA gestational age Table 5Labour and fetal factors associated with stillbirthExplanatory variablesBirth outcome (n = 1,419)Multivariable analysisLive birth (n = 1,151)Stillbirth (n = 268)UORAORPlace of birthBBA70(6.08)32(11.94)2.09(1.35–3.26) ***1.87(1.75–2.51) ***Health facility1081(93.92)236(88.06)11MSLNone1060 (92.09)225 (83.96)1-MSL I71 (6.17)14 (5.22)0.93(0.51–1.68.51.68)1.29(0.57–2.94.57.94)MSL II16 (1.39)19 (7.09)5.59(2.83–11.05) ***11.42(4.30–30.35.30.35) ***MSL III4 (0.35)10 (3.73)11.78(3.66–37.89) ***7.12(1.49–33.91) *GAPreterm95 (8.25)169 (63.06)26.43(11.81–59.15) ***14.27(4.71–43.19) ***Term952 (82.71)92 (34.33)1.44(0.65–3.18.65.18)3.58(0.12–10.02.12.02)Postterm104 (9.04)7 (2.61)11Fetal complicationNone982 (85.32)144 (53.73)11Present169 (14.68)124 (46.27)5.00(3.74–6.69) ***3.17(1.99–5.11) **** p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, BBA birth before arival, MSL meconium−stained liquor, GA gestational age Labour and fetal factors associated with stillbirth * p ˂0.05, ** p ˂0.01, *** p˂0.001, uOR unadjusted odds ratio, aOR adjusted odds ratio, BBA birth before arival, MSL meconium−stained liquor, GA gestational age Discussion This study identified several key factors significantly associated with stillbirth in Eswatini. Using adjusted logistic regression analysis with backward selection, the following variables remained significant predictors of stillbirth. At p˂0.001, there is a lack of IFAS supplementation, a positive RPR test, HDP, APH, grade II MSL, fetal complications, preterm, and BBA at a health facility. At p ˂0.01, alcohol consumption during pregnancy, having more than eight ANC visits/contacts, and abnormal amniotic fluid levels were also significantly associated with stillbirth. Additionally, at p ˂0.05, the intake of traditional remedies, gestational diabetes mellitus, anemia in pregnancy, and grade III MSL were identified as significant factors. The current study found a strong association between lack of IFAS and stillbirth. While our effect size (aOR = 2.32) was lower than that reported in Ethiopia [26] (aOR = 8.26), both studies found a significant positive association between lack of IFAS and stillbirth, reinforcing the consistency of this relationship across different populations. This was also similar to a study conducted in Northern India [27]. The observed higher odds of stillbirth could be attributed to that lack of iron and folic acid could lead to maternal anemia and fetal developmental defects respectively, both possibly compromising proper fetal growth, increasing the risk of stillbirth. Therefore, this underscores the importance of promoting and ensuring universal access to IFAS for pregnant women, as per the daily intake recommendation of 30 mg to 60 mg of elemental iron and 400 g (0.4 mg) of folic acid (WHO, 28). Additionally, women who had anemia had higher odds of stillbirth in our study consistent with findings reported in a prospective cohort in India [28], systematic review and meta-analysis in SSA [12]. As with other factors, the association likely stems from compromised maternal and fetal function, which disrupts fetal physiology and may lead to stillbirth. However, our finding also contradicts those of other studies where no association between anemia and stillbirth was found [29]. Regarding syphilis, women who had a positive RPR test had higher odds of having a stillbirth in the study. A higher effect size (aOR = 7.30) was observed in the current study compared to aOR = 2.03 and aOR = 1.78 for non-latent syphilis and maternal titers over 1.4 reported in China [20], where the difference may be attributed to the duration of data collection, with our study analyzing 6 months of data compared to 6 years. Similarly, our results are consistent with those of studies conducted in Malawi [30], and China [20] which found that syphilis was associated with stillbirth. This result may be explained by that syphilis can cross the placenta causing several effects such as impairing placental function in gaseous exchange and nutrient transfer and causing fetal inflammation and organ damage, resulting in fetal demise. This finding highlights the significance of screening and treating pregnant women for syphilis, promoting safe sexual practices during pregnancy, and ensuring the continuous availability of testing kits at all levels of healthcare facilities. In addition, the current study revealed that HDP increased the odds of stillbirth. Although consistent in direction, our effect size (aOR = 3.49) was smaller than that reported in Cameroon (aOR = 18.0) [31] despite the similarities in study design. These findings align with several studies, including a multi-center case-control study in Kenya [29]. Specific HDP conditions, such as pre-eclampsia [13, 16, 32] and eclampsia [33], have also been independently associated with stillbirth in multiple studies. The observed association between HDP and stillbirth may have to do with the possible abnormal development and narrowing of the placental blood vessels, leading to poor placental perfusion restricting nutrient and oxygen supply to the fetus hence growth restriction, hypoxia, and possible stillbirth. This finding underscores the need for close monitoring and timely and effective management of HDP. Women who had APH, encompassing placenta abruption or placenta previa had higher odds of stillbirth. Compared