Abstract

Background

Age-appropriate breastfeeding and adequate complementary feeding play a crucial role in reducing under-5 mortality rates, malnutrition, and mortality worldwide. Given the high prevalence of adolescent pregnancy in Uganda, we aimed to assess the competence of adolescent mothers in infant and young child feeding (IYCF) using data from the 2016 Uganda Demographic Health Survey.

Method

Our sample comprised the last children aged 6鈥�23 months of mothers aged 15鈥�29 years, totaling 2594 mother-infant pairs; 402 were aged 15鈥�19 years, 1,388 were aged 20鈥�29 years with their first delivery before the age of 20, and 804 were aged 20鈥�29 years with their first delivery at 20 years or older. We utilized complex sample logistic regression analysis to examine the interactions between adolescent motherhood and IYCF indicators during the previous day of the survey.

Results

The rates of ongoing breastfeeding, bottle feeding, minimum meal frequency, minimum dietary diversity, minimum acceptable diet, egg and/or flesh food consumption, zero vegetable or fruit consumption, and unhealthy food consumption were found to be 77.7%, 16.3%, 37.5%, 28.7%, 14.2%, 41.7%, 31.6%, and 14.7%, respectively. Multivariate complex sample logistic regression revealed that women with a history of adolescent pregnancy had lower odds of achieving minimum meal frequency, and higher odds of zero vegetable or fruit consumption.

Conclusion

Adolescent mothers and women who give birth during their teenage years are particularly vulnerable to inadequate infant and young child nutrition practices. Targeted interventions and education programs should be implemented to promote optimal complementary feeding practices among mothers in Uganda.

Introduction

Adolescent pregnancy is a major public health problem, especially in low- and middle-income countries [1]. Pregnant adolescents often encounter heightened vulnerabilities, facing obstacles in accessing healthcare services and experiencing a higher likelihood of adverse outcomes compared to adult pregnant women [2, 3]. Various factors, such as socio-economic status, limited education, substance abuse, and inadequate antenatal care, contribute to the complexity of adolescent pregnancies [2, 4, 5]. An international study across 24 countries in Latin America, Africa, and Asia found that the adolescent group had higher rates of low education level, lack of antenatal care, maternal death, preterm delivery, and perinatal mortality compared to the adult group [4]. Namutebi et al. found through in-depth interviews with 42 first-time teenage mothers aged 14 to 19 years in Uganda that many lacked information about postpartum care for themselves and their babies, and while some appreciated the guidance they received, others found it insufficient, leading to confusion and frustration in implementing self-care and infant care practices after hospital discharge [6]. Indeed, adolescents often fail to consume a balanced diet, as highlighted in a study involving 598 adolescents from the African Research, Implementation Science, and Education (ARISE) Network in Uganda, which found that 45.3% had a low dietary diversity score [7]. The study reported that common food categories consumed included cereals/roots/tubers (99.7%), fats & oils (87.0%), and sweets (77.1%), while less frequent consumption was observed for nutrient-dense foods like meat, dairy, eggs, and vitamin A-rich fruits and vegetables. Adolescents from low socio-economic backgrounds and those living with a single parent or guardian, along with those dependent on home-cooked meals, were more likely to have low dietary diversity [7]. Given these patterns in adolescent behavior, it is plausible that similar dietary challenges may be present in the children they care for, influencing their nutritional outcomes as well.

Against the backdrop of a global landscape marked by diverse reproductive health challenges, adolescent pregnancies stand out as a significant concern. Annually, an estimated 16听million girls between the ages of 15 and 19 experience childbirth worldwide, underscoring the urgency of addressing this complex issue. Regional disparities further emphasize the nuanced nature of this phenomenon. Between 2015 and 2020, 14.5% of women aged 20鈥�24 globally gave birth before the age of 18. However, the prevalence varied starkly across regions, with rates of 10.8% in South Asia, 25.2% in East and Southern Africa, 27.2% in West and Central Africa, and 26.3% in Sub-Saharan Africa [8, 9]. The 2016 Uganda Demographic Health Survey (UDHS) highlighted the specific challenges faced by Uganda, revealing that 25% of 15鈥�19-year-olds had begun childbearing [10]. Additionally, over half of women in Uganda who have a first birth before 18 report a repeat adolescent birth before 20, with no decline observed in the past 30 years, influenced by factors such as poverty, low education, early union, lack of contraceptive use, and previous pregnancies [11].

