Determinants of newborn birth size in Somaliland: an ordinal logistic regression analysis Authors: Abdirahman Omar Ibrahim, Bashaar Ali Daud, Waleed Salahudiin Ahmed, Suhaib Muse Jama, Nagaad Abdinasir Habane, Rahma Yasin Ismail, Umulkhayr Mohamed Abadir, Amaani Abdilahi Sahal, Mohamed Said Hassan Published: 2025 ABSTRACT BackgroundUnderstanding the predictors of birth size is essential for designing effective interventions to reduce neonatal morbidity and mortality in Somaliland.MethodsA comprehensive cross-sectional study was conducted using the 2020 Somali Health and Demographic Survey (SHDS). An ordinal logistic regression model, a robust statistical method, was applied to identify determinants of birth size (Low, Normal, High) among 2,137 mother-infant pairs, ensuring a thorough and reliable analysis.ResultsThe prevalence of reported small, average, and large birth sizes was 9.0%, 75.0%, and 16.0%, respectively. Our study identified significant predictors of birth size, including household wealth, maternal age, and birth order. The adjusted odds of having a larger baby were significantly higher for infants from the wealthiest households compared to the poorest (AOR = 1.45; 95% CI: 1.12–1.87), for second-order births compared to first-borns (AOR = 1.38; 95% CI: 1.10–1.74), and for infants of mothers aged 30–34 compared to adolescent mothers (AOR = 2.09; 95% CI: 1.25–3.48). These findings provide valuable insights for designing targeted interventions to improve maternal and child health in Somaliland.ConclusionOur findings underscore the importance of socioeconomic status and maternal reproductive patterns in determining birth size. We recommend tailored interventions that target the identified risk factors, with a focus on nutritional support for poor households and reproductive health education for young women. Implementing these recommendations has the potential to significantly improve maternal and child health outcomes in Somaliland, offering hope for a healthier future and a brighter outlook for the region’s health. However, our findings also suggest that while factors such as antenatal care attendance are important, their impact may be limited without concurrent improvements in service quality and the empowerment of women, underscoring the need for a multifaceted public health strategy.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13690-025-01765-3. Text box 1. Contributions to the literature• Provides crucial, up-to-date evidence on newborn health from Somaliland, a region with limited health data• Identifies key vulnerable groups for poor birth outcomes—specifically, infants of adolescent mothers and those from the poorest households—to guide targeted public health interventions.Identifies key vulnerable groups for poor birth outcomes—specifically, infants of adolescent mothers and those from the poorest households—to guide targeted public health interventions.• Highlights that while socioeconomic status is a powerful predictor, the impact of factors like maternal education and antenatal care attendance may be limited without improvements in service quality and women’s autonomy. Text box 1. Contributions to the literature• Provides crucial, up-to-date evidence on newborn health from Somaliland, a region with limited health data• Identifies key vulnerable groups for poor birth outcomes—specifically, infants of adolescent mothers and those from the poorest households—to guide targeted public health interventions.Identifies key vulnerable groups for poor birth outcomes—specifically, infants of adolescent mothers and those from the poorest households—to guide targeted public health interventions.