Abstract
Background
Skin-to-skin contact (SSC) is recognized as an affordable and effective strategy to reduce neonatal mortality; however, its implementation in fragile health systems remains limited.
Objective
This study aimed to determine the prevalence of mother–newborn SSC and to identify its individual- and community-level predictors in Somalia.
Design
A nationally representative cross-sectional study.
Methods
A secondary analysis was conducted using data from the 2020 Somali Demographic and Health Survey involving 6,817 mothers aged 15–49 who had given birth to a live child in the five years preceding the survey. Multilevel binary logistic regression was used to identify predictors.
Results
Findings revealed that only 18.2% of women reported engaging in SSC after birth. Delivery in a health facility is significantly associated with the likelihood of SSC (aOR = 6.01; 95% CI: 5.12–7.05). Individual-level factors such as higher education, household wealth, maternal employment, and media exposure were positively associated with SSC practice, while divorced mothers had reduced odds. Community-level socioeconomic status also influenced SSC uptake, with mothers from wealthier communities being more likely to practice it.
Conclusion
SSC is practiced at a very low level with substantial disparities tied to social and regional factors. Improving neonatal survival requires prioritizing interventions that increase access to institutional deliveries, train healthcare providers in essential newborn care, and launch targeted public awareness campaigns to reach vulnerable populations.
Introduction
Skin-to-skin contact (SSC) is defined by the World Health Organization (WHO) as placing a naked newborn on the mother's chest immediately after birth (WHO, 2017). Originating from Kangaroo Mother Care in Bogotá, Colombia in the 1970s, SSC has since been recognized as a simple, low-cost intervention that improves neonatal survival (Charpak et al., 2005). Evidence consistently demonstrates that SSC stabilizes neonatal thermoregulation, cardiorespiratory function, glycemic control, enhances immunity, and significantly reduces neonatal mortality (Moore et al., 2016; Mose et al., 2021). For mothers, SSC promotes early initiation of breastfeeding, reduces physiological stress and pain, lowers the risk of postpartum depression, and strengthens maternal-infant bonding through oxytocin-mediated pathways (Karimi et al., 2020; Widström et al., 2019). Recent neurobiological research further demonstrates that tactile contact supports neurodevelopmental pathways critical for early life adaptation (La Rosa et al., 2024)
Despite these well-documented benefits, SSC implementation remains inconsistent globally. A systematic review reported prevalence ranging from less than 10% in some African nations to over 95% in parts of Europe, with consistently lower rates in low-and middle-income countries (Abdulghani et al., 2018). Within sub-Saharan Africa, SSC prevalence averages around 45%, though national estimates vary widely, reflecting differences in health system capacity and delivery practices (Aboagye et al., 2022a), for instance, prevalence has been reported as low as 12% in Nigeria (Ekholuenetale et al., 2021). Outside the African region, similarly low coverage has been documented in other low- and middle-income settings, including Bangladesh, underscoring that suboptimal SSC uptake is not confined to a single geographic context (Sharif et al., 2025).
Somalia represents an extreme case within this global landscape. The country faces one of the highest neonatal mortality rates worldwide, compounded by prolonged conflict, population displacement, weak health infrastructure, and low rates of institutional delivery (Hassan et al., 2026; Omar et al., 2026; Sheikh et al., 2024; UNICEF, 2025). In such settings, SSC has the potential to function not merely as a complementary practice, but as a critical survival strategy. Yet, to date, no nationally representative study has examined SSC prevalence or its determinants in Somalia. This lack of evidence limits policymakers’ and practitioners’ ability to integrate SSC effectively into maternal and newborn health programming.
Previous multilevel studies from Ethiopia, Nigeria, and Papua New Guinea demonstrate that SSC practice is shaped by both individual-level factors and broader contextual influences, including community norms and health system characteristics (Aboagye et al., 2022b; Debella et al., 2024; Ekholuenetale et al., 2021). However, Somalia's prolonged instability and fragmented service delivery suggest that contextual determinants may play an even more decisive role. This study addresses this gap by providing the first nationally representative multilevel analysis of SSC in Somalia, using Demographic and Health Survey data to estimate prevalence and identify individual- and community-level predictors. By situating SSC within the realities of a fragile health system, the study generates evidence directly relevant to strengthening essential newborn care and advancing progress toward Sustainable Development Goal 3.2 on neonatal mortality reduction (United Nations General Assembly, 2015; World Health Organization, 2014).
