Somalian mothers’ perspectives on breast milk while their infants are cared for in the neonatal intensive care unit: a qualitative study Authors: Derya Kılınç, Remziye Semerci Şahin, Siham Abdikarim Omar, Fatma Hacıoglu Published: 2026 ABSTRACT BackgroundBreastfeeding is essential for neonatal survival and development, yet it is often disrupted when infants are hospitalized in neonatal intensive care units (NICU). In low-resource settings, structural barriers such as limited access to lactation support, equipment, and storage facilities further complicate the maintenance of breastfeeding. Despite its importance, evidence on mothers’ experiences and challenges in sustaining breast milk provision in these contexts remains limited. The aim of this study was to explore the lived experiences of Somali mothers regarding breast milk while their infants were hospitalized in the NICU.MethodA descriptive phenomenological design was used. Data were collected through individual, semi-structured, in-depth interviews with 12 mothers whose infants were receiving care in the NICU of a tertiary hospital in Mogadishu, Somalia. Participants were selected using purposive sampling. Interviews were audio-recorded, transcribed verbatim, and analyzed using thematic content analysis following Thomas and Harden’s approach. The study was conducted in accordance with the COREQ reporting guidelines.ResultsFour main themes were identified: (1) Breast Milk as an Absolute Good, but a Vague Knowledge Domain; (2) Milk Production Shaped More by Living Conditions Than by the Maternal Body; (3) Breastfeeding Knowledge Learned Through Trial and Error Rather Than Intergenerational Transfer; and (4) Motherhood as a Well-Intentioned but Poorly Supported Effort. Mothers viewed breast milk as beneficial but had limited biomedical knowledge and low risk perception. Breastfeeding practices were shaped by socioeconomic barriers, with pumping mainly during hospital visits and knowledge largely acquired through trial and error despite limited support.ConclusionBreastfeeding during NICU hospitalization in Somalia is shaped more by structural and health-system constraints than by maternal motivation or physiological capacity. Positive attitudes toward breast milk alone are insufficient to sustain lactation without consistent professional guidance and basic infrastructure support. NICU nurses should deliver structured, skills-based lactation counseling on pumping, milk handling, and transport, and ensure access to practical resources to minimize milk loss and support breastfeeding continuity in low-resource settings. Introduction Breast milk is the most fundamental source of nutrition, supporting healthy growth and development in infants [1–5], protecting against illness, and providing antibodies that strengthen the immune system [2]. By optimally meeting both the biological and psychological needs of newborns, breast milk contains numerous bioactive components that protect against infection and inflammation while supporting immune maturation, organ development, and the establishment of a healthy microbiome [2, 6]. The World Health Organization (WHO) and United Nations Children’s Fund (UNICEF) recommend initiating breastfeeding within the first hour after birth, exclusive breastfeeding for the first 6 months, and continued breastfeeding with complementary feeding up to 2 years of age [6–7]. However, according to 2023 global data, only 46% of infants are exclusively breastfed during the first five months of life, and only 48% initiate breastfeeding within the first hour after birth, indicating substantial gaps between recommendations and actual practices worldwide [6–8]. In Somalia, breastfeeding rates vary widely across regions, ranging from as low as 5.2% to as high as 68%, reflecting major inequities in access to maternal health education, healthcare services, and social support systems [9]. Breastfeeding is widely recognized as the optimal feeding method, providing both physical and psychological benefits to the mother–infant dyad and contributing to maternal well-being and infant survival [3]. Nevertheless, many newborns require admission to NICU due to prematurity, low birth weight, perinatal infections, or congenital anomalies, which significantly disrupts early breastfeeding practices [10]. Separation of mothers and infants in NICU settings exposes mothers to emotional stress, physical exhaustion, and logistical challenges related to milk expression and transport, which can negatively affect lactation and breastfeeding continuation [5]. This separation may suppress milk production and reduce breastfeeding rates, despite breast milk being considered a life-saving medical intervention for hospitalized newborns, particularly through its protective effects against necrotizing enterocolitis (NEC) and its role in promoting maternal–infant bonding [11–12]. Early initiation of milk expression has been shown to enhance milk supply and establish a strong foundation for later breastfeeding, making sustained pumping during NICU hospitalization a critical component of neonatal care [13–14]. Therefore, mothers must be supported in acquiring appropriate knowledge and skills related to milk expression techniques, safe storage conditions, and proper transport procedures to ensure continued access to breast milk for their hospitalized infants [13–14]. Available clinical data from tertiary-level neonatal intensive care units in Mogadishu indicate that prematurity represents the most common reason for NICU admission (27.0%), followed by birth asphyxia (18.0%), neonatal sepsis (14.6%), and acute respiratory distress syndrome (12.2%) [10]. In addition, a substantial proportion of admitted neonates are born preterm or with very low birth weight (< 1500 g), both of which are strongly associated with increased neonatal mortality risk in the region. These findings highlight the considerable burden of high-risk neonatal conditions requiring intensive care in Somalia and underline the critical importance of optimal nutritional support, including access to breast milk, for hospitalized newborns in this setting [10]. Despite the recognized importance of maternal experiences and perceptions in sustaining breastfeeding during NICU hospitalization, most existing studies are derived from high-income countries and healthcare systems with relatively well-established lactation support infrastructures [4–6]. In contrast, low-resource and conflict-affected settings such as Somalia face unique barriers including limited healthcare access, shortages of trained health professionals, sociocultural beliefs surrounding breastfeeding, and structural challenges in maternal–child health services, all of which may significantly shape mothers’ breastfeeding experiences [5–9]. However, qualitative evidence exploring how mothers in Somalia perceive breast milk and navigate breastfeeding during NICU hospitalization remains extremely limited in the