• Highlights that while socioeconomic status is a powerful predictor, the impact of factors like maternal education and antenatal care attendance may be limited without improvements in service quality and women’s autonomy. Introduction Newborn birth weight is a critical indicator of maternal health, nutrition, and the quality of the intrauterine environment. The World Health Organization (WHO) categorizes birth weight as Low Birth Weight (LBW) (< 2.5 kg), Normal Birth Weight (NBW) (2.5–4 kg), and High Birth Weight (HBW) (>4 kg) [1]. LBW is a significant public health problem globally, contributing significantly to neonatal morbidity and mortality, as well as long-term health consequences such as stunting, developmental delays, and increased risk of chronic diseases in adulthood [2, 3]. In sub-Saharan Africa, the prevalence of LBW is estimated to be between 13% and 15% and is a primary driver of the region’s high neonatal mortality rates [1]. In Somalia and Somaliland, where conflict and poverty have weakened health systems, understanding the local determinants of birth size is not only crucial but also urgent for developing effective maternal and child health policies. This study aims to identify the sociodemographic and maternal factors that influence newborn birth size in Somaliland, addressing a pressing need for actionable research. Literature review The determinants of birth weight are multifactorial, involving a complex interplay of maternal, socioeconomic, and environmental factors. Maternal factors are paramount. Adolescent motherhood (< 20 years) is a well-established risk factor for LBW, often linked to biological immaturity, poor nutrition, and lower utilization of health services [4, 5]. Similarly, short birth intervals (< 24 months) are strongly associated with adverse birth outcomes, as they do not allow for sufficient maternal nutritional recovery between pregnancies [3]. Parity also plays a role, with first-born infants often having a lower birth weight compared to subsequent births [1]. Socioeconomic status is a significant contributor to disparities in birth outcomes. Numerous studies have demonstrated a strong link between low household wealth and an increased risk of LBW [4, 5]. Poverty limits access to nutritious food, quality healthcare, and adequate housing, all of which are essential for a healthy pregnancy. Maternal education is another powerful predictor; educated women are more likely to have better health literacy, make informed decisions about their health, and utilize services like Antenatal Care (ANC) [6]. Health service utilization, particularly ANC, is a key mediating factor. Regular ANC visits provide opportunities for health professionals to monitor fetal growth, provide nutritional counseling, and manage pregnancy-related complications, thereby reducing the risk of LBW [7]. This study will use an ordinal logistic regression model to simultaneously assess these determinants and understand their relative importance in the context of Somaliland. Data and methodology Data source This study used data from the 2020 Somali Health and Demographic Survey (SHDS). The survey covered all six administrative regions of Somaliland. A total of n = 2,137 mother-infant pairs were included. The analysis was restricted to the last live birth for each mother in the five years preceding the survey to ensure data quality and minimize recall bias. Variable Description Dependent variable: Newborn Birth Size, an ordinal variable with three categories (1 = Low, 2 = Normal, 3 = High).Independent variables: Maternal age, education level, household wealth index (a composite measure of a household’s economic status based on factors such as income, assets, and living conditions), ANC visits (recategorized as 0 visits, 1–3 visits, and 4 or more visits based on WHO recommendations), birth interval, birth order, place of residence, and sex of the child.As a sensitivity analysis to test the robustness of the proportional odds assumption, a multinomial logistic