Introduction

Maternal underweight is a condition caused by inadequate dietary intake, poor nutritional absorption or poor biological utilisation of nutrients.1 2 When a woman is underweight during lactation until the infant reaches two, her health as well as the future survival, growth, development and production of her offspring are at risk.3

Nutrition has an enormous effect on health throughout life.4 Thus, a healthy diet throughout the first 1000 days of life is essential for long-term physical and mental development as well as overall health.5 The WHO recommends breastfeeding for up to 2 years, including 6 months of exclusive breastfeeding,6 7 but because breast milk production is dependent on the mother’s nutritional condition, it is critical to provide appropriate nutrition among lactating women.6 7

Maternal underweight is a severe health problem worldwide8 9 and is particularly prevalent in low-income countries.10 All lactating women are at risk of malnutrition, but returnee lactating women suffer more because they have higher energy and nutrient requirements alongside the lack of easily accessible food that satisfies their needs.8 11 According to the Ethiopia Demographic and Health Survey (2016) report, 22% of women aged 15–49 years were undernourished.12 A study conducted in Dilla town depicted that the prevalence of undernutrition among lactating mothers was 23.86%.8

A number of complex conditions, including severe illness, breast feeding and a short birth interval, interact to cause maternal malnutrition.13 In turn, inadequate maternal nutrition increases the chance of miscarriage, preterm birth, low birth weight, stillbirth, overall growth impairment and developmental delay in the offspring.14 15 In addition, the health and nutritional status of women are highly linked with the overall health and nutritional status of the population.1 3

Ethiopia’s government has made a strong commitment to tackle maternal undernutrition by launching the National Nutrition Programme, which was updated and relaunched by the government in 2013, as well as the Health Sector Development Plan IV (2010–2015) and other measures.16 17 The interethnic dispute in Ethiopia displaced an estimated 800 000 individuals from the Guji and Gedeo populations. After several attempts at peace and reconciliation by the Aba Gadas, or traditional chiefs, of Gedeo and West Guji, a significant number of internally displaced people (IDPs) returned to their woredas under the direction of the government.18 Breastfeeding women in the IDP camps subsisted on the relief supplies and medical costs supplied by donor organisations. Food assistance programmes, health services and agency withdrawal from the area would all end on return. Although humanitarian aid is provided by the agencies, it is limited to individuals listed as internally displaced persons on official lists of beneficiaries.19 This would most likely compromise the dietary intake and medical care provided to lactating mothers, which would have a detrimental effect on their nutritional status. The determinants of underweight among returnees lactating mothers are different from place to place so it is still critical to research the predictors in order to prioritise, plan and implement intervention programmes based on the most up-to-date evidence. Therefore, the aim of this study was to assess determinants of underweight among returnees lactating mothers in governmental health institutions of Gedeb District, Gedeo zone, South Ethiopia 2022.

MethodsStudy area and period

The study was conducted in public health institutions in Gedeb District Southern Ethiopia. The district is located 435 km south of Addis Ababa, the capital city of Ethiopia. Recently, there was a conflict in study area and the residents were displaced to the adjacent districts and now the displaced population are returned to their homes. Based on the 2020 report of the district administrative office, the total population was 139 183 (69 649 were females and 69 534 were males), from which the total lactating mothers having children below 24 months of age accounts for (4408) 7.5% of the population. Gedeb District had 13 kebeles (the lowest administrative office). In the study area, there were a total of 17 government health facilities (1 primary-level hospital, 3 health centres and 13 health posts). On average, every month 1265 returnees lactating mothers with their children visited health facilities.

Study design

An institutional-based unmatched case–control study was carried out from 11 July 2022 to 20 September 2022. Due to resource constraints, we employed this study design. If there is a large resource cost associated with the matching process, a matching design will likely be less efficient than an unmatched one

Source population and study population

All returnees lactating mothers who visited the selected health institutions for expanded programme for immunisation (EPI) programme, postnatal care service (PNC), family planning (FP) and other services were the source population. The study population comprised randomly selected returnee lactating mothers who were visiting selected health institutions for PNC, FP, EPI programmes and other services during the data collection period.

Inclusion and exclusion criteria

For cases: Returnee lactating mothers with a body mass index (BMI) of less than 18.5 kg/m², aged 15–49 years, who are at least 2 months post partum and who have resided in the study area for 6 months or more.