to previous studies with similar effect direction (aOR = 2.70) [12], (aOR = 3.33) [32], the effect size in our study was higher (aOR = 17.04) Similarly, our findings are also consistent with findings from Namibia [34], Kenya [29], Ethiopia [32, 35], and Malawi [30] among others, which reported a significant association of APH or its specific conditions with stillbirth. The possible interference of either placenta previa or abruption on placenta functions cannot be ruled out as it can cause acute or chronic events to either the fetus or mother which can contribute to stillbirth. This finding emphasizes the significance of enhancing midwives’ skills in obstetric emergencies to ensure timely recognition and effective management of APH to improve maternal and fetal outcomes. Regarding meconium-stained liquor, MSL grades II and III were significantly associated with increased odds of stillbirth. This study identified a greater magnitude of effect (aOR = 17.04) relative to (aOR = 5.01) reported in Ethiopia [16] which may be attributed to the differences in data sources used. The positive significant association was also consistent with findings from studies conducted in China [36], Ethiopia [16, 35], and Tanzania [37], where MSL was also linked to a higher risk of stillbirth. Passage of meconium in utero reflects underlying fetal compromise such as fetal distress due to varying complications such as umbilical cord compression, intrauterine infection, or placental insufficiency. Also, when the meconium is inhaled it can cause airway obstruction and inflammation. These factors could explain the relative or high level of significant association with stillbirth. Therefore, these findings underscore the importance of enhancing prenatal and intrapartum care to prevent conditions that could lead to fetal distress and ensure timely identification and management of fetal distress to minimize or mitigate its effects on the fetus. The current study also found that the presence of fetal complications significantly increased the odds of stillbirth. This supports findings from other studies worldwide, such as a systematic review of observational studies in South Asia [38] and a case-control study in Ethiopia [16] which reported fetal complications such as umbilical cord problems, fetal distress as factors significantly associated with stillbirth. The implication of this finding in practice is the implementation of appropriate interventions, including improved maternal health, education and awareness, regular ANC contacts, and enhanced fetal surveillance (e.g., ultrasound, non-stress test, biophysical profiles), to identify potential fetal complications and ensure timely management. Also, fetus born preterm had higher odds of stillbirth in the study (aOR = 14.27), congruent with findings from Ethiopia (aOR = 2.08; aOR = 10) [32, 33], Ghana (aOR = 9.37) [39], SSA [40] and India [28]. This may be due to the immaturity of fetal organs and systems making them prone to infections and effects from altered maternal health, underscoring the strengthening of preventive measures to prevent preterm labour and birth. The study also found an association between birth before arrival (BBA) and increased odds of stillbirth. Comparable data for this variable is limited. This finding may be translated to delays in seeking care, possibly due to factors such as lack of prenatal care, inadequate awareness, distance to healthcare facilities, and lack of birth and emergency preparedness. This aligns with a study in Papua New Guinea, which reported that delays in seeking care during labor significantly raise the risk of stillbirth [41]. The implication of this finding to practice involves improving antenatal care, birth preparedness, and women’s education while enhancing healthcare accessibility. Additionally, further research is needed to investigate the reasons behind BBA and the factors contributing to delays in seeking care. The study found that the odds of stillbirths were two times higher in women who consumed alcohol during pregnancy compared to those who did not (aOR = 4.05). This is consistent with earlier studies conducted in Cameroon (aOR = 9.00) [31] and SSA (aOR = 1.58) [42], Ethiopia [35, 43], Kenya [44],. A possible explanation for this result may be based on previous work that revealed alcohol-induced changes in the intrauterine environment [45] which may cause detrimental effects such as placental insufficiency and congenital abnormalities [46]. Moreover, maternal nutrition and physical health play a crucial role in fetal development [47] and similarly, alcohol can contribute to maternal nutritional deficiencies and high blood pressure ultimately affecting the fetus. Therefore, the toxic effect of alcohol on both mother and fetus likely contributes to the increased odds of stillbirth. The findings on alcohol underscore the need for patient education and behavioral counseling and support. Regarding ANC visits or contacts, the study found women who had eight or more had significantly lower odds of stillbirth (aOR = 0.66). This finding is congruent with findings reported in Ghana (aOR = 0.60) [19] and a systematic review and meta-analysis in SSA (aOR = 0.45) [12] where a higher number of visits (four or more) was a protective factor against stillbirth. Other studies such as in Ethiopia [33], India [28] and SSA [40] reported higher odds of stillbirth in women who had no or minimal (less than 4) ANC visits, which further shed light on our findings. It seems possible that these results are due to that more ANC means better monitoring, early detection, prevention, and timely interventions, all key in reducing the risk of stillbirth. Therefore, findings underscore ANC as a critical intervention, guiding both clinical decision-making and public health strategies. Moreover, the study identified that abnormal AFV, either polyhydramnios or oligohydramnios, increased the odds of stillbirth. The effect size in our study (aOR = 7.44) exceeds that reported in systematic and meta-analysis (RR = 4.75) [48], although the association remains positive. While our estimates differ in strength, it support the majority of trends, including those conducted in Ethiopia [43] and Latvia [49], which reported significant associations between one or both of these conditions and stillbirth These results could be explained by the disruption of the protective and regulatory factors of the intrauterine environment such as placental insufficiency in oligohydramnios and fetal anomalies or premature rupture of membranes in polyhydramnios, significantly increasing the risk of fetal compromise and ultimate demise. This finding emphasizes the significance of evaluating and monitoring amniotic fluid during pregnancy, which can be addressed by following the recommended ultrasound before 24 weeks for all pregnancies [50] and additional scans for high-risk pregnancies, along with ensuring timely birth when necessary. Also, the study found that consuming traditional remedies during pregnancy and/or labour increased the odds of stillbirth. However, limited research has directly examined this phenomenon, but other studies have explored other cultural practices, such as female genital mutilation (FGM) which were reported to be significantly associated with stillbirth [51]. Future research is recommended to explore the types of traditional remedies used during pregnancy and labor in various regions and cultures, along with their potentially harmful effects on fetal health, to enhance understanding of their relationship with stillbirth risk and improve maternal and fetal health outcomes. Regarding diabetes, our study found that women who had GDM had higher odds of stillbirth, consistent with findings reported in Malawi [30]. However, this finding contrasts with a systematic review of SSA, which found no significant association [40]. Notably, a systematic review reported inconsistent findings on GDM and stillbirth, concluding that due to substantial study heterogeneity, the association remains unclear [52]. Therefore, the current results must be interpreted with caution and underscores the need for more robust research to further understand the association. These results should be interpreted with consideration of limitations. The use of secondary data limited control over data quality and lacked information on exposure timing; however, several steps were taken to address potential biases, including sampling files with complete information on certain key variables, imputation of missing data, and exclusion of variables with more than 50% missing values during analysis. While unmatched case-control design limited causal inference and control of confounding by design, multivariable logistic regression was conducted and built systematically using backward selection to adjust for confounders. Additionally, the quantitative approach limited the in-depth exploration of variables, its strengths such as multi-center design, large sample size, use of recent data, and inclusion of emerging variables like COVID-19 infection and use of traditional remedies, enhanced statistical power, relevance, and generalizability. Conclusions This study highlights key stillbirth-associated factors with varying degrees of significance, emphasizing the need for targeted interventions. The most significant factors were lack of IFAS, positive RPR status, HDP, APH, MSL II, fetal complications, and BBA. These findings underscore the importance of comprehensive antenatal screening and management. The broader implication suggests that improving maternal healthcare access and quality, particularly in the provision of preventive measures, screening, and managing infections and obstetric complications, could substantially reduce the stillbirth rate. These interventions are low-cost, highly feasible, and have demonstrated cost-effectiveness in improving maternal-fetal outcomes, particularly in low-resource settings. This study contributes to the growing body of evidence by identifying context-specific factors with a higher predictive value for stillbirth informing local and regional maternal health policies. Future research should focus on longitudinal studies to assess causality, as well as interventional studies to evaluate the effectiveness of integrated antenatal care packages to address the identified factors among other research initiatives. Immediate action is needed from healthcare systems, policymakers, and national health programs such as reallocating resources towards proven, cost-effective maternal health services to strengthen ANC and ensure early detection and management of these preventable factors to reduce the burden of stillbirth and improve maternal and perinatal outcomes.