existing literature, representing a critical gap in global maternal and neonatal health research [4–9]. Understanding mothers’ perspectives in this context is essential for designing culturally appropriate, feasible, and effective breastfeeding support interventions that are responsive to both clinical and social realities in fragile health systems [6–7]. By addressing this gap, the present study contributes context-specific insights that can inform nursing practices, policy development, and targeted maternal support strategies in low-resource neonatal care settings [5–6]. Methods Aim The aim of this study was to explore the lived experiences, perceptions, and breastfeeding practices of Somali mothers regarding breast milk while their infants were hospitalized in the NICU. Study design A descriptive-phenomenological design was employed to explore mothers’ lived experiences and perceptions of breast milk in the neonatal intensive care context. This approach aims to describe and interpret the essence of participants’ experiences as they perceive and make meaning of a particular phenomenon. The study was theoretically informed by Husserl’s descriptive phenomenology, which emphasizes the exploration of lived human experience and the principle of “returning to the things themselves” in order to understand how individuals perceive and make meaning of their life-world [15]. This approach aims to describe experiences as expressed by participants, without imposing preconceived interpretations. It enables the generation of rich descriptive data and facilitates an in-depth understanding of participants’ perspectives, particularly in complex and sensitive clinical environments [16]. Data were collected through individual, in-depth, semi-structured interviews. The study was conducted in accordance with the Consolidated Criteria for Reporting Qualitative Research (COREQ) to ensure methodological rigor and transparent reporting [17]. Ethical considerations Ethical approval for the study was obtained from the relevant institutional ethics committee and the hospital administration (Decision No: MSTH/177–940; Date: 24.03.2023), and the study was conducted in accordance with the principles of the Declaration of Helsinki. All participants were fully informed that participation in the study was voluntary and that they could withdraw at any time without consequences. Written and verbal informed consent was obtained from all participants prior to participation. Confidentiality and privacy were strictly maintained by anonymizing all data and securely storing all research materials, and no personally identifiable information was used at any stage of the study. Informed consent was obtained from all participants, and no financial support was received for the research, authorship, and/or publication of this study. Clinical trial number: not applicable. Study setting and participants The study was conducted in the NICU of a training and research hospital located in Mogadishu, Somalia, between March 24 and April 24, 2023. Participants were recruited through purposive sampling to identify mothers who met the predefined inclusion criteria and could provide rich, relevant experiential data related to the research objectives. Eligible participants were mothers of infants hospitalized in the NICU for at least 24 h, who voluntarily agreed to participate in the study, were not receiving psychiatric or psychological treatment, and had no medical contraindications to breast milk expression. Mothers who declined to provide informed consent or who chose to withdraw from the study at any stage of the research process were excluded. Mothers who visited their infants in the NICU between 08:00 and 16:00 (East Africa Time, UTC + 3) were approached by a researcher, informed about the study, and eligible participants were provided with detailed information and invited to participate voluntarily. The target sample size was guided by the principle of data saturation, as recommended in qualitative research methodology. Consistent with Morse’s (1995) assertion that adequacy and appropriateness are the primary determinants of sample size in qualitative studies, the initial target sample size was set at 9–17 participants [18]. This range was informed by empirical evidence indicating that data saturation in qualitative research is frequently achieved within 9–17 interviews when samples are relatively homogeneous, and research aims are narrowly defined [19]. However, data collection continued until data saturation was reached, operationally defined as the point at which no new themes or substantive insights emerged in consecutive interviews. In the present study, data saturation was achieved after interviews with 12 participants, at which point further data collection was deemed unnecessary. Data collection tools Two data collection tools were used in this study: a Sociodemographic Information Form and a Semi-Structured Interview Guide. Sociodemographic information form The researchers developed the form, which consisted of 13 items to capture participants’ sociodemographic and obstetric characteristics. The form included questions regarding maternal age, educational level, number of children, number of deceased children, infant sex, previous breastfeeding experience, and mode of pregnancy, among other variables. This form enabled the contextual interpretation of mothers’ perceptions and experiences regarding breastfeeding. The structured data collection form In line with the descriptive phenomenological design, data were collected using broad open-ended questions rather than a strictly structured interview guide. The initial questions and prompts were developed by the research team in alignment with the study objectives to facilitate discussion of mothers’ perceptions and lived experiences regarding breast milk while their infants were hospitalized in the NICU [16–19–20]. These questions served as flexible starting points for conversation, allowing participants to freely describe their experiences in their own words. The questions were pilot-tested with three mothers who were not included in the main study sample to ensure clarity and relevance. No revisions were required following the pilot interviews. The interviews were initiated with broad questions such as. What does breastfeeding mean to you, and what are your overall views about breastfeeding?What benefits do you believe breast milk provides for the baby and for the mother?Do you have any concerns or perceived harms related to breast milk?In your view, what should mothers do regarding expressing, storing, and transporting breast milk when their infant is hospitalized in the NICU?Is there anything else you would like to share about your experiences or views? What does breastfeeding mean to you, and what are your overall views about breastfeeding? What does breastfeeding mean to you, and what are your overall views about breastfeeding? What benefits do you