The nutritional impacts experienced by children born to adolescent mothers can initiate a cycle of adverse health conditions that transcends generations. The repercussions of early motherhood on early childhood nutrition have the potential to set in motion a pattern of intergenerational transmission, wherein health challenges are passed from one generation to the next. Chilot et al. [12] examined the 鈥榯riple burden of malnutrition,鈥� which encompasses overnutrition, undernutrition, and micronutrient deficiencies, in 22 low- and middle-income countries (LMICs) that had Demographic and Health Surveys (DHS) conducted between 2016 and 2020. Among mother鈥揷hild pairs in these countries, the combined prevalence of this triple burden was 11.39% (95% CI: 9.56鈥�13.23%). In Uganda specifically, the prevalence stood at 6.92% (95% CI: 5.49鈥�8.35%). In that study, maternal age and high parity were associated with an increased triple burden of malnutrition, and parity was identified as a key determinant [12]. Adolescent pregnancy may further increase the number of children, potentially leading to a higher triple burden in later years.

The World Health Organization (WHO) and the United Nations Children鈥檚 Fund (UNICEF) recommend that infants be breastfed within the first hour of birth and continue breastfeeding for a minimum of 2 years, while also introducing appropriate complementary feeding at 6 months [13]. Age-appropriate and adequate complementary feeding plays a pivotal role in mitigating triple burden of malnutrition and mortality among children [14]. The WHO and UNICEF aimed to promote breastfeeding in Uganda through the Baby-Friendly Hospital Initiative, which was initiated in 1991 by the Ministry of Health and involved training healthcare workers in childbirth and newborn services [15]. The median duration of breastfeeding in Uganda, according to the World Breastfeeding Trends Initiative, is 19 months [16]. During this process, 52.5% of children were breastfed within the first hour after birth, 41.1% received prelacteal foods, and 15% were fed with bottles. Despite these international guidelines, child malnutrition remains a pressing issue, as evidenced by UNICEF鈥檚 2023 report, indicating that 22.3% of children under 5 years were affected by stunting in 2022, with a substantial proportion (43%) residing in Africa [17].

In light of the high prevalence of adolescent motherhood in Uganda, this study seeks to assess the adequacy of infant and young child feeding (IYCF) practices among adolescent mothers, utilizing data from the 2016 UDHS. The hypothesis driving this research is that adolescent motherhood may have insufficient support from healthcare professionals [6], a negative impact on IYCF practices, potentially contributing to adverse health outcomes for both mothers and their offspring. Understanding the nuances of this association is crucial for informing targeted interventions and policies aimed at improving the well-being of adolescent mothers and their children in Uganda.

Methods

Data collection and participants

In conducting this study, we used data from the UDHS of 2016, a comprehensive national survey executed between June and December of that year [10]. The household survey of UDHS, designed to be nationally representative, adopted a two-stage cluster sampling methodology. In the first stage, a nationally representative sample of 20,910 households from 697 sample clusters- referred to as enumeration areas (EAs)- were meticulously selected from the 2014 Uganda National Population and Housing Census (NPHC). This included 162 EAs in urban areas and 535 in rural regions, with one cluster from Acholi subregion excluded due to land disputes. In the second stage, households were randomly selected, with 30 households chosen per EA segment to ensure a balanced and representative sample. All women aged 15鈥�49 who were usual residents or who had stayed in the selected households the night before the survey were eligible for interview. The 2016 UDHS sample was specifically designed to provide estimates across the country, for urban and rural areas separately, and for each of Uganda鈥檚 15 regions (South Central, North Central, Busoga, Kampala, Lango, Acholi, Tooro, Bunyoro, Bukedi, Bugisu, Karamoja, Teso, Kigezi, Ankole, and West Nile). Of the 20,880 selected households, interviews were successfully conducted with 18,506 women, averaging around 1,200 completed interviews per region.

The recruitment criteria for our study focused on infants under 2 years of age and mothers aged 15鈥�29 years at the time of their last pregnancy. These criteria were diligently applied during the fieldwork period, resulting in the identification of 3,815 mother-baby pairs meeting the recruitment criteria. In addition to the adolescent group (15鈥�19 years), the 20鈥�29 age range was included in the study, as it is widely recognized as the optimal reproductive period for childbearing from both biological and social perspectives. This age range provides a suitable baseline for identifying risks specific to adolescence. Women aged 30 years and above were excluded from the study, as this group鈥攑articularly those aged 35 and older鈥攆aces distinct biological and clinical risks that differ significantly from those of younger age groups [18].