Literature Review
Mother–infant SSC is recognized globally as a vital, low-cost intervention for reducing neonatal mortality and fostering positive health outcomes, including stabilizing thermoregulation, enhancing early breastfeeding initiation, and strengthening maternal-infant bonding (Moore et al., 2016). Individual-level determinants shaping SSC practice are varied; for instance, higher maternal education has consistently been shown to be positively associated with the likelihood of SSC, as educated women may be more informed about its benefits (Girma et al., 2024; Sharif et al., 2025). Moreover, factors such as lower birth weight infants have been associated with increased SSC odds, potentially due to heightened awareness among mothers of vulnerable newborns (Aboagye et al., 2022b). In addition, prevailing cultural norms emphasizing modesty can present significant barriers, leading to mother–newborn separation even in settings where SSC is recommended by the World Health Organization (Ekholuenetale et al., 2020; Mose et al., 2021). Consequently, the consistency of SSC adoption is strongly influenced by these deeply rooted socio-cultural factors alongside individual maternal characteristics (Abdulghani et al., 2018; Karimi et al., 2020).
Examining the studies from fragile health systems and Sub-Saharan Africa reveals critical multilevel determinants that shape SSC practice. Facility-level factors, such as institutional deliveries versus home births, are major predictors, with public health facilities often demonstrating higher SSC rates due to staff training and established protocols (Aboagye et al., 2022a; Sharif et al., 2025). Conversely, delivery outside formal settings or via cesarean section is frequently associated with lower SSC practice (Aboagye et al., 2022a; Ekholuenetale et al., 2020). In addition, health system readiness, including provider training and exposure to health education via media or mobile phones, directly impacts compliance (Aboagye et al., 2022a; Ekholuenetale et al., 2021). Studies in fragile settings like Somalia suggest that implementation remains limited despite the recognized benefits, likely compounded by weak infrastructure (Hassan et al., 2026). This heterogeneous landscape, evidenced by varying prevalence across regions like East Africa, highlights variability in coverage. Therefore, the current knowledge gap regarding Somalia's specific context justifies a nationally representative multilevel analysis to precisely map the key individual- and community-level factors that drive or inhibit SSC practice, a critical step for informing essential newborn care programming.
Methods
Study Design and Data Source
This study used a cross-sectional design and conducted a secondary analysis of data from the 2020 national Demographic and Health Survey. This DHS is a nationally representative population-based survey that provides reliable data on a wide range of health, demographic, and nutritional indicators. The survey employed a two-stage stratified cluster-sampling methodology to ensure a representative sample of the population. The 2020 DHS sample included 380 enumeration areas (clusters) distributed across all 16 regions, with cluster sizes ranging from 8 to 35 households. In the first stage, the enumeration areas (clusters) are selected with a probability proportional to their size. In the second stage, a fixed number of households was systematically selected from each cluster. Trained interviewers collected the data using standardized questionnaires. The full methodological details of the survey are available in the final DHS report. The data were publicly available for research purposes on the DHS Program website. This analysis focused on a weighted sample of 6,817 women aged 15–49 who had given birth to a live child in the five years preceding the survey. The study period was from Feb. 2018 to Jan. 2019 (Hassan et al., 2026).
Study Variables
Outcome Variable
The primary outcome of this study was mother–infant SSC. This practice is defined as placing a naked newborn on the mother's chest immediately after delivery (WHO, 2017). In the DHS survey, this was assessed based on the mothers’ responses to questions about their most recent birth. Following the operationalization used in similar studies (Aboagye et al., 2022b; Sharif et al., 2025). The outcome was coded as a binary variable: ‘1’ if the mother reported that her baby was placed on her bare chest immediately after birth, and ‘0’ otherwise.