believe breast milk provides for the baby and for the mother? What benefits do you believe breast milk provides for the baby and for the mother? Do you have any concerns or perceived harms related to breast milk? Do you have any concerns or perceived harms related to breast milk? In your view, what should mothers do regarding expressing, storing, and transporting breast milk when their infant is hospitalized in the NICU? In your view, what should mothers do regarding expressing, storing, and transporting breast milk when their infant is hospitalized in the NICU? Is there anything else you would like to share about your experiences or views? Is there anything else you would like to share about your experiences or views? Follow-up prompts and probing questions were used as needed to encourage participants to elaborate on their experiences and to clarify emerging meanings during the interviews. Data collection procedure Mothers of infants hospitalized in the NICU who met the inclusion criteria were approached during their unit visits. Eligible participants were informed about the study, and written and verbal informed consent was obtained prior to participation. Data were collected through face-to-face, individual, semi-structured in-depth interviews. Interviews were conducted in a quiet and private area within the NICU to ensure participant comfort, confidentiality, and an environment conducive to open expression. Each participant was interviewed once during the study. Each interview lasted approximately 20–40 min. The interviews were moderated by a researcher fluent in Somali and trained in qualitative research methods. A second researcher, trained in qualitative research methods, observed the interviews, took field notes, and contributed probing questions to the primary interviewer when necessary. The second researcher was not fluent in Somali; therefore, all probing questions were communicated via the primary interviewer to ensure accurate interpretation and continuity of the interview process. No prior relationship existed between the researchers and the participants before the commencement of the study. All interviews were audio-recorded with participants’ permission and subsequently transcribed verbatim. To ensure anonymity, each participant was assigned a unique identification code (e.g., M1, M2, M3). Data collection continued until data saturation was achieved, defined as the point at which no new themes or meaningful insights emerged in consecutive interviews. Data analysis This study was underpinned by a descriptive-phenomenological approach, which was used to explore and describe mothers’ lived experiences and perceptions of breast milk while their infants were hospitalized in the NICU. Rather than testing a formal theory, the study was guided by phenomenological assumptions that prioritize participants’ subjective meanings and everyday experiences of the phenomenon under investigation. Although the study was conceptually informed by descriptive phenomenology, the data were analyzed using thematic content analysis to systematically identify patterns and themes in participants’ narratives. Data were analyzed using qualitative thematic content analysis, employing a systematic and iterative approach to organize and interpret textual data. The analytic process was informed by the eight-step framework described by Downe-Wamboldt (1992), which provides a structured method for reducing and interpreting qualitative data [21]. This approach facilitates the identification of patterns, themes, and categories that meaningfully reflect participants’ experiences, while allowing for analytic flexibility and depth [22]. Thus, the phenomenological orientation informed the overall purpose of understanding lived experience, whereas thematic content analysis provided the analytic structure for coding, organizing, and interpreting the interview data. In the present study, the analysis aimed to capture mothers’ perceptions and lived experiences related to breast milk while their infants were hospitalized in the Neonatal Intensive Care Unit (NICU). To ensure analytic rigor and transparency, the analysis was conducted systematically and in a cyclical manner, organized into 8 sequential steps outlined in Fig. 1. Fig. 1Eight-step thematic content analysis process Eight-step thematic content analysis process Qualitative content analysis was conducted following an eight-step analytic process, as illustrated in Fig. 1. Interviews were transcribed verbatim and repeatedly reviewed by the researchers to achieve immersion in the data and develop an overall understanding of participants’ experiences. Meaningful words, phrases, and statements were identified as units of analysis and organized into codes, sub-themes, and themes through an iterative coding process. Two researchers independently coded the transcripts, and discrepancies were discussed until consensus was reached to enhance analytic credibility. Categories and themes were continuously refined through comparison and discussion among the research team to ensure conceptual clarity and consistency. All transcripts were managed and analyzed using MAXQDA qualitative data analysis software (2020 version). The final themes were structured to comprehensively reflect mothers’ perceptions and lived experiences of breast milk in the NICU. Detailed steps of the analytic procedure are presented in Fig. 1 to avoid redundancy in the text. Interviews were conducted by DK and SAO. DK is a pediatric nursing academic with experience in qualitative research and maternal–child health, while SAO works at a tertiary-level hospital in Mogadishu and is familiar with the local healthcare context and cultural environment. SAO was fluent in Somali and facilitated communication with participants, whereas DK contributed methodological expertise in qualitative interviewing and research design. The research team had no prior relationship with the participants before the study, and reflexive discussions were conducted during the analytic process to minimize potential researcher bias and ensure that interpretations remained grounded in participants’ accounts. Methodological rigor The trustworthiness of the study was ensured by applying the criteria of credibility, transferability, dependability, and confirmability, as proposed by Lincoln and Guba (1985). Credibility was enhanced through prolonged engagement with participants, the use of open-ended and non-directive interview questions, independent coding by multiple researchers, and member checking. Member checking was conducted with a small subset of participants who were available and willing to be contacted after the interviews. These participants were provided with a brief summary of the preliminary themes and were invited to comment on whether the interpretations resonated with their experiences. This process was conducted through follow-up communication rather than a second formal interview. Participants