To ensure the precision of the study, certain exclusion criteria were also applied, leading to the exclusion of babies not cohabiting with the mothers, including babies who had unfortunately passed away before the study (n鈥�=鈥�116), multiple births (n鈥�=鈥�54), and preterm births occurring before the 32nd week of pregnancy (n鈥�=鈥�86) (Fig.听1). Following the application of these exclusion criteria, the participating mothers were categorized into three groups: 603 mothers aged 15鈥�19 years, 1,848 mothers aged 20鈥�29 years who had their first birth before the age of 20, and 1,113 mothers aged 20 and above who had their first birth after the age of 20. Simultaneously, the children were classified into two age groups: 0鈥�5 months and 6鈥�23 months, with only those in the latter age group included in the study; among the enrolled mothers having children aged 6鈥�23 months, 402 were aged 15鈥�19 years, 1,388 were aged 20鈥�29 years with their first delivery before the age of 20, and 804 were aged 20鈥�29 years with their first delivery at 20 years or older. Consequently, our study meticulously examined a total of 2,594 mother-baby pairs, ensuring a robust and focused analysis of the targeted demographic (Fig.听1).

Fig. 1

Flowchart of Excluded Cases and Distribution by Group [One case might have more than one exclusion criteria]

Outcome variables

The primary variables for infants aged 6鈥�23 months under scrutiny encompassed the ongoing breastfeeding (OBF), the promotion of minimum dietary diversity (MDD), the promotion of minimum meal frequency (MMF), Minimum acceptable diet (MAD), the consumption of eggs and/or meat (EFF), the consumption of unhealthy foods (UF), Zero vegetable or fruit consumption (ZVF), use of Bottle feeding (BoF) during the previous day (Table听1), partially adapted from key IYCF indicators aligned with UNICEF to ensure consistency and relevance in the assessment process [13].

Table 1 Feeding indicators for children 6鈥�23 months of age

As for the independent variables, they encompassed the general characteristics of both the mothers and the children, offering a broader contextual understanding of the factors influencing the studied feeding practices.

Statistical analysis

The analysis was conducted utilizing the IBM-SPSS 23.0 and Stata 13.0 software packages. The CSPLAN file was employed and a weighted approach was implemented to meticulously organize the sample distribution and to ensure the accuracy and representativeness of the findings.

The characteristics of the mothers and children were presented as counts (n) and percentage distributions. The following aspects were examined and presented as frequencies (n) and percentage distributions of OBF, MDD, MMF, MAD, EFF, UF, ZVF, BoF. The chi-square test was used to analyze the association between the variables and the feeding indicators. Variables with more than two subgroups and statistically significant differences between subgroups were further evaluated for subgroup interactions using adjusted residuals.

The examination of differences in feeding indicators based on maternal-child variables was conducted through complex sample logistic regression, providing odds ratios (OR) and 95% confidence intervals (CI). As previously recommended for selecting right variable in logistic regression [19], the mother-child factors with p value鈥�<鈥�0.1 in each feeding indicators were included in multivariable complex sample logistic regression analysis to determine the associated factors for that indicator.

In all analyses, a significance level of p鈥�<鈥�0.05 was considered statistically significant.

Results

A total of 2,594 mother-infant pairs were included in the study, with data sourced from the 2016 Uganda DHS (Table听2). Among the mothers, 15.5% belonged to the 15鈥�19 age group (Group 1), 53.5% in the 20鈥�29 age group had their first pregnancy before 20 years (Group 2), and 31% had their first pregnancy at or after the age of 20 (Group 3). Mother-infant characteristics were given in Table听2.

Table 2 General characteristics of mother-child pairs (n鈥�=鈥�2594)

Ongoing breastfeeding (OBF)

At least three-fourths of children aged 6鈥�23 months continued breastfeeding in the last 24听h (Table听3). When analyzing the connection between the continuation of breastfeeding and maternal characteristics, the study revealed that the odds of Group 1 had a 71% times higher for OBF than Group 3 (%95 Cl: 1.56鈥�2.52, Table听3). Rural residence was associated with a 2.27 times greater likelihood of OBF (%95 Cl: 1.76鈥�2.92). Higher household welfare index was linked to lower OBF [OR (95%CI): 0.45 (0.35鈥�0.57)] compared to lower one. The percentage of OBF was decreased with increased maternal education (<鈥�0.001). Mothers engaged in family work were 46% more likely to continue breastfeeding compared to never worked (95%CI: 1.13鈥�1.88). OBF was significantly lower among currently pregnant mothers [OR (95%CI): 0.05 (0.04鈥�0.07)]. Additionally, OBF was 34% higher in children born as the third child or later compared to the first child. Compared to home, delivery in private institutions were associated with lower odds of OBF [OR (95%CI): 0.62 (0.46鈥�0.84)]. OBF rates were the lowest among children aged 18鈥�23 months compared to those aged 6鈥�8 months (p鈥�<鈥�0.001). Children with a small size at birth had higher percentage for OBF compared to average and large size (p鈥�=鈥�0.004, Table听3).