Explanatory Variables
The selection of explanatory variables was guided by a comprehensive review of existing literature on the determinants of newborn care practices and maternal health service utilization (Aboagye et al., 2022b; Ekholuenetale et al., 2021; Tawfiq et al., 2025). To align with the multilevel statistical approach, these variables were categorized as either individual- or community-level.
The individual-level factors were mother's age in years (15–19, 20–24, 25–29, 30–34, 35–49), maternal education (no education, primary, secondary/higher), household wealth index (categorized as poor, middle, rich), marital status (married, divorced/widowed), maternal employment (worked in last 12 months: yes/no), exposure to mass media, place of delivery (home, health facility), sex of the child (male, female), child's size at birth (large, normal, small), birth type (single or multiple), and birth order (1st, 2nd-3rd, 4th and above).
Community-level factors derived from the survey's cluster data included place of residence (urban, rural, or nomadic), region, community socioeconomic status, and community-level education. Community-level socioeconomic variables were derived by aggregating individual responses within each primary sampling unit (PSU). Household wealth (V190) was recoded into poor (lowest two quintiles), middle (third quintile), and rich (highest two quintiles), while maternal education (V106) was categorized as no education, primary, or secondary/higher. Weighted cluster-level summaries were calculated to account for the DHS sampling design. The resulting continuous measures were categorized into tertiles (low, moderate, and high).
Statistical Analysis
All statistical analyses were performed using Stata/SE version 17.0 (StataCorp, College Station, TX, USA). All calculations were weighted with the sampling weights provided in the dataset to account for the complex survey design and to ensure national representativeness. Initially, descriptive statistics (weighted frequencies and percentages) were used to summarize the background characteristics of the study population. Subsequently, survey-adjusted chi-square tests were employed to examine the bivariate association between each explanatory variable and the practice of SSC. To identify independent predictors of SSC while accounting for the hierarchical data structure, a multivariable multilevel binary logistic regression model was fitted. All explanatory variables from the bivariate analysis with p-values less than .25 were included as candidates in the multivariable model to prevent premature exclusion of potentially important confounders (Mickey & Greenland, 1989). Listwise deletion was employed for handling missing data, with incomplete observations excluded from the analysis, ensuring that only complete cases were considered for the final models. Multicollinearity among predictor variables was assessed using variance inflation factors (VIF). All VIF values were below 2.92 (mean VIF = 1.49), well below the commonly used threshold of 10, indicating that multicollinearity did not significantly affect the stability of the regression estimates (O’brien, 2007). The final results were presented as adjusted odds ratios (aORs) with their corresponding 95% confidence intervals (CIs). The intraclass correlation coefficient (ICC) was calculated using an empty, unconditional model to quantify the proportion of the total variance in SSC attributable to community-level clustering. Statistical significance for the final multivariable model was set at p < .05.
Results
Prevalence of Mother and Infant Skin-to-Skin Contact
The overall prevalence of mother–infant SSC in the sample was 18.2% (95% CI: 16.1–20.5%).
Distribution of SSC Across Explanatory Variables
Table 1 shows the prevalence of SSC across different individual and community characteristics. The practice of SSC was significantly more common among mothers with primary (27.8%), secondary (34.1%), or higher (41.4%) education compared to those with no education (15.8%). A strong positive association with wealth was observed, with SSC prevalence rising from 9.3% among the poorest mothers to 27.7% among the wealthiest mothers.
Prevalence and Distribution of Mother and Newborn Skin-to-Skin Contact Across Explanatory Variables.
Mothers who delivered at a health facility had a substantially higher prevalence of SSC (44.6%) than those who gave birth at home (9.4%). Exposure to media was also a significant factor, with 30.4% of mothers with media exposure practising SSC, compared to only 16.1% of those without. Substantial regional variations were observed at the community level, with SSC practices ranging from 3.3% in Gedo to 37.6% in Togdheer. Both community socioeconomic status and community educational levels were strongly and positively linked to SSC practices.
Predictors of Mother and Newborn Skin-to-Skin Contact
Table 2 presents the results of the mixed effects logistic regression model. Finally, in the full model (Model III), which controlled individual and community factors simultaneously, several variables remained significant predictors of SSC.