For controls: Returnee lactating mothers with a BMI between 18.5 and 24.9 kg/m², aged 15–49 years, who are at least 2 months post partum and who have resided in the study area for 6 months or more.

Exclusion criteria

Cases and/or controls were excluded if they had confirmed pregnancies (pregnancy status was verified using Human Chorionic Gonadotropin tests, conducted with the consent of the head of the primary healthcare unit if there was any doubt about pregnancy status), or if they had physical deformities that made them unsuitable for anthropometric measurements.

Sample size determination

The sample size was calculated using the double population proportion formula with Epi Info V.7 software. The calculation was based on the following assumptions: a prevalence of family size greater than 5 was 49.6% among cases and 30.9% among controls.20 The calculation considered a maximum disparity of 5% between groups, 80% power, a 10% non-response rate and a 1:2 case-to-control ratio. Consequently, the final sample size was determined to be 264, consisting of 88 cases and 176 controls.

Sampling techniques

All public health institutions in the study area, including one primary-level hospital and three health centres, were included in the study. The number of cases and controls was allocated proportionally to each institution. Four weeks prior to the actual data collection, an anthropometric measurement survey was conducted at these institutions. This survey involved all returnee lactating mothers who visited for FP, PNC, EPI, and other services, and a list of these mothers was prepared. For the actual data collection, appointments were scheduled for each participant on the list. Using this list as the sampling frame, participants were selected through a simple random sampling technique.

Dependent variable: Underweight (case/control).

Independent variables: Sociodemographic factors such as age, place of residence, marital status, types of marriage, educational status, occupation, family size, wealth index. Obstetric and reproductive factors such as number of under-5 children, parity, age at first pregnancy, type of pregnancy, antenatal care (ANC) follow-up and frequency of ANC, PNC follow-up, place of delivery, age of current breastfeed child, birth interval from the preceding child. Feeding and dietary habit factors: number of meals per day, household food insecurity and dietary diversity.

Data collection tools and procedures

Data were collected using a structured questionnaire after reviewing different literature.7 16 21 22 The questionnaire consists of sociodemographic characteristics, obstetric and other characteristics of study participants. Data collector team comprise eight diploma nurses and two supervisors.

Anthropometric measurement

A weight scale and a height board were used to measure the weight and height of lactating women, respectively. A calibrated scale accurate to 0.1 kg was used to weigh lactating women, and a portable board accurate to 0.1 cm was used to measure their height. For height measurements, participants were instructed to remove any head coverings, stand upright and barefoot and ensure their heads, shoulders, buttocks, knees and heels were in contact with the height board. Prior to weight measurements, the scale was calibrated to standard settings. Participants were asked to remove extra clothing and wear only light clothing. Weight measurements were taken while barefoot. All measurements were performed in duplicate to ensure accuracy, and the average value was recorded.

Data quality management

A structured questionnaire was initially prepared in English and then translated into the local languages, Amharic and Gedeoffa. Data collectors underwent 2 days of training on data collection procedures. A pretest was conducted on 5% of the sample size at a health centre in an adjacent district. Calibration of instruments was performed before each measurement. To ensure data reliability, the investigator and field supervisors repeated some anthropometric measurements daily on randomly selected participants

Operational definitions

Underweight returnees lactating mothers: Mothers whose BMI is less than 18.5 kg/m2.8

Normal weight returnees lactating mothers were mothers whose BMI is between 18.5 and 24.9 kg/m2.3

Returnees lactating mothers: Mothers who were breastfed their infant or child less than 24 months of age and returned from internal displacement.

Controls: Returnees lactating mothers who had normal weight.

Cases: Returnees lactating mothers who were underweight.

Inadequate Dietary Diversity Score (DDS): When a study participant consumed less than 5 food groups among the 10 food groups within 24 hours.23

Index child: The most recent child born to a lactating mother who is participating in the study. Specifically, the index child is defined as the child who is under 24 months of age and is the subject of the mother’s current lactation period. This child serves as the focal point for assessing the maternal nutritional status and its associated factors, with the assumption that the mother’s health and nutritional practices during this period directly impact the well-being of the index child. The selection of the index child is based on the mother’s current lactation status, ensuring relevance to the study’s focus on maternal underweight among lactating women.