regression was also performed. Dependent variable: Newborn Birth Size, an ordinal variable with three categories (1 = Low, 2 = Normal, 3 = High). Dependent variable: Newborn Birth Size, an ordinal variable with three categories (1 = Low, 2 = Normal, 3 = High). Independent variables: Maternal age, education level, household wealth index (a composite measure of a household’s economic status based on factors such as income, assets, and living conditions), ANC visits (recategorized as 0 visits, 1–3 visits, and 4 or more visits based on WHO recommendations), birth interval, birth order, place of residence, and sex of the child. Independent variables: Maternal age, education level, household wealth index (a composite measure of a household’s economic status based on factors such as income, assets, and living conditions), ANC visits (recategorized as 0 visits, 1–3 visits, and 4 or more visits based on WHO recommendations), birth interval, birth order, place of residence, and sex of the child. As a sensitivity analysis to test the robustness of the proportional odds assumption, a multinomial logistic regression was also performed. As a sensitivity analysis to test the robustness of the proportional odds assumption, a multinomial logistic regression was also performed. Results Distribution of newborn birth size The distribution of newborn birth size as reported by mothers in the Somaliland survey. The findings show that the vast majority of infants, 75%, were perceived to be of Average Birth Size. A smaller proportion was reported as Large Birth size (16%), while the Small Birth Size category, which includes infants at risk for low birth weight, constituted 9% of the sample. This distribution provides the baseline for the subsequent ordinal logistic regression analysis, which aims to identify the key maternal and socioeconomic factors that predict whether a child is more likely to be in a smaller or larger birth size category (Fig. 1).Fig. 1Distribution of Mother-Reported Newborn Birth Size in Somaliland (2020 Somali Health and Demographic Survey, n = 2,137) Distribution of Mother-Reported Newborn Birth Size in Somaliland (2020 Somali Health and Demographic Survey, n = 2,137) Descriptive statistics The characteristics of the 2,137 participants are presented in Table 1. The majority of mothers (82.0%) had no formal education, and a significant proportion (52.4%) came from low-income households. The prevalence of small birth size was 9.0%. There were substantial bivariate associations (p < 0.001) between birth size and maternal education, wealth, and attendance at ANC.Table 1Socio-Demographic characteristics of study participants in Somaliland (2020 Somali health and demographic Survey, n = 2,137)VariableCategoryFrequency (%)No ANC Visit n (%)Yes ANC Visit n (%)χ²P-value Type of ResidenceRural1124 (52.60%)671 (59.7%)453 (40.3%)0.27500.600Urban1013 (47.40%)616 (60.8%)397 (39.2%) Age (in 5-year groups)15–19122 (5.71%)83 (68.0%)39 (32.0%)11.92010.06420–24389 (18.20%)235 (60.4%)154 (39.6%)25–29578 (27.05%)328 (56.7%)250 (43.3%)30–34458 (21.43%)272 (59.4%)186 (40.6%)35–39374 (17.50%)223 (59.6%)151 (40.4%)40–44157 (7.35%)109 (69.4%)48 (30.6%)45–4959 (2.76%)37 (62.7%)22 (37.3%) RegionAwdal275 (12.87%)121 (44.0%)154 (56.0%)125.48200.000Marodijeh240 (11.23%)115 (47.9%)125 (52.1%)Sahil239 (11.18%)117 (49.0%)122 (51.0%)Togdheer400 (18.72%)221 (55.3%)179 (44.7%)Sool448 (20.96%)333 (74.3%)115 (25.7%)Sanaag535 (25.04%)380 (71.0%)155 (29.0%) Education LevelNo education1752 (81.98%)1171 (66.8%)581 (33.2%)186.56530.000Primary289 (13.52%)99 (34.3%)190 (65.7%)Secondary62 (2.90%)13 (21.0%)49 (79.0%)Higher34 (1.59%)4 (11.8%)30 (88.2%) Household Radio OwnershipYes323 (15.11%)206 (63.8%)117 (36.2%)2.00470.157No1814 (84.89%)1081 (59.6%)733 (40.4%) Household TV OwnershipYes368 (17.22%)204 (55.4%)164 (44.6%)4.25770.039No1769 (82.78%)1083 (61.2%)686 (38.8%) Sex of Household HeadMale1300 (60.83%)773 (59.5%)527 (40.5%)0.80680.369Female837 (39.17%)514 (61.4%)323 (38.6%) Sex of ChildMale1132 (52.97%)684 (60.4%)448 (39.6%)------Female1005 (47.03%)603 (60.0%)402 (40.0%) Vitamin A SupplementationNo1862 (87.13%)1207 (64.8%)655 (35.2%)127.71170.000Yes275 (12.87%)80 (29.1%)195 (70.9%) Household Size (HHM)0–3460 (21.53%)286 (62.2%)174 (37.8%)1.28100.5274–6935 (43.75%)564 (60.3%)371 (39.7%)7–9742 (34.72%)437 (58.9%)305 (41.1%) Birth Order (BO)First birth273 (12.77%)140 (51.3%)133 (48.7%)11.46250.003Second birth340 (15.91%)201 (59.1%)139 (40.9%)Third or more1524 (71.31%)946 (62.1%)578 (37.9%) Birth Size (BS)Large351 (16.42%)185 (52.7%)166 (47.3%)13.11350.001Average1594 (74.59%)995 (62.4%)599 (37.6%)Small192 (8.98%)107 (55.7%)85 (44.3%) Income LevelPoor1120 (52.41%)876 (78.2%)244 (21.8%)351.92640.000Middle243 (11.37%)137 (56.4%)106 (43.6%)Rich774 (36.22%)274 (35.4%)500 (64.6%) Table 1Socio-Demographic characteristics of study participants in Somaliland (2020 Somali health and demographic Survey, n = 2,137)VariableCategoryFrequency (%)No ANC Visit n (%)Yes ANC Visit n (%)χ²P-value Type of ResidenceRural1124 (52.60%)671 (59.7%)453 (40.3%)0.27500.600Urban1013 (47.40%)616 (60.8%)397 (39.2%) Age (in 5-year groups)15–19122 (5.71%)83 (68.0%)39 (32.0%)11.92010.06420–24389 (18.20%)235 (60.4%)154 (39.6%)25–29578 (27.05%)328 (56.7%)250 (43.3%)30–34458 (21.43%)272 (59.4%)186 (40.6%)35–39374 (17.50%)223 (59.6%)151 (40.4%)40–44157 (7.35%)109 (69.4%)48 (30.6%)45–4959 (2.76%)37 (62.7%)22 (37.3%) RegionAwdal275 (12.87%)121 (44.0%)154 (56.0%)125.48200.000Marodijeh240 (11.23%)115 (47.9%)125 (52.1%)Sahil239 (11.18%)117 (49.0%)122 (51.0%)Togdheer400 (18.72%)221 (55.3%)179 (44.7%)Sool448 (20.96%)333 (74.3%)115 (25.7%)Sanaag535 (25.04%)380 (71.0%)155 (29.0%) Education LevelNo education1752 (81.98%)1171 (66.8%)581 (33.2%)186.56530.000Primary289 (13.52%)99 (34.3%)190 (65.7%)Secondary62 (2.90%)13 (21.0%)49 (79.0%)Higher34 (1.59%)4 (11.8%)30 (88.2%) Household Radio OwnershipYes323 (15.11%)206 (63.8%)117 (36.2%)2.00470.157No1814 (84.89%)1081 (59.6%)733 (40.4%) Household TV OwnershipYes368 (17.22%)204 (55.4%)164 (44.6%)4.25770.039No1769 (82.78%)1083 (61.2%)686 (38.8%) Sex of Household HeadMale1300 (60.83%)773 (59.5%)527 (40.5%)0.80680.369Female837 (39.17%)514 (61.4%)323 (38.6%) Sex of ChildMale1132 (52.97%)684 (60.4%)448 (39.6%)------Female1005 (47.03%)603 (60.0%)402 (40.0%) Vitamin A SupplementationNo1862 (87.13%)1207 (64.8%)655 (35.2%)127.71170.000Yes275 (12.87%)80 (29.1%)195 (70.9%) Household Size (HHM)0–3460 (21.53%)286 (62.2%)174 (37.8%)1.28100.5274–6935 (43.75%)564 (60.3%)371 (39.7%)7–9742 (34.72%)437 (58.9%)305 (41.1%) Birth Order (BO)First birth273 (12.77%)140 (51.3%)133 (48.7%)11.46250.003Second birth340 (15.91%)201 (59.1%)139 (40.9%)Third or more1524 (71.31%)946 (62.1%)578 (37.9%) Birth Size (BS)Large351 (16.42%)185 (52.7%)166 (47.3%)13.11350.001Average1594 (74.59%)995 (62.4%)599 (37.6%)Small192 (8.98%)107 (55.7%)85 (44.3%) Income LevelPoor1120 (52.41%)876 (78.2%)244 (21.8%)351.92640.000Middle243 (11.37%)137 (56.4%)106 (43.6%)Rich774 (36.22%)274 (35.4%)500 (64.6%) Socio-Demographic characteristics of study participants in Somaliland (2020 Somali health and demographic Survey, n = 2,137) Bivariate and multivariate ordinal logistic regression analysis Table 2 presents the results from the ordinal logistic regression models, showing both the crude odds ratios (COR) from the bivariate analysis and the adjusted odds ratios (AOR) from the final multivariate model. The odds ratios represent the odds of a newborn being in a higher birth size category (e.g., moving from Low to Normal, or Normal to High). Table 2Crude and adjusted ordinal logistic regression of factors associated with newborn birth size in Somaliland (2020 Somali health