were not asked to validate the “accuracy” of the analysis but rather to reflect on whether the interpretations were understandable and meaningful in relation to their experiences. No participants requested major changes or additions to the interpretations. Transferability was supported through the provision of rich, detailed descriptions of the research context, participant characteristics, data collection procedures, and analytic process, enabling readers to assess the applicability of the findings to similar settings or populations. Dependability and confirmability were strengthened by maintaining an audit trail throughout the research process, including documentation of analytic decisions, coding revisions, and reflective notes. Additionally, peer debriefing was conducted within the research team to discuss emerging themes and resolve discrepancies. All stages of data collection and analysis were transparently reported to allow for external evaluation of the study’s methodological consistency, rigor, and objectivity. Results Characteristics of the participants A total of 12 mothers participated in the study. Most mothers (n = 9, 75%) reported bringing breast milk to the hospital once daily, while 25% brought it twice daily. All participants (n = 12, 100%) had spontaneous pregnancies, and no parental consanguinity was reported. A history of infant loss was present in 41.7% of the mothers, whereas 25% reported a history of preterm birth. Regarding sociodemographic characteristics, 83.4% of the mothers were housewives and 16.6% were physicians. Educational levels varied: 33.6% had completed primary school, 16.6% had madrasa education, 16.6% had completed secondary school, and 16.6% had completed high school or held a bachelor’s degree. The mean age of the mothers was 26.70 ± 3.12 years (range: 23–30). The mean number of pregnancies and deliveries was 3.90 ± 1.85, and the average breastfeeding duration was 2.35 ± 1.50 months (See Table 1). Table 1Sociodemographic and obstetric characteristics of mothers (n = 12)Variable n % Frequency of bringing breast milk to hospital Once daily975%Twice daily325% Type of pregnancy Spontaneous12100% Parental consanguinity Absent12100% History of infant loss Yes541.6%No758.4% History of preterm birth Yes325%No975% Occupation Housewife1083.4%Physician216.6% Education level Madrasa216.6%Primary school433.6%Secondary school216.6%High school216.6%Bachelor’s degree216.6% Variables Mean ± Sd Min-max Age (years)26.70 ± 3.1223–30 yearsNumber of children4.30 ± 1.941–7Number of pregnancy3.90 ± 1.851–7Number of deliver3.90 ± 1.851–7Number of live children3.80 ± 1.471–6Number of deceased children1.40 ± 0.511–2Average of breastfeeding2.35 ± 1.501–24 month Table 1Sociodemographic and obstetric characteristics of mothers (n = 12)Variable n % Frequency of bringing breast milk to hospital Once daily975%Twice daily325% Type of pregnancy Spontaneous12100% Parental consanguinity Absent12100% History of infant loss Yes541.6%No758.4% History of preterm birth Yes325%No975% Occupation Housewife1083.4%Physician216.6% Education level Madrasa216.6%Primary school433.6%Secondary school216.6%High school216.6%Bachelor’s degree216.6% Variables Mean ± Sd Min-max Age (years)26.70 ± 3.1223–30 yearsNumber of children4.30 ± 1.941–7Number of pregnancy3.90 ± 1.851–7Number of deliver3.90 ± 1.851–7Number of live children3.80 ± 1.471–6Number of deceased children1.40 ± 0.511–2Average of breastfeeding2.35 ± 1.501–24 month Sociodemographic and obstetric characteristics of mothers (n = 12) Analysis of interviews with 12 mothers identified four main themes and seven subthemes reflecting mothers’ perceptions of breast milk, breastfeeding practices, and the structural conditions shaping lactation during neonatal hospitalization. These themes captured mothers’ understanding of the benefits and perceived risks of breast milk, the influence of socioeconomic and environmental constraints on milk expression and storage, the experiential nature of breastfeeding knowledge acquisition, and the emotional commitment to breastfeeding in the absence of adequate professional and systemic support (Table 2). Table 2Main themes and sub-themes of the studyMain ThemesSub-Themes1. Breast Milk as an Absolute Good, but a Vague Knowledge Domain1.1. The Discourse of Benefit: “It Is Good, but We Do Not Know How or Why”1.2. Absence of Risk Perception: “Nothing Happens If Breast Milk Is Not Given”2. Milk Production Shaped More by Living Conditions Than by the Maternal Body2.1. Poverty and Infrastructure Barriers: “If I Cannot Store It, I Discard It”Subtheme 2.2. Breastfeeding and Pumping Limited to Hospital Visits3. Breastfeeding Knowledge Learned Through Trial and Error Rather Than Intergenerational Transfer3.1. Knowledge Construction Through Bodily Experience3.2. Disconnect Between Medical Recommendations and Everyday Practices4. Motherhood as a Well-Intentioned but Poorly Supported Effort4.1. Emotional Motivation Without Informational or Practical Support Table 2Main themes and sub-themes of the studyMain ThemesSub-Themes1. Breast Milk as an Absolute Good, but a Vague Knowledge Domain1.1. The Discourse of Benefit: “It Is Good, but We Do Not Know How or Why”1.2. Absence of Risk Perception: “Nothing Happens If Breast Milk Is Not Given”2. Milk Production Shaped More by Living Conditions Than by the Maternal Body2.1. Poverty and Infrastructure Barriers: “If I Cannot Store It, I Discard It”Subtheme 2.2. Breastfeeding and Pumping Limited to Hospital Visits3. Breastfeeding Knowledge Learned Through Trial and Error Rather Than Intergenerational Transfer3.1. Knowledge Construction Through Bodily Experience3.2. Disconnect Between Medical Recommendations and Everyday Practices4. Motherhood as a Well-Intentioned but Poorly Supported Effort4.1. Emotional Motivation Without Informational or Practical Support Main themes and sub-themes of the study Theme 1. Breast milk as an absolute good, but a vague knowledge domain Across all interviews, mothers consistently described breast milk as inherently beneficial for infants, frequently emphasizing its role in strengthening the body, improving brain development, and protecting against illnesses. However, these benefits were articulated in highly generalized terms, and most mothers were unable to specify which illnesses were prevented, how breast milk contributed to development, or how long exclusive breastfeeding should continue. Breast milk was thus framed as “good” based on collective beliefs and social discourse rather than concrete biomedical understanding. At the same time, mothers did not perceive the absence of breast milk as harmful. Many stated that infants fed formula also grow well and develop normally, often citing their previous children as examples. Consequently, breastfeeding was viewed as advantageous but not essential, and a lack of breast milk was not associated with perceived health risks. This dual perception of breast milk as beneficial