Table 3 Associations between mother-infant characteristics and the 鈥渙ngoing breastfeeding鈥� and 鈥渂ottle feeding practices鈥�

Bottle feeding (BoF)

Overall, 16.3% of infants were bottle-fed (Table听3). When examining the relationship between maternal characteristics and BoF in our study, it was found that BoF was lower in Group 2 than Group 3 and Group 1 (p鈥�=鈥�0.001). Rural areas had lower odds for BoF [OR (95%CI): 0.47 (0.35鈥�0.63)]. The Lango and Iteso ethnic backgrounds exhibited the lowest BoF (p鈥�<鈥�0.001). Percentages of BoF were increased with level of maternal education (p鈥�<鈥�0.001). Mothers working in family businesses had lower odds for BoF [OR (95%CI): 0.52 (0.36鈥�0.73)]. Children born as the 3rd child and higher BoF had odds ratio 0.66 (Ref: the first child). Children of mothers with 4 or more ANC visits had 43% higher odds for BoF than no ANC. BoF percentages was higher in children born in institutions compared to home births (p鈥�<鈥�0.001). BoF percentage was higher in children receiving breastfeeding counseling within 2 days after delivery than in absence of counseling (p鈥�=鈥�0.034). BoF was lowest among children aged 12鈥�17 and 18鈥�23 months compared to younger ages (p鈥�&濒迟;鈥�0.001, Table听3).

Minimum meal frequency (MMF) status

Of children aged 6鈥�23 months, 37.5% met the MMF appropriate for their age group in the last 24听h (Table听4). Group 2 had lower achieving MMF percentage compared to Group 3 (p鈥�=鈥�0.043). MMF was positively associated with rural residence, Iteso ethnic group, those engaged in family work, and with 4 or more prenatal care visits, and breastfeeding counseling within 2 days (p鈥�<鈥�0.05). Negative relation was found for MMR with currently pregnant, older children (p鈥�<鈥�0.05, Table听4).

Table 4 Associations between mother-infant characteristics and compliance with 鈥渕inimum meal frequency鈥� and 鈥渕inimum dietary diversity鈥� guidelines

Minimum dietary diversity (MDD) status

Overall, 28.7% of children had a diet diversity appropriate for their age and breastfeeding status in the last 24听h (Table听4). Group 1 and Group 2 had lower odds for MDD compared to Group 3 [respectively, 0.60(0.44鈥�0.83), 0.81 (0.65-1.00)]. Rural areas had lower odds for MDD [OR (95%CI): 0.80 (0.66鈥�0.97)]. MDD was changed with ethnicity (p鈥�<鈥�0.001). Higher wealth index and higher maternal education were associated with higher achieving MDD (p鈥�<鈥�0.001 for both). Mothers working for someone else and mothers working in family businesses had higher percentages for MDD (p鈥�=鈥�0.001). Counseling on breastfeeding within the first 48听h was associated with 26% higher odds for achieving MDD (95% CI: 1.01鈥�1.56). MDD percentages were higher among children aged 9鈥�11 months and 12鈥�17 months (p鈥�&濒迟;鈥�0.001, Table听4).

Egg and/or flesh food (EFF) consumption

Less than half of the children (41.7%) had EFF consumption in the last 24听h (Table听5). The percentages of EFF consumption did not change significantly among maternal age groups. Rural areas had lower odds of EFF consumption [OR (95%CI): 0.66 (0.55鈥�0.79)]. The Bagisu and Lango ethnic group exhibited the lowest percentages for EFF consumption and Iteso the highest (p鈥�<鈥�0.001). Higher and lower wealth index was associated with higher percentages for EFF consumption than medium index (p鈥�<鈥�0.001). Counseling on breastfeeding within the first two postpartum days was associated with higher percentages for EFF consumption (p鈥�<鈥�0.001). EFF consumption percentage was lowest in children aged 6鈥�8 and highest in those aged 18鈥�23 months and (p鈥�&濒迟;鈥�0.001, Table听5).

Table 5 Associations between mother characteristics and compliance with 鈥渆gg and/or flesh food consumption鈥� and 鈥渕inimum acceptable diet鈥� guidelines

Minimum acceptable diet (MAD) status

Only 14.2% of children were reported to have been fed with MAD appropriate for their age in the last 24听h (Table听5). Group 2 mothers had lower odds for MAD compared to group 3 [OR (95%CI): 0.75 (0.57鈥�0.98)]. Among ethnicities, Banyankore had the highest MAD percentages, while Lango had the lowest (p鈥�=鈥�0.002). Maternal education being at least incomplete secondary school, as well as maternal employment were associated with higher odds for MAD. The MAD percentages was the lowest among infants aged 18鈥�23 months (p鈥�&濒迟;鈥�0.001, Table听5).