and demographic Survey, n = 2,137)VariableCrude Odds Ratio (95% CI)P-valueAdjusted Odds Ratio (95% CI)P-valueAge Group (years) 15–19 (Ref)1.00 (—)—1.00 (—)— 20–241.71 (1.13–2.58)0.011*1.81 (1.15–2.85)0.010* 25–291.63 (1.09–2.43)0.017*1.61 (1.00–2.61)0.050* 30–342.19 (1.46–3.29)< 0.001***2.09 (1.25–3.48)0.005** 35–392.01 (1.33–3.04)0.001**2.05 (1.23–3.43)0.006** 40–441.82 (1.14–2.90)0.012*1.72 (0.96–3.08)0.068 45–491.31 (0.69–2.50)0.4131.19 (0.59–2.40)0.627Place of Residence Rural (Ref)1.00 (—)—1.00 (—)— Urban1.12 (0.92–1.37)0.2581.06 (0.83–1.36)0.629Education Level No education (Ref)1.00 (—)—1.00 (—)— Primary1.79 (1.40–2.29)< 0.001***1.38 (0.96–1.98)0.082 Secondary3.91 (2.45–6.25)< 0.001***2.51 (1.41–4.49)0.002** Higher3.25 (1.68–6.27)0.001**1.10 (0.54–2.25)0.796Birth Order First (Ref)1.00 (—)—1.00 (—)— Second1.51 (1.16–1.96)0.002**1.38 (1.10–1.74)0.006** Third or more1.25 (1.01–1.56)0.041*1.12 (0.84–1.50)0.428ANC Visit No visit (Ref)1.00 (—)—1.00 (—)— Visited2.15 (1.75–2.64)< 0.001***1.15 (0.91–1.44)0.245Income Level Poor (Ref)1.00 (—)—1.00 (—)— Middle1.43 (1.07–1.91)0.015*1.21 (0.88–1.67)0.239 Rich2.58 (2.13–3.12)< 0.001***1.45 (1.12–1.87)0.005** Odds Ratios (ORs) indicate the likelihood of having a newborn in a higher birth size category. CI = Confidence Interval. Ref = Reference category. Significance levels: *p < 0.05, **p < 0.01, ***p < 0.001. Table 2Crude and adjusted ordinal logistic regression of factors associated with newborn birth size in Somaliland (2020 Somali health and demographic Survey, n = 2,137)VariableCrude Odds Ratio (95% CI)P-valueAdjusted Odds Ratio (95% CI)P-valueAge Group (years) 15–19 (Ref)1.00 (—)—1.00 (—)— 20–241.71 (1.13–2.58)0.011*1.81 (1.15–2.85)0.010* 25–291.63 (1.09–2.43)0.017*1.61 (1.00–2.61)0.050* 30–342.19 (1.46–3.29)< 0.001***2.09 (1.25–3.48)0.005** 35–392.01 (1.33–3.04)0.001**2.05 (1.23–3.43)0.006** 40–441.82 (1.14–2.90)0.012*1.72 (0.96–3.08)0.068 45–491.31 (0.69–2.50)0.4131.19 (0.59–2.40)0.627Place of Residence Rural (Ref)1.00 (—)—1.00 (—)— Urban1.12 (0.92–1.37)0.2581.06 (0.83–1.36)0.629Education Level No education (Ref)1.00 (—)—1.00 (—)— Primary1.79 (1.40–2.29)< 0.001***1.38 (0.96–1.98)0.082 Secondary3.91 (2.45–6.25)< 0.001***2.51 (1.41–4.49)0.002** Higher3.25 (1.68–6.27)0.001**1.10 (0.54–2.25)0.796Birth Order First (Ref)1.00 (—)—1.00 (—)— Second1.51 (1.16–1.96)0.002**1.38 (1.10–1.74)0.006** Third or more1.25 (1.01–1.56)0.041*1.12 (0.84–1.50)0.428ANC Visit No visit (Ref)1.00 (—)—1.00 (—)— Visited2.15 (1.75–2.64)< 0.001***1.15 (0.91–1.44)0.245Income Level Poor (Ref)1.00 (—)—1.00 (—)— Middle1.43 (1.07–1.91)0.015*1.21 (0.88–1.67)0.239 Rich2.58 (2.13–3.12)< 0.001***1.45 (1.12–1.87)0.005** Odds Ratios (ORs) indicate the likelihood of having a newborn in a higher birth size category. CI = Confidence Interval. Ref = Reference category. Significance levels: *p < 0.05, **p < 0.01, ***p < 0.001. Crude and adjusted ordinal logistic regression of factors associated with newborn birth size in Somaliland (2020 Somali health and demographic Survey, n = 2,137) Odds Ratios (ORs) indicate the likelihood of having a newborn in a higher birth size category. CI = Confidence Interval. Ref = Reference category. Significance levels: *p < 0.05, **p < 0.01, ***p < 0.001. Model Diagnostics Proportional Odds Assumption Test (Brant Test): p > 0.05 → Assumption met, ordinal logistic regression is appropriate.Classification Accuracy: 75.1%.VIF values: All < 2.0 → No significant multicollinearity detected. Proportional Odds Assumption Test (Brant Test): p > 0.05 → Assumption met, ordinal logistic regression is appropriate. Proportional Odds Assumption Test (Brant Test): p > 0.05 → Assumption met, ordinal logistic regression is appropriate. Classification Accuracy: 75.1%. Classification Accuracy: 75.1%. VIF values: All < 2.0 → No significant multicollinearity detected. VIF