but not necessarily suggests that while positive attitudes toward breastfeeding exist, they are not accompanied by a strong sense of urgency or perceived vulnerability when breastfeeding is disrupted due to hospitalization or logistical barriers. Subtheme 1.1. The discourse of benefit: “It is good, but we do not know how or why” Mothers consistently described breast milk as highly beneficial for infants, frequently stating that it supports brain development, strengthens the body, and protects against illnesses. However, these benefits were expressed in generalized and non-specific terms, and most mothers were unable to identify concrete health mechanisms, specific diseases prevented, or recommended durations of breastfeeding. References to protection against “many diseases” or to improving “intelligence” were common, yet without further elaboration. This suggests that breastfeeding is socially constructed as inherently positive, but not grounded in detailed biomedical understanding, indicating reliance on community narratives rather than health education. Breast milk protects the baby from many diseases, I don’t know exactly which ones, but it is very good for the brain and the body. (M1, 23 years old) Breast milk protects the baby from many diseases, I don’t know exactly which ones, but it is very good for the brain and the body. (M1, 23 years old) It helps the baby grow and become smarter, that’s what I know, but I don’t really know how it works. (M12, 26 years old) It helps the baby grow and become smarter, that’s what I know, but I don’t really know how it works. (M12, 26 years old) They say breast milk strengthens the baby and helps the brain develop, but I don’t know the details, I just know it is good.” (M11, 24 years old) They say breast milk strengthens the baby and helps the brain develop, but I don’t know the details, I just know it is good.” (M11, 24 years old) Subtheme 1.2. Absence of risk perception: “Nothing happens if breast milk is not given” Although breast milk was described as beneficial, many mothers also reported that breastfeeding would not harm the infant, noting that formula-fed children also grow and develop normally. Several mothers cited previous children who were not breastfed extensively and were perceived as healthy and intelligent. As a result, breastfeeding was framed as advantageous but not essential, and the absence of breast milk was not viewed as a health risk. This reflects a lack of perceived vulnerability related to non-breastfeeding, which may reduce urgency or persistence in milk expression during infant hospitalization. If the baby does not get breast milk, nothing bad happens. We give formula, and the baby still grows. (M12, 27 years old) If the baby does not get breast milk, nothing bad happens. We give formula, and the baby still grows. (M12, 27 years old) I didn’t breastfeed my other children much, and they are fine, they are smart and healthy. (M3, 30 years old) I didn’t breastfeed my other children much, and they are fine, they are smart and healthy. (M3, 30 years old) Breast milk is good, but if you don’t give it, it’s not a problem, you can feed with formula. (M4, 23 years old) Breast milk is good, but if you don’t give it, it’s not a problem, you can feed with formula. (M4, 23 years old) Theme 2. Milk production shaped more by living conditions than by the maternal body Mothers’ ability to provide breast milk was strongly influenced by their living conditions rather than by physiological capacity or motivation. Many mothers reported lacking basic infrastructure such as refrigerators, appropriate storage containers, or reliable transport to the hospital, which directly affected whether expressed milk could be preserved or delivered to their infants. As a result, milk was often discarded when immediate hospital visits were not possible. Mothers framed this not as a choice, but as an unavoidable consequence of their environment. Furthermore, breastfeeding and pumping practices were frequently confined to hospital visits. Several mothers reported that they only expressed milk when physically present in the neonatal unit, using hospital equipment, and did not pump regularly at home. This resulted in infrequent milk expression, reduced lactation stimulation, and a declining milk supply over time. Breastfeeding thus became an institutionally bounded activity rather than a continuous maternal practice integrated into daily life. These findings highlight that structural and socioeconomic constraints played a dominant role in shaping lactation behaviors, often overriding maternal intentions or beliefs about breastfeeding. Subtheme 2.1. Poverty and infrastructure barriers: “If i cannot store it, i discard it” Structural barriers strongly influenced breastfeeding practices. Many mothers reported not owning refrigerators, lacking appropriate containers, or being unaware of safe storage guidelines. As a result, expressed milk was often discarded if immediate transport to the hospital was not possible. These accounts indicate that milk loss was not due to unwillingness to breastfeed but rather to material deprivation and inadequate household infrastructure. Mothers frequently described discarding milk as unavoidable, framing it as a consequence of practical limitations rather than choice. I do not have a refrigerator at home, so when I pump, I usually have no choice but to throw the milk away. Even if I want to save it for my baby, I cannot keep it safely. (M6, 30 years old) I do not have a refrigerator at home, so when I pump, I usually have no choice but to throw the milk away. Even if I want to save it for my baby, I cannot keep it safely. (M6, 30 years old) If I cannot go to the hospital that day, I have to discard the milk, because there is no way for me to store it at home. I feel bad when I pour it out, but I do not have any other option. (M7, 24 years old) If I cannot go to the hospital that day, I have to discard the milk, because there is no way for me to store it at home. I feel bad when I pour it out, but I do not have any other option. (M7, 24 years old) I pump milk, but if I cannot bring it to the hospital, I pour it out, because there is no way to store it at home. Sometimes I pump and still have to throw it away, even though I know my baby needs it. (M11, 31 years old) I pump milk, but if I cannot bring it to the hospital, I pour it out, because there is no way to store it at home. Sometimes I pump and still have to throw it away, even though I know my baby needs it. (M11, 31 years old) Subtheme 2.2. Breastfeeding and pumping limited to hospital visits For many mothers, milk expression occurred primarily or exclusively during hospital visits rather than at home. Some mothers reported pumping only when physically present in the neonatal unit, using hospital-provided equipment, and not engaging in regular expression at home. This led to irregular pumping schedules and reduced stimulation of milk