Zero Vegetable and Fruit (ZVF) Consumption

Nearly one-third of children had not consumed any fruits or vegetables in the last 24听h (Table听6). Group 2 had higher odds for ZVF consumption [OR (95%CI): 1.30 (1.05鈥�1.62)]. The Lango ethnic group had the highest ZVF consumption and Banyankore the lowest (p鈥�<鈥�0.001). ZVF consumption percentages were changed with mothers education (p鈥�<鈥�0.001); those having no education had the highest. Mothers working in family businesses showed lower ZVF consumption than those never worked (p鈥�=鈥�0.007). ZVF consumption was lower in pregnant women (p鈥�<鈥�0.001). ZVF consumption was highest in children aged 6鈥�8 months and lowest among children aged 18鈥�23 months (p鈥�&濒迟;鈥�0.001, Table听6).

Table 6 Associations between mother-infant characteristics and compliance with 鈥渮ero vegetable or fruit consumption鈥� and 鈥渦nhealthy food consumption鈥� guidelines

Unhealthy food (UF) Consumption

Unhealthy food consumption was reported in 14.7% of children (Table听6). UF consumption rates were similar in maternal age groups. Rural dwellers had lower percentages for UF consumption (p鈥�<鈥�0.001). The Lango and Bakiga ethnic groups showed the lowest UF consumption and Baganda and Iteso had the highest (p鈥�<鈥�0.001). Higher wealth index, higher maternal education, maternal employment, and birth in institutions were associated with higher percentages for UF consumption (p鈥�&濒迟;鈥�0.001, p鈥�&濒迟;鈥�0.001, p鈥�=鈥�0.003, and p鈥�=鈥�0.002; respectively). Unmarried mother, delayed initiation of BF, breastfeeding counseling in the first two days after birth, and older child age were associated with higher odds for UF consumption (p鈥�=鈥�0.004, p鈥�=鈥�0.022, p鈥�&濒迟;鈥�0.001, respectively, Table听6).

Multivariable complex sample logistic regression

In multivariable complex sample logistic regression, after controlling for confounding factors, Group 2 mothers showed lower odds for achieving MMF [OR (95%CI): 0.74 (0.60鈥�0.93)] and higher odds for ZVF consumption [OR (95%CI): 1.37 (1.08鈥�1.74)]. However, other feeding indicators did not change significantly according to mothers鈥� age groups (Table听7).

Table 7 Mother-infant factors related with feeding indicators, complex logistic regression*

The multivariable analysis revealed that the odds of OBF was higher in rural residences (ref: urban), among the Bagisu ethnicity (compared to Baganda), and lower among educated mothers (ref: no education), those currently pregnant (ref: not pregnant), child from unwanted pregnancy (ref: wanted from the begining), infants with larger birth size (ref: average), and infants older than 12 mo (ref:6鈥�8 mo). For BoF, infants from the Iteso ethnicity (compared to Baganda) and infants older than 12 months (ref: 6鈥�8 mo) had lower odds, while those from the wealthy households (ref: medium) had higher odds of lower BoF. Regarding MMF, Bakiga and Iteso ethnicities (compared to Baganda), maternal employment in family businesses (ref: never worked), and attending four or more antenatal care (ref:<4) sessions were associated with higher MMF, while infants older than 9 months (ref: 6鈥�8 mo) had lower MMF. For MDD, the Lango ethnicity (ref: Baganda) showed lower odds, while maternal employment (ref: never worked), small size baby at birth, and infants older than 9 months (ref: 6-8mo) were associated with higher MDD. An association was found between MAD and maternal education. The odds of ZVF varied according to mothers鈥� age groups, ethnicity, maternal education, maternal employment status, and the child鈥檚 age (Table听7).

Discussion

In this study, we observed that the likelihood of breastfeeding continuation is higher among teenage mothers, those residing in rural areas, households with the lowest socioeconomic status, mothers with no formal education, and mothers engaged in family businesses. However, multivariate analysis revealed a positive association between OBF and rural residence, Bagisu ethnicity, lack of maternal education, maternal involvement in family businesses, and the desire for the child from the beginning. Conversely, there was a negative association with being pregnant during the survey period, older infant age, large birth in size, and no significant relation with mother age groups. A study conducted in Uganda in 2014 supported our findings, highlighting higher breastfeeding rates in rural areas and among mothers with lower education levels. It also indicated that mothers who gave birth in hospitals were more likely to stop breastfeeding early [15]. In a qualitative study conducted in Uganda, it was noted that mothers believe their breast milk is insufficient and that exclusively breastfeeding is socially associated with poverty [20]. In a study covering 27 sub-Saharan African countries, a higher likelihood of exclusive breastfeeding (EBF) was associated with higher maternal education (secondary level or above), mothers aged 25鈥�34, living in rural areas, belonging to wealthier households, having 4 or more antenatal visits, giving birth in a health facility, singleton births, female infants, early initiation of breastfeeding (EIBF), and younger infant age [21]. When examining factors related to pregnancy and childbirth in our study, we found that the number of antenatal care visits, timing of breastfeeding initiation, and skin-to-skin contact at birth did not significantly influence breastfeeding continuation in children aged 6鈥�23 months. However, previous research has indicated a notable improvement in breastfeeding rates when antenatal breastfeeding education is integrated into routine antenatal care [22, 23].