Fixed Effect Estimates from Multilevel Logistic Regression Models Predicting SSC.
*p < .05, **p < .01, ***p < .001. aOR, adjusted odds ratio; CI, confidence interval.
Among the individual-level factors, the strongest predictor was the place of delivery; mothers who delivered at a health facility had six times the odds of practising SSC compared to those who delivered at home (aOR = 6.01; 95% CI: 5.12–7.05). Sociodemographic characteristics were also significant. Mothers with secondary education were 1.41 times more likely to practise SSC than those with no education (aOR = 1.41; 95% CI: 1.03–1.92). Compared with the poorest mothers, those in the middle (aOR = 1.46; 95% CI: 1.13–1.89) and the richest (aOR = 1.37; 95% CI: 1.05–1.79) wealth quintiles had higher odds of practising SSC. Conversely, divorced mothers had approximately 28% lower odds of practising SSC than married mothers (aOR = 0.72; 95% CI: 0.55–0.94). Similarly, mothers who had not worked in the past year had significantly lower odds of practising SSC than those who had (aOR = 0.54; 95% CI: 0.32–0.92). Regarding media exposure, mothers with no media exposure had 18% lower odds of practising SSC than those with media exposure (aOR = 0.82; 95% CI: 0.68–0.98).
At the community level, mothers living in communities with moderate socioeconomic status had 1.65 times the odds of practising SSC compared to those in communities with low socioeconomic status (aOR = 1.65; 95% CI: 1.08–2.50). However, significant regional variation persisted. Compared with the Awdal region, mothers in Togdheer (aOR = 1.80) and Nugaal (aOR = 2.09) had significantly higher odds of practising SSC, whereas those in Banadir (aOR = 0.66), Gedo (aOR = 0.31), and Lower Juba (aOR = 0.47) had significantly lower odds.
Random Effects (Measures of Variation)
The results of the random-effects model are shown in Table 3. The null model (Model 0), which included no explanatory variables, demonstrated significant variation in the odds of practising SSC across communities (σ2 = 1.39, p < .001). The intraclass correlation coefficient (ICC) derived from this model was 29.7%, indicating that almost 30% of the total variance in SSC practices was due to differences between communities.
Random Effect Estimates and Model Fit Statistics.
After including individual-level factors (Model I) and community-level factors (Model II), the community-level variance was notably reduced. In the final model (Model III), the ICC dropped to 5.9%, indicating that the combination of individual and community variables explained a significant portion of the initial community-level variation. The consistent decline in the Akaike information criterion (AIC) across the models, with Model III having the lowest value (5242.32), confirms it as the best-fitting model for data.
Discussion
This study aimed to determine the prevalence of mother–newborn SSC and identify its associated multilevel predictors. The analysis revealed that the overall prevalence of SSC is extremely low (18.2%). This finding is significantly lower than the pooled prevalence of 48% reported in a meta-analysis from neighboring Ethiopia (Debella et al., 2024) and rates in countries such as Papua New Guinea (45.2%) (Aboagye et al., 2022b). Nevertheless, this study's prevalence is slightly higher than the rates observed in other low-resource settings, such as Nigeria (12.0%) (Ekholuenetale et al., 2021) and Bangladesh (15.6%) (Sharif et al., 2025). These findings suggest that conflict-affected contexts face particularly severe implementation challenges
This finding provides updated evidence on the influence of maternal education on newborn care in Somalia, revealing a strong positive association between higher levels of education and SSC. This finding is consistent with the extensive body of research conducted in diverse settings, including Afghanistan, Ethiopia, and Nigeria (Ekholuenetale et al., 2021; Mose et al., 2021; Tawfiq et al., 2025). Educated mothers typically possess greater health literacy, demonstrate more autonomy in health-related decision-making, and are more likely to utilize formal healthcare services where they can receive information on best practices for newborn care (Kassie et al., 2025; Neves et al., 2021).