production. Consequently, breastfeeding became an activity institutionally confined to the hospital setting, disrupting continuity of lactation practices and shifting maternal care routines away from the home environment. I usually pump only when I come to the hospital. At home I do not pump regularly, because I do not have the same conditions or equipment. (M9, 27 years old) I usually pump only when I come to the hospital. At home I do not pump regularly, because I do not have the same conditions or equipment. (M9, 27 years old) I pump when I arrive at the hospital and again before I leave, but at home I do not pump. So there are long hours when I do not express any milk. (M3, 30 years old) I pump when I arrive at the hospital and again before I leave, but at home I do not pump. So there are long hours when I do not express any milk. (M3, 30 years old) I use the pump in the hospital, because at home I do not have a machine. That is why I mostly pump only when I am here at the hospital. (M12, 27 years old) I use the pump in the hospital, because at home I do not have a machine. That is why I mostly pump only when I am here at the hospital. (M12, 27 years old) Theme 3. Breastfeeding knowledge learned through trial and error rather than intergenerational transfer Rather than relying on formal health education or advice from family elders, mothers primarily constructed breastfeeding knowledge through personal experience and bodily feedback. Many described learning that milk increased when they pumped more frequently or after eating and decreased when they were unable to express milk regularly. These experiential observations formed the basis of their understanding of milk production and guided their practices. However, there was limited awareness of standardized recommendations regarding pumping frequency, storage duration, or safe transportation of expressed milk. Although some mothers reported practices that resembled medical guidelines such as freezing milk or using ice during transport, these actions were not based on formal instruction but rather on intuition or improvisation. This disconnect indicates that breastfeeding knowledge remained largely individualized and situational, rather than systematically transmitted through healthcare services or community-based education, resulting in inconsistent and sometimes inefficient lactation management. Subtheme 3.1. Knowledge construction through bodily experience Rather than following standardized recommendations, mothers described learning about breastfeeding through observing bodily responses, such as increased milk output after eating or after pumping more frequently. Statements such as “when I pump, the milk increases” or “if I eat, my milk comes” reflect a self-developed physiological model of lactation. This experiential learning served as the primary source of breastfeeding knowledge, particularly in the absence of professional counseling, and shaped mothers’ beliefs about how milk production could be maintained or improved. I noticed that when I pump more often, my milk increases. I learned this by watching my own body and seeing what happens when I pump more. (M7, 24 years old) I noticed that when I pump more often, my milk increases. I learned this by watching my own body and seeing what happens when I pump more. (M7, 24 years old) When I eat well, my milk comes more, and when I do not eat enough, there is less milk. I understand this only from my own experience, not because someone explained it to me. (M1, 23 years old) When I eat well, my milk comes more, and when I do not eat enough, there is less milk. I understand this only from my own experience, not because someone explained it to me. (M1, 23 years old) When I express milk, I see that it increases, so I try to pump whenever I can. This is something I learned by myself over time. (M2, 29 years old) When I express milk, I see that it increases, so I try to pump whenever I can. This is something I learned by myself over time. (M2, 29 years old) Subtheme 3.2. Disconnect between medical recommendations and everyday practices Despite some mothers reporting practices that coincidentally aligned with medical recommendations (e.g., freezing milk, using ice during transport), these behaviors were not guided by formal instruction. Instead, mothers frequently stated that they did not know how milk should be stored, how often pumping should occur, or how long milk could be kept safely. This gap between recommended practices and actual behaviors highlights a lack of structured lactation guidance, resulting in inconsistent and sometimes suboptimal milk handling practices. I really do not know how long the milk can be kept or how it should be stored safely. (M8, 31 years old) I really do not know how long the milk can be kept or how it should be stored safely. (M8, 31 years old) I put the milk in the freezer, but I do not know the correct way to do it; I just do it like this because no one showed me. (M2, 29 years old) I put the milk in the freezer, but I do not know the correct way to do it; I just do it like this because no one showed me. (M2, 29 years old) No one told me how often I should pump or how I should bring the milk to the hospital, so I just do what I think is right. (M6, 30 years old) No one told me how often I should pump or how I should bring the milk to the hospital, so I just do what I think is right. (M6, 30 years old) Theme 4. Motherhood as a well-intentioned but poorly supported effort Mothers consistently expressed strong emotional attachment to breastfeeding, describing it as a meaningful and satisfying aspect of motherhood that fostered closeness and emotional well-being. Feelings of happiness, calmness, and fulfillment were commonly associated with providing breast milk to their infants. This emotional motivation suggests that reluctance to breastfeed was not a driving factor behind reduced milk provision. Despite this strong intrinsic motivation, mothers received limited informational, practical, or material support to sustain breastfeeding while their infants were hospitalized. Many were left to manage pumping schedules, storage, and transportation of milk independently, often without guidance from healthcare professionals. Responsibility for maintaining lactation was therefore individualized, and systemic support mechanisms were largely absent. This placed mothers in a position where breastfeeding success depended heavily on personal resilience rather than structured institutional assistance, reinforcing the perception of motherhood as a solitary effort rather than a shared caregiving responsibility. Subtheme 4.1. Emotional motivation without informational or practical support Across interviews, mothers described breastfeeding as emotionally meaningful, reporting feelings of happiness, calmness, and maternal fulfillment when able to provide breast milk. However, this emotional commitment was not matched by access to consistent