Our study revealed that 36.8% of mothers received breastfeeding counseling within the first 48听h after delivery. However, despite this counseling, breastfeeding rates did not show a significant increase. This lack of improvement may be linked to societal perceptions that EBF is associated with poverty, as well as insufficient community education on breastfeeding practices [5, 20]. Comprehensive evaluations of breastfeeding counseling have demonstrated that community-based counseling positively influences infant feeding during the first six months, as shown in studies from Burkina Faso, Uganda, and South Africa [24]. In Tanzania, Maonga et al. [25] found that although 91% of mothers attended antenatal care, only 39% received breastfeeding counseling during pregnancy, and just 25% in the postpartum period. Mothers who did not practice EBF often believed that breast milk was insufficient and supplemented their infants鈥� diets with water and herbal products [25]. According to the WHO Global Analysis, barriers to breastfeeding in low- and middle-income countries include low levels of education, limited access to healthcare services, and cultural beliefs [26]. Additionally, practices such as separating mothers and newborns for rest or using pacifiers to soothe infants, as reported among Syrian mothers, negatively impact breastfeeding [23]. Prelacteal feeding is also known to adversely affect breastfeeding [27].

Both WHO and UNICEF aimed to promote breastfeeding through the Baby-Friendly Hospital Initiative, which was launched in Uganda in 1991 under the Ministry of Health. However, our study revealed that infants of mothers who received breastfeeding counseling in hospitals exhibited lower rates of continued breastfeeding and higher rates of bottle-feeding. To enhance breastfeeding outcomes, regular inspections of hospitals adhering to 鈥淏aby-Friendly鈥� standards, expert breastfeeding counseling, and consistent monitoring are essential [16]. Recognizing that the Baby-Friendly Hospital Initiative primarily benefits mothers accessing antenatal care and delivering in hospitals, the Baby-Friendly Community Initiative was introduced in Kenya. This community-based approach led to higher rates of exclusive breastfeeding during the first six months and longer breastfeeding durations in the intervention group compared to the control group [28].

In a qualitative study conducted in Uganda, mothers mentioned avoiding breastfeeding due to fear of being stigmatized as HIV-positive and confusion resulting from changing recommendations for HIV-positive mothers [20]. According to the 2016 UDHS report, 76% of women aged 15鈥�49 who gave birth in the past two years received counseling and testing for HIV during antenatal care visits [10]. In our study, the lower rates of breastfeeding among mothers with higher education levels, those in households with a higher wealth index, and those giving birth in hospitals can be attributed to these factors.

In our study, we found that teenage motherhood did not significantly impact the use of baby bottles. Instead, the likelihood of bottle usage was lower in rural areas, among households with the lowest socioeconomic status, mothers without formal education, mothers who delivered at home, and infants born through methods other than vaginal delivery. Interestingly, these factors seemed to support the continuation of breastfeeding. In studies conducted in Ethiopia by Kebebe et al., the lowest rates of bottle usage were found among mothers aged 15鈥�19, with bottle feeding increasing as the mother鈥檚 education level rose [29]. Using DHS data from 20 sub-Saharan African countries, including Uganda, the pooled prevalence of BoF among mothers of children aged 0 to 23 months was 13.7% [30]. Factors such as maternal age, education, marital status, occupation, media exposure, wealth index, household head鈥檚 gender, family size, number of children under five, place of birth, delivery method, breastfeeding counseling, child鈥檚 age, and residence were all significantly associated with BoF practices. Mothers aged 25鈥�34 years and 35鈥�49 years were respectively 1.09 and 1.08 times more likely to bottle-feed their infants compared to mothers aged 15鈥�24 years [30]. However, in our study, further analyses revealed that maternal age was not significantly associated with BoF, while wealth index, ethnicity, and child鈥檚 age emerged as relevant factors. A study in Namibia, a Sub-Saharan African country, found that women giving birth in healthcare facilities鈥攚hether through vaginal delivery or cesarean section鈥攚ere more likely to use bottles compared to those who gave birth at home [31]. These studies, all conducted in low-income countries, align with our findings. In a study analyzing 30 years of Turkey DHS data, no association was found between hospital births and increasing BoF trends [32].