This study showed a significant dose-response relationship between household wealth and SSC practice, with mothers from the richest quintiles having markedly higher odds of practicing SSC. This finding is consistent with those of research conducted in sub-Saharan Africa and Papua New Guinea (Aboagye et al., 2022a, 2022b). The strong association observed in Somalia likely reflects the profound influence of economic status on healthcare access. Families with greater financial resources are better positioned to overcome geographic and financial barriers to health facilities, are more likely to be educated, and have increased exposure to health-related media (Wuneh et al., 2019). This economic gradient reveals a significant equity gap: the most vulnerable newborns from the poorest households are least likely to receive protective interventions such as SSC. These disparities are compounded by limited maternal autonomy, lower health literacy, and reduced access to facilities staffed with trained personnel. Addressing this gap requires targeted public health strategies that prioritize outreach and support for disadvantaged communities.
The findings revealed that mothers who were not employed had significantly lower odds of practicing SSC. This finding aligns with research indicating that maternal employment can act as a proxy for greater financial independence, social mobility, and exposure to information and networks outside the immediate household (Sharif et al., 2025). It differs, however, from contexts where non-working mothers might be assumed to have more time for childcare practices. In this context, employed women may have greater autonomy to seek facility-based care or be more exposed to modern health messaging, which may contribute to higher odds of SSC practice (Dickson et al., 2021). The low employment rate reflects prevailing cultural norms, where women predominantly manage household and childcare duties.
This study provides updated evidence on the role of media in health behavior in Somalia, revealing that mothers exposed to media are significantly more likely to practice SSC. This finding strongly aligns with studies from Nigeria and The Gambia, which underscore the vital role of the mass media in disseminating public health information (Ekholuenetale et al., 2021). In contexts such as Somalia, where literacy rates are low and physical access to healthcare professionals is limited, radio can be a particularly effective and far-reaching tool for delivering essential health messages about newborn care, including the benefits of SSC (Aboagye et al., 2022c).
This study identified the place of delivery as the most potent predictor of SSC, with mothers delivering in a health facility being six times more likely to practice it. This finding is consistent with the overwhelming body of evidence from Bangladesh, Afghanistan, and other African countries (Ali et al., 2021; Mose et al., 2021; Sharif et al., 2025). The significant difference is likely to reflect that facility births are attended by skilled health personnel trained in evidence-based newborn care practices, including SSC. In contrast, home births in the study context are typically attended by traditional birth attendants or family members who may not have received formal training in immediate postpartum care protocols (Atiqzai et al., 2019; Schueller et al., 2022). In Somalia's underserved regions, births are mainly attended by traditional birth attendants rather than skilled health professionals, which further restrict opportunities for SSC and highlight the structural barriers that must be addressed alongside efforts to promote institutional delivery.
This study found that divorced mothers had lower odds of practising SSC than married mothers. This aligns with the existing literature emphasizing the social and economic vulnerability of divorced women, who often face reduced family support, financial hardship, and psychological distress (Arnold et al., 2015). These stressors can create a challenging environment for mother–infant bonding.
This study confirmed the existence of significant regional- and community-level influences on SSC. The finding that practice varies widely by region, with rates in Togdheer being over ten times higher than those in Gedo, is consistent with reports of deep-seated inequalities in healthcare infrastructure, security, and humanitarian access across Somalia (Lydon et al., 2021). The observed regional disparities in SSC (ranging from 3.3% in Gedo to 37.6% in Togdheer) likely reflect the uneven impact of conflict across Somalia. Regions such as Gedo and Lower Juba, which have experienced prolonged insecurity and displacement, show the lowest SSC rates, potentially due to destroyed health infrastructure, healthcare worker displacement, and limited humanitarian access. Conversely, relatively stable regions such as Togdheer and Nugaal, where health facilities have maintained operations and received more consistent international support, demonstrate higher SSC rates. Similarly, the strong association between community socioeconomic status and SSC aligns with research showing that community-level resources, collective social norms, and better access to services create an enabling environment for positive health behaviors (Girma et al., 2024).