education, counseling, or material resources. Mothers were often left to manage milk expression, storage, and transport independently. This indicates that while motivation to breastfeed was strong, institutional and community support systems were insufficient, placing the burden of maintaining lactation almost entirely on the individual mother. Breastfeeding makes me feel good; it is a very beautiful and meaningful feeling for me as a mother. (M2, 29 years old) Breastfeeding makes me feel good; it is a very beautiful and meaningful feeling for me as a mother. (M2, 29 years old) When I am able to give my own milk to my baby, I feel happy and relieved, but I honestly do not know what else I should be doing to help my baby more. (M8, 31 years old) When I am able to give my own milk to my baby, I feel happy and relieved, but I honestly do not know what else I should be doing to help my baby more. (M8, 31 years old) I really want to give my baby breast milk, but I do not know the correct way to do it, and there is no one who explains it to me or supports me. (M6, 30 years old) I really want to give my baby breast milk, but I do not know the correct way to do it, and there is no one who explains it to me or supports me. (M6, 30 years old) Discussion The aim of this study was to explore mothers’ perceptions and experiences regarding breast milk while their infants were hospitalized in the NICU. One of the principal findings was that mothers consistently perceived breast milk as beneficial; however, this perception was largely based on generalized beliefs rather than detailed biomedical knowledge. While breastfeeding is globally endorsed as the optimal method for infant nutrition, mothers often operate within a “vague knowledge domain” where the benefits of breast milk are acknowledged as an “absolute good” without a specific understanding of biomedical mechanisms or recommended durations [23]. This pattern is not unique to the Somali context: a systematic review of qualitative research across eleven sub-Saharan African countries found that knowledge of the benefits of exclusive breastfeeding was among the most commonly reported maternal-infant factors shaping breastfeeding behavior, yet this knowledge remained largely surface-level and disconnected from informed risk perception [24]. Consistent with our findings, mothers described breast milk as beneficial for infant growth, brain development, and protection from diseases, yet they were generally unable to explain how these benefits occur or which specific health outcomes are affected [23, 26]. Public health messaging has been highly effective in disseminating the “Breast is Best” slogan, yet many low-income mothers do not perceive the absence of breast milk as a significant health risk, often viewing formula as a “good enough” substitute [14, 23, 27]. Similarly, many mothers in this study reported that infants fed with formula could grow and develop normally, indicating a limited perception of risk when breast milk is not provided. These findings suggest that although positive attitudes toward breastfeeding exist, public health messaging may not always translate into a clear understanding of the health implications of not providing breast milk [23, 24]. Beyond knowledge and risk perception, breastfeeding practices are also shaped by social norms and structural barriers. For mothers of hospitalized or preterm infants, the provision of expressed breast milk serves as a marker of “good motherhood” and an emotional connection to the infant who is otherwise under the care of medical staff [29–30]. However, the inability to produce sufficient milk may lead to feelings of guilt, inadequacy, and failure [26–31]. In addition, logistical challenges such as transportation costs, limited access to refrigeration, and the demanding and isolating nature of pumping may further hinder mothers’ ability to maintain lactation during infant hospitalization [30–31]. Structural and socioeconomic barriers, therefore, play an important role in shaping breastfeeding practices, suggesting that interventions should extend beyond education to address the practical and social constraints mothers face [23–27]. The maintenance of lactation among low-income mothers is predominantly dictated by structural and infrastructural constraints rather than maternal physiology. Consistent with the findings of this study, breastfeeding was often shaped by living conditions such as access to refrigeration, storage containers, pumping equipment, and reliable transportation to the hospital [14–231. Successful breastfeeding is frequently characterized as a marker of socioeconomic privilege, as it is more easily attained by those with financial security and flexible schedules [23–27]. In line with our results, mothers reported that expressed milk was sometimes discarded when it could not be safely stored or transported, indicating that milk loss was often a consequence of material limitations rather than lack of motivation. Mothers report intense feelings of guilt and sadness when they are forced to pour out milk they know their infants need due to these storage limitations. Another important finding was that milk expression was often limited to hospital visits, as some mothers relied on hospital-based pumping equipment and did not pump regularly at home [14–31]. These constraints result in prolonged intervals between expressions, which physiologically suppresses milk production by failing to provide the frequent hormonal stimulation necessary to maintain supply [25–31]. Furthermore, the high cost and logistical challenges of transporting infants to the neonatal intensive care unit are significant determinants of health, often forcing mothers to choose between visiting their infants and maintaining their milk supply [14–30]. The material realities of poverty, including rigid workplace schedules and limited access to family-integrated care facilities, make sustained lactation feel unsustainable or even “impossible” for many [14– [27–32]. Overall, these findings suggest that sustaining lactation during infant hospitalization requires not only maternal motivation but also adequate environmental and institutional support that addresses structural barriers faced by mothers. Breastfeeding knowledge among mothers is increasingly constructed through individual bodily experience and trial-and-error rather than through the traditional intergenerational transfer of maternal wisdom. Consistent with this study’s findings, mothers often described learning about lactation by observing their own bodily responses, such as increased milk output after more frequent pumping or after eating, rather than through formal education or structured family guidance [14–31]. Mothers often develop an understanding of lactation mechanics by observing biological feedback loops, such as the correlation between increased pumping frequency or specific nutritional intake and fluctuations in milk volume [14]. In the absence of direct nursing, mothers frequently utilize the volume of mechanical output as a physiological “gauge” to interpret their internal supply and guide subsequent feeding and expression behaviors [31]. Another important finding was the limited awareness of standardized recommendations regarding pumping frequency, milk storage, and safe transportation [14–31]. Although some mothers reported practices resembling medical guidance, these were often based on intuition or improvisation rather than professional instruction. However, this reliance on personal experimentation highlights a significant disconnect between clinical recommendations and everyday practices, leaving many mothers to manage complex tasks solely on intuition [28–31]. Specific knowledge gaps are prevalent, with mothers often reporting a lack of standardized guidance on safe milk storage times or appropriate freezer protocols [25]. This deficit is frequently exacerbated by systemic constraints on healthcare providers, who are often perceived as too busy to provide the necessary one-on-one guidance or to repeatedly reinforce educational content [25–32]. Overall, these findings suggest that breastfeeding knowledge during infant hospitalization remains highly individualized and situational, and that the transition to effective lactation management may be hindered when mothers do not receive accessible, practical, and consistently reinforced professional support [33–34]. Mothers often conceptualize breast milk as a foundational marker of “good motherhood,” viewing it as a vital emotional bridge that fosters a sense of relief and connection with their infants [29]. Consistent with this study’s findings, mothers described breastfeeding as emotionally meaningful and associated it with happiness, calmness, and maternal fulfillment, indicating that reduced milk provision was not primarily due to a lack of motivation [29–31]. While breastfeeding is perceived as an “absolute good” that yields significant psychological rewards for the mother, this strong emotional motivation is frequently unsupported by a corresponding level of technical or informational guidance [23–25]. In line with our results, many mothers reported receiving limited informational, practical, or material support while trying to maintain lactation during infant hospitalization [26, 27]. As a result, they were often left to manage pumping, storage, and transportation of milk on their own, without consistent professional guidance [14, 29]. Consequently, many mothers operate within a “vague knowledge domain,” possessing an intense desire to nourish their babies but lacking specific biomedical understanding regarding the mechanical processes of expression or the duration required for exclusivity [23–26]. These findings suggest that breastfeeding during infant hospitalization was experienced as a well-intentioned but poorly supported maternal effort, with responsibility placed largely on the mother rather than shared through structured institutional support [30, 32–36]. Implications The findings indicate that nursing support in NICUs should prioritize structured, skills-based lactation counseling focused on pumping frequency, milk handling, and transport rather than relying solely on general breastfeeding promotion messages. Given that breastfeeding practices in this study were strongly shaped by structural and environmental constraints, breastfeeding interventions should also address the practical and contextual barriers mothers face. Effective breastfeeding support should incorporate practical resource facilitation, such as access to pumps, appropriate containers, and feasible cooling or transport options. These findings also highlight the importance of culturally tailored breastfeeding interventions that account for local cultural beliefs, social norms, and socioeconomic realities that influence mothers’ breastfeeding practices. Integrating brief, standardized lactation guidance into routine NICU workflows and providing repeated reinforcement during visits may help reduce reliance on trial-and-error learning and support continuity of milk expression during hospitalization. Such context-sensitive, culturally responsive approaches may improve the feasibility and effectiveness of breastfeeding support for mothers of infants hospitalized in NICUs. Beyond individual-level clinical interventions, these findings carry broader implications for policy and systems-level change. A fundamental shift is needed away from mother-infant separation as the default model of NICU care. Dyadic care practices, most notably kangaroo mother care (KMC), which has robust evidence for improving neonatal outcomes and supporting breastfeeding [8], should be actively promoted and structurally embedded within NICU environments in alignment with WHO implementation guidance [WHO KMC Implementation Strategy; WHO KMC Clinical Practice Guide]. Achieving this requires not only facility-level practice changes but also social and financial supports for families navigating medical crises, such as supplemental income protection or work relief for parents whose employment is disrupted by a prolonged NICU admission. Removing the structural conditions that force extended separation, while simultaneously facilitating breastfeeding and strengthening family-centered care, would be transformative for fragile newborns and their families. Implementing these recommendations across health systems has the potential to reduce inequities in neonatal outcomes and fundamentally reshape how institutions support families during one of the most vulnerable periods of early life. Limitations This study was conducted in a single tertiary NICU in Mogadishu with a purposive sample of 12 mothers, which may limit transferability to other regions, rural settings, or different levels of neonatal care in Somalia. Recruitment during daytime visiting hours may have excluded mothers with greater socioeconomic or transportation barriers, potentially underrepresenting the most vulnerable populations. Additionally, although rigorous qualitative procedures were followed, potential nuances may have been lost during the transcription and translation processes. Conclusion This study found that Somali mothers of NICU-hospitalized infants strongly valued breast milk but often had limited biomedical understanding of breastfeeding, relying instead on generalized social knowledge and personal experience. Breastfeeding practices were predominantly shaped by structural and environmental constraints, including limited access to refrigeration, storage materials, pumping equipment, and transportation, which frequently disrupted milk expression and continuity of lactation. These findings highlight that improving breastfeeding outcomes in NICU settings requires culturally and contextually tailored interventions that address structural barriers and provide consistent practical and professional support for mothers.