When examining MMF, it was observed that adolescent mothers had lower rates of meeting MMF requirements. Mothers over the age of 20 who had their first pregnancy before 20 were found to have a lower prevalence of meeting MMF. This indicates that the negative effects of adolescent pregnancies may persist into subsequent pregnancies in adulthood. No other studies addressing this specific issue were found in the literature. Our study indicates that households with the highest socioeconomic status had a lower likelihood of achieving MMF. A cross-sectional study using secondary data from a USAID-funded project in Uganda, involving 384 caregivers of children aged 6鈥�23 months in the Kisoro district (part of the Kigezi region with stunting rates above the national average), found that 76.6% of caregivers met the MMF [33]. A meta-analysis conducted in Ethiopia showed that wealthy families were 2.11 times more likely to meet the recommended MMF for children aged 6鈥�23 months compared to poor families [34]. Furthermore, our study revealed that children in rural areas had a higher likelihood of meeting MMF. In contrast, a study by Dadzie et al. [35] in Ghana found that rural women were less likely to meet MMF for their children due to a lack of awareness regarding the importance of feeding frequency. As maternal education level increased, the likelihood of meeting MMF also increased. Similarly, highly educated mothers were more likely to provide a higher frequency of daily meals for their children, as they were more informed about the importance of good nutrition for their children鈥檚 health [35]. In our study, the probability of meeting MMF decreased as the child鈥檚 age increased. This contrasts with a study by Beyene et al. in Ethiopia, where it was observed that children aged 12鈥�17 months and 18鈥�23 months had a higher likelihood of meeting MMF in their daily meals compared to children aged 6鈥�11 months [36].

The study found that 28.7% of children were provided with MDD. Univariate analysis revealed the lowest MDD percentage among adolescent mothers (group 1) and a low percentage among adult mothers with a history of adolescent pregnancies compared to adult mothers without such history. Multivariate analysis showed higher odds of meeting MDD with increasing infant age, maternal occupation, smaller infant size at birth, and breastfeeding counseling during the first two postpartum days. A cross-sectional study in Uganda reported that despite 95% of infants being introduced to semi-solid foods on time, only 4.4% met the MDD and MAD standards [33]. In Ethiopia, the MDD score was found to be 59.9%, positively associated with maternal educational attainment and higher household income [37]. A study reported that the prevalence of MDD ranged from 12% in Ethiopia to 38.5% in South Africa, with Uganda reporting 24% [38]. It was also noted that mothers with sufficient knowledge about dietary diversity and child nutrition had higher rates of providing MDD [37]. However, our study revealed a nonsignificant increase in MDD percentages with higher maternal education. Rakotomanana et al. found that in Madagascar, the likelihood of inadequate dietary diversity decreased when mothers completed primary school or further education [39]. Lencha et al., reported that 52.1% of included children in Ethiopia were provided with MDD, and as the mother鈥檚 education level increased, the rate of providing MDD also increased [40]. In a study conducted across nine African countries [38], grain consumption was 79.8%, legumes 49.1%, flesh foods 33.0%, eggs 13.4%, dairy 29.2%, vitamin A-rich fruits and vegetables 49.4%, and other fruits and vegetables 20.6% in Uganda. This study found that MDD prevalence was statistically higher among children whose mothers were older, had wage jobs, had higher educational levels, and read newspapers [38]. The equity gap for MDD by wealth status [the difference in prevalence (percentage points, pp) between the top and bottom quintiles] was reported to be widest in Kenya (40 pp), Zimbabwe (31.1 pp), and Rwanda (30.5 pp), while smaller gaps were observed in Uganda (15.2 pp) [38].

Our study found that 41.7% of children aged 6鈥�23 months consumed EFF, and we noted that this consumption increased with the child鈥檚 age, with no association observed with adolescent mothers. Additionally, we observed that the consumption of EFF was lower among rural children. This corresponds with the findings of Hailu et al., revealed that around 14% of urban children and 7% of rural children consumed both EFF in Sub-Saharan African countries [41].