Implications for Practice
The unacceptably low overall prevalence of mother–infant SSC in Somalia mandates an urgent, multi-pronged clinical and health system response focused on strengthening the points of care where intervention is most effective. The study's identification of institutional delivery as the strongest predictor (aOR = 6.01) signals that improving access to and uptake of facility-based births is paramount. Consequently, healthcare providers must be mandated and continuously trained in essential newborn care protocols to ensure SSC is an uninterrupted standard practice immediately following birth, regardless of the delivery mode. Moreover, since maternal education, household wealth, employment, and media exposure were positively associated with SSC, health literacy interventions must be integrated into both antenatal and postnatal care to equalize knowledge and improve the utilization of evidence-based newborn practices.
These findings emphasize the essential role of nurses and midwives in institutionalizing SSC during the “golden hour.” Nursing leadership should establish clinical guidelines and documentation protocols to mandate SSC as a standard, auditable postpartum practice. Furthermore, incorporating SSC competencies into Somalia's nursing curricula and in-service training is vital. As trusted health educators, nurses can bridge literacy gaps through routine counseling, while nursing-led research into facility-specific barriers will further strengthen targeted quality improvement initiatives.
Addressing the significant equity gaps, particularly the reduced odds for divorced mothers and the influence of community socioeconomic status and regional disparities, requires targeted community-level strategies. In addition, given the positive link between media exposure and SSC practice, public health campaigns should leverage locally trusted mass media, such as radio, to deliver essential newborn care messages to women with lower education or those in socioeconomically disadvantaged areas. This layered approach, combining robust systemic improvements within facilities with targeted community education that addresses vulnerabilities in wealth, education, and regional access, is essential. By implementing these practice recommendations, Somalia can effectively enhance the uptake of this life-saving intervention, thereby making direct progress toward achieving Sustainable Development Goal 3.2 on reducing neonatal mortality.
Strengths and Limitations
The primary strength of this study lies in its use of data from the Demographic and Health Survey. This large, nationally representative dataset enables the findings to be generalized to the broader Somali population. The application of multilevel modelling is another key strength as it effectively accounts for the hierarchical nature of the data and provides more robust estimates by considering both individual- and community-level influences.
However, this study has several limitations. First, its cross-sectional design precludes causal inferences. Second, the SSC data were based on maternal self-reports, which may be subject to recall bias, particularly for births further in the past. Third, the DHS does not capture critical potential predictors such as specific cultural beliefs about newborn care, maternal mental health status, or healthcare provider attitudes, which could significantly influence SSC practice.
Conclusion
The practice of mother–infant skin-to-skin contact in Somalia is unacceptably low, leaving the majority of newborns without lifesaving evidence-based interventions. The key determinants, maternal education, wealth, media exposure, place of delivery, and region, highlight profound inequities in access to health and knowledge. To improve neonatal survival in Somalia, a multi-pronged approach is urgently needed to strengthen health systems to ensure universal access to institutional delivery, train healthcare providers, and implement targeted community-based health education. Future research should employ qualitative methods to explore the specific sociocultural barriers and enablers of SSC in the Somali context, thereby providing deeper insights into more effective, culturally sensitive interventions.
Footnotes
Acknowledgments
The authors would like to acknowledge the Demographic and Health Survey (DHS) Program and the Somalia National Bureau of Statistics (SNBS) for providing access to the 2020 Somali Health and Demographic Survey dataset. We also thank the field staff and participants of the SHDS 2020 for their contribution to the primary data collection.
Ethics Approval and Consent to Participate
The study was conducted through a secondary analysis of the 2020 Somalia Health and Demographic Survey (SHDS). The primary survey protocols received ethical clearance from the Somali Federal Ministry of Health and relevant Institutional Review Boards. Access to the de-identified raw data was formally authorized by the DHS program following the submission of a research proposal. Participant confidentiality was strictly maintained, as the dataset was anonymized prior to access, ensuring that no individual or household identifiers were visible. Consequently, the study adhered to international ethical standards for secondary data utilization, requiring no further primary consent from respondents.
Authors’ Contributions
YSAH: Conceptualization, study design, data analysis, and manuscript drafting. MHM: Data interpretation and critical review of the manuscript. ANO: Critical review of the manuscript and intellectual contribution. ASA: Data interpretation, manuscript review, and validation. All authors have read and approved the final manuscript for publication.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Availability of Data and Materials
Use of AI Software
The authors did not use AI software in the preparation of this manuscript.