values: All < 2.0 → No significant multicollinearity detected. Discussion This study identified that fundamental socioeconomic and biological factors are the paramount drivers of newborn birth size in Somaliland. Specifically, the odds of a larger birth size were significantly higher for infants from the wealthiest households (AOR = 1.45; 95% CI: 1.12–1.87), those born to mothers aged 30–34 (AOR = 2.09; 95% CI: 1.25–3.48), and second-order births (AOR = 1.38; 95% CI: 1.10–1.74). The strong, graded relationship between household wealth and birth size aligns with the theory of social conditions as “fundamental causes” of health inequalities [8]. Wealth enables better maternal nutrition, reduces exposure to physical and psychological stress, and facilitates access to a broader range of health resources, which collectively contribute to a healthier intrauterine environment and optimal fetal growth [9, 10]. The fact that this association remained strong after adjusting for other factors underscores the pervasive impact of poverty on neonatal health in Somaliland. The significant role of maternal age and birth order is consistent with a vast body of global research. Our results, which indicate that adolescent mothers are at a higher risk of having smaller babies, align with studies from across sub-Saharan Africa [11–13]. Teenage pregnancy is often associated with biological immaturity and competition for nutrients between the growing mother and the fetus, alongside social disadvantages such as lower rates of healthcare utilization [11]. Conversely, the finding that second and higher-order births tend to be larger than first births is a well-documented physiological phenomenon, potentially related to improved uterine blood flow and maternal adaptation after the first pregnancy [1, 14]. These reproductive patterns are crucial targets for public health intervention. A particularly noteworthy finding from our study is the lack of a statistically significant association between ANC attendance and birth size in the final adjusted model, while maternal education showed a mixed effect. While initially appearing counterintuitive, this result offers essential insights. The non-significance of ANC attendance does not necessarily mean it is unimportant; instead, it may suggest that the quality and content of the care received, rather than mere attendance, are what truly impact birth weight [7, 15]. In resource-limited settings, ANC visits may be brief and often lack essential components, such as nutritional counseling or adequate screening for complications. Similarly, the benefits of maternal education may be mitigated by overriding contextual factors, such as extreme poverty, food insecurity, or cultural norms that limit a woman’s autonomy —a complexity noted in other fragile settings [6, 16, 17]. The implications are that policy must move beyond simply promoting access to services and focus on improving service quality and addressing the deep-seated structural barriers that prevent women from translating knowledge into better health outcomes. To further assess the robustness of our findings, we conducted a multinomial logistic regression as a sensitivity analysis, with ‘Average’ birth size as the reference category (see Supplementary Table S1). The results corroborated our primary findings from the ordinal model. For example, infants from the wealthiest households had significantly lower odds of being born small (AOR = 0.68; 95% CI: 0.49–0.94) and higher odds of being born large (AOR = 1.35; 95% CI: 1.03–1.77) compared to those born into the poorest households. This consistency across different analytical approaches reinforces the validity of our conclusions regarding the key determinants of birth size in Somaliland. Strengths and Limitations Strengths: Use of a nationally representative dataset; appropriate