In our study, 31.6% of children aged 6 to 23 months did not consume any vegetables. ZVF consumption was more prevalent among adolescent mothers, although no association was found between UF consumption and maternal age. The prevalence of ZVF consumption was higher among mothers over 20 years old who had a history of adolescent pregnancy. The loss of educational opportunities due to pregnancy during adolescence may have played a role in this outcome. Supporting this hypothesis, our study found that educated mothers and those engaged in family work were less likely to consume ZVF. Additionally, there was a negative association between ZVF consumption and infant age. In both Zimbabwe and Uganda poorer households were more likely to consume vitamin A-rich fruits and vegetables compared to wealthier households (鈭掆��13.3 and 鈭掆��15.2 pp, respectively; all p鈥�<鈥�0.05) [38]. Allen et al. [42] calculated the prevalence of ZVF consumption by country, region, and globally using data from 125 Demographic and Health Surveys across 64 countries conducted between 2006 and 2020, assessing whether children consumed vegetables or fruits on the previous day. They found a global prevalence of ZVF consumption at 45.7%, with the highest rate in West and Central Africa (56.1%). In Uganda, ZVF prevalence remained stable in 2006 and 2011 but showed a decline in 2016 [42]. Their findings indicated that breastfed children had lower ZVF consumption compared to non-breastfed children. Furthermore, children whose mothers had at least a secondary level of education and children of employed mothers had significantly lower odds of ZVF consumption. This may be influenced by mothers鈥� exposure to media messages, which was associated with lower ZVF consumption compared to those with no media exposure鈥攁 significant finding across all regions in this multicountry study [42].

We did not find a statistically significant effect of maternal age on UF consumption. However, our study revealed that with child ages, the mother鈥檚 level of education, welfare index, and employment status, the rate of UF consumption also increased.

In our study, three separate groups of women with children under the age of 24 months were included. Adolescent mothers currently with a baby/children, women whose first pregnancy occurred during adolescence but are currently over 20 years old, and women whose first pregnancy occurred after the age of 20 were recruited. This allowed us to assess the effects of women starting their reproductive life during adolescence on IYCF indicators compared to adulthood. In Uganda, short pregnancy intervals among young women are common, as adolescent pregnancies are prevalent [43]. The lower MMF and higher ZFV Consumption observed in Group 2 can be attributed to these mothers starting their reproductive life earlier, having a higher number of children, and consequently having less time for childcare.

Strengths and limitations

The comprehensive dataset from the 2016 UDHS provided a detailed overview of IYCF prevalent in Uganda during that period. By examining both dependent and independent variables, our study enabled a nuanced analysis, revealing patterns, associations, and potential factors influencing the nutritional landscape for infants and young children in Uganda. As a strength of the study design, we demonstrated the impact of adolescent pregnancy on IYCF indicators both during adolescence and adulthood.

To focus on breastfeeding patterns and outcomes in singletons, infants born from multiple pregnancies were excluded to ensure more consistent and comparable data. Mothers of multiples often face unique challenges, including increased demands on milk supply, time constraints, and complications with latch and positioning, making them less likely to exclusively breastfeed [44]. Adequate analysis of multiple pregnancies requires a larger sample size to capture the unique dynamics and challenges associated with breastfeeding multiples. Similar studies often exclude multiple births to maintain sample consistency and ensure reliable results [27, 45].

Conclusions

Our study indicates that despite the implementation of baby-friendly hospital practices in Uganda, there were no significant differences in the prevalence of OBF among mothers with antenatal care and hospital deliveries. This suggests a need for ongoing updates in the training of healthcare personnel and periodic reevaluations of baby-friendly hospital assessments to maintain their effectiveness in promoting optimal breastfeeding practices.

Adult mothers who had their first pregnancy before the age of 20 were found to have a lower prevalence of MFF, while the prevalence of ZVF consumption was higher. It highlights the need for closer monitoring of these mothers and the development of tailored interventions and recommendations to address the ongoing challenges they face. Special attention should be given to ensuring that these women receive adequate nutritional support and counseling, both during pregnancy and in the postpartum period, to improve their overall health and the well-being of their children.

Furthermore, we found that the impact of adolescent pregnancy extends beyond the pregnancy itself, with lasting effects on the mother鈥檚 subsequent pregnancies. This highlights the importance of providing comprehensive support and education to adolescent mothers, not only during pregnancy but also in the postnatal period and beyond.

Additionally, variations in IYCF indicators were observed based on rural residence and ethnic background. This underscores the necessity of incorporating ethnic diversity into maternal health education programs. Tailoring interventions to the unique cultural and ethnic contexts of different communities could improve the effectiveness of these programs, leading to better maternal and child health outcomes.

In conclusion, addressing the challenges associated with adolescent pregnancy, enhancing breastfeeding practices, and developing culturally sensitive maternal health education programs are critical for improving maternal and child health outcomes in Uganda. Ongoing research, continuous monitoring, and the adaptation of healthcare policies and interventions will be key to ensuring the well-being of mothers and children in the country.