Adequate DDS: When a study participant consumed 5 or more food groups among the 10 food groups within 24 hours.23

Food secure household: Experiences none of the food insecurity (access) conditions or just experiences worry, but rarely.24

Food insecure household: If the household encounters any of the circumstances (uncertainty, inadequate food quality and quantity) during the recall time. (If any of the questions had a ‘rarely’, ‘sometimes’ or ‘frequently’ response. The only exception is among households where the respondent answered ‘rarely’ to question one but ‘never’ to all other questions.24

Data analysis and processing

The collected data were reviewed, coded and entered into EpiData V.4.6 software, then exported to SPSS (Statistical Package for Social Sciences) V.23 for analysis. Model fit was assessed using the Hosmer and Lemeshow test. Variables with a p≤0.25 in the bivariable logistic regression analysis were selected for multivariable logistic regression to control for potential confounders. In the multivariable logistic regression, predictor variables with a p≤0.05 at a 95% confidence level were identified as determinants of maternal underweight.

Data analysis procedure for socioeconomic status

Wealth index: Principal component analysis was conducted using 16 dichotomous variables, where each variable was scored as ‘1’ for possession and ‘0’ for non-possession. The summed scores were classified into tertiles: low, medium and high.

Food security status: Food security was assessed using 18 questions adapted from FANTA V.3, with a recall period of 4 weeks (30 days). Participants were first asked if the condition occurred in the past 4 weeks (yes or no). If ‘yes’, they were further asked how often the condition occurred: rarely (once or twice), sometimes (3–10 times) or often (more than 10 times). Scores were classified as either food secure or insecure.25

DDS: DDS was calculated by summing the number of unique food groups consumed by the mother on the day preceding the survey. Using the 10 categorised food groups from the latest Women’s DDS recommendations.23 Each food group consumed contributed 1 point, with a maximum score of 10. A score of 5 or more food groups indicated good dietary diversity while consuming fewer than 5 food groups was classified as low dietary diversity.23

Patient and public involvement

This study did not involve direct patient or public participation in the design, conduct or dissemination phases. The research was conducted without engagement from patients or the public, focusing solely on analysing secondary data collected from health records and surveys. While the study did not include patient input, the findings aim to contribute to improving healthcare policies and practices that could benefit patient outcomes in the future.

ResultCharacteristics of the study participants

A total of 264 lactating women were initially recruited for the study, achieving a high response rate of 255 participants (96.59%). Among these, 76 (90.5%) of the case group and 157 (91.8%) of the control group were aged between 20 and 34 years. The median age of the participants was 28 years, with an IQR of 7 years. Approximately 49 (58.3%) of the cases and 93 (54.4%) of the controls resided in rural areas. Nearly all participants were married, with 77 (91.7%) of the cases and 168 (98.2%) of the controls being married. Additionally, 33 (39.3%) of the cases and 62 (36.3%) of the controls were housewives (table 1).

Table 1

Sociodemographic characteristics of returnees lactating mothers visiting government health facilities in Gedeb District, Southern Ethiopia, 2022

Reproductive and obstetric characteristics of women

Regarding the birth interval between children, the majority of cases, 43 (51.2%), had a short birth interval of less than 24 months from the preceding child. In contrast, most controls, 145 (84.8%), had a birth interval of 24 months or more. Concerning the place of delivery, more than half of the cases, 48 (57.1%), delivered their index child at a health facility, while almost all controls, 148 (86.5%), delivered at a health facility (table 2).

Table 2

Reproductive and obstetric characteristics of returnee lactating mothers visiting government health facilities in Gedeb District, Southern Ethiopia, 2022

Socioeconomic and dietary diversity characteristics

The majority of cases, 45 (53.6%), had inadequate DDSs, compared with approximately one in five controls, 36 (21.1%). Regarding food security, about 71 (84.5%) of the case households and 105 (61.4%) of the control households experienced moderate food insecurity. Additionally, a higher proportion of cases, 46 (54.8%), were from poor socioeconomic backgrounds compared with 25 (14.6%) of the controls (table 3).

Table 3

Dietary diversity, food security and socioeconomic status of returnee lactating mothers visiting government health facilities in Gedeb District, Southern Ethiopia, 2022

Determinants of underweight among returnees lactating mothers

In the bivariable logistic regression analysis, several factors were associated with maternal underweight, including the age of the index child, age at first pregnancy, consumption of additional food during lactation, birth interval, place of delivery, type of pregnancy, number of meals per day, household food security, wealth index and dietary diversity. Variables with a p≤0.25 were selected for multivariable logistic regression analysis.

The multivariable logistic regression revealed several significant determinants of maternal underweight. Returnee lactating mothers who gave birth to the index child within 24 months of the preceding child were found to be four times more likely to be underweight (adjusted OR (AOR) 4.04, 95% CI 1.68 to 9.74) compared with those with a birth interval of over 24 months. Mothers who had an unplanned pregnancy for the index child were 7.6 times more likely to be underweight (AOR 7.60, 95% CI 3.14 to 18.41) than those with a planned pregnancy.

Regarding the place of delivery, the odds of being underweight were 5.81 times higher for those who delivered at home compared with those who gave birth at health facilities (AOR 5.81, 95% CI 2.09 to 16.14). Additionally, lactating mothers who consumed fewer than three meals a day were 2.51 times more likely to be underweight (AOR 2.51, 95% CI 1.12 to 5.63) compared with those who ate three or more meals daily.

In terms of dietary diversity, mothers with inadequate dietary diversity had 3.92 times higher odds of being underweight (AOR 3.92, 95% CI 1.52 to 10.15) compared with those with adequate dietary diversity. The likelihood of being underweight was 4.72 times greater for mothers living in food-insecure households compared with those in food-secure households (AOR 4.72, 95% CI 1.77 to 12.61). Finally, returnee lactating mothers from poor and medium wealth index households were 4.38 times more likely to be underweight than those from rich wealth index households (AOR 4.38, 95% CI 1.49 to 12.82) (table 4).

Table 4

Bivariate and multivariate analysis of determinants of underweight among returnee lactating mothers in Gedeb District, South Ethiopia, 2022

Discussion

This study investigated determinants of underweight among returnee lactating mothers in Gedeb District, South Ethiopia. Using a case–control design, it aimed to identify factors contributing to maternal underweight in a context marked by recent conflict and displacement. Our analysis reveals significant associations between maternal underweight and various sociodemographic, reproductive and nutritional factors.

Our results show that among returnee lactating women, the place of delivery considerably increases the likelihood of being underweight. The risk of underweight was more than five times higher for women who gave birth at home compared with those who did so in health facilities. This result emphasises how important institutional delivery is to ensuring greater benefits for maternal health. Mothers giving birth at home often do not receive enough medical support, which can put them at risk for both short-term and long-term nutritional problems. The likely explanation is that institutional deliveries often come with better access to nutrition and health education, which can positively influence maternal health behaviours. In contrast, home births may lack such support, increasing susceptibility to malnutrition due to inadequate information and potential exposure to infections. This observation is consistent with previous studies in Dangila and Wonberma woreda, which also found that home deliveries were associated with higher rates of maternal underweight.26 27

Unplanned pregnancies also emerged as a significant risk factor for maternal underweight. Women with unplanned pregnancies were over seven times more likely to be underweight than those with planned pregnancies. This finding aligns with the hypothesis that unplanned pregnancies can lead to inadequate preparation and support, both psychologically and economically. Conversely, unintended pregnancies may cause underweight by encouraging unsafe behaviours including skipping or using insufficient prenatal care, causing more stress and despair and reducing social support during lactation.

The effect of birth interval on maternal underweight was also highlighted by this study. Women who had a short birth interval were more likely to be underweight, a finding that is in line with research from the Afar region and Arba Minch.28 29 Short intervals between pregnancies do not allow sufficient time for maternal nutritional recovery, exacerbating the strain on maternal health. This finding supports the maternal nutritional depletion hypothesis, which suggests that frequent pregnancies without adequate recovery periods contribute to nutritional deficiencies.30

Household food insecurity is also positively associated with maternal underweight. This finding is consistent with previous studies done in Ghana, Borena zone, Arbaminch, Ethiopia.16 28 31 This probably because households with food insecurity might not be able to access sufficient food and limited availability of food result in underweight. Furthermore, in the study area, there was recent internal displacement due to conflict. So, most of the participants in this study were food insecure.

Moreover, returnees lactating mothers who had inadequate DDSs were risky for underweight than lactating mothers who had adequate dietary diversity. This finding is also supported by studies in Dire Dawa, Degadamot and Afar.2 3 29 This might be because women with low dietary diversity could not get diversified diets. In addition to these, lactating women who do not get enough energy and nutrients in their diets faced risks for nutrient depletion that increases the risk of being underweight.

The frequency of meals per day was another factor