use of ordinal logistic regression for an ordered outcome; control for multiple confounders.Limitations: Cross-sectional design limits causal inference; birth size was based on maternal recall for a portion of the sample, which is subject to bias; data may not capture unmeasured confounders like maternal nutrition or infectious diseases during pregnancy, and potential interaction effects between predictors were not assessed. Strengths: Use of a nationally representative dataset; appropriate use of ordinal logistic regression for an ordered outcome; control for multiple confounders. Strengths: Use of a nationally representative dataset; appropriate use of ordinal logistic regression for an ordered outcome; control for multiple confounders. Limitations: Cross-sectional design limits causal inference; birth size was based on maternal recall for a portion of the sample, which is subject to bias; data may not capture unmeasured confounders like maternal nutrition or infectious diseases during pregnancy, and potential interaction effects between predictors were not assessed. Limitations: Cross-sectional design limits causal inference; birth size was based on maternal recall for a portion of the sample, which is subject to bias; data may not capture unmeasured confounders like maternal nutrition or infectious diseases during pregnancy, and potential interaction effects between predictors were not assessed. Policy recommendations Based on these findings: Develop socioeconomic support programs for pregnant women from poor households, such as cash transfers or food supplementation, to improve maternal nutrition.Strengthen reproductive health education and family planning services, with a focus on adolescent girls, to reduce the incidence of high-risk teenage pregnancies.Enhance the quality of ANC services to ensure they include comprehensive nutritional counseling and support, rather than just routine check-ups. Develop socioeconomic support programs for pregnant women from poor households, such as cash transfers or food supplementation, to improve maternal nutrition. Develop socioeconomic support programs for pregnant women from poor households, such as cash transfers or food supplementation, to improve maternal nutrition. Strengthen reproductive health education and family planning services, with a focus on adolescent girls, to reduce the incidence of high-risk teenage pregnancies. Strengthen reproductive health education and family planning services, with a focus on adolescent girls, to reduce the incidence of high-risk teenage pregnancies. Enhance the quality of ANC services to ensure they include comprehensive nutritional counseling and support, rather than just routine check-ups. Enhance the quality of ANC services to ensure they include comprehensive nutritional counseling and support, rather than just routine check-ups. Conclusion This study found that household wealth, maternal age, and birth order are significant predictors of newborn birth size in Somaliland. The findings underscore the critical importance of addressing socioeconomic inequities and strengthening maternal health programs, particularly for young and first-time mothers. Policymakers must design targeted interventions that provide economic support and enhance the quality of care to improve birth outcomes and ensure every child has a healthy start to life. Importantly, this study also highlights that simply promoting service utilization may be insufficient; efforts must also focus on improving the quality of maternal health services and addressing the underlying structural barriers that limit women’s autonomy and ability to act on health knowledge. Implementing these recommendations could contribute to improved maternal and child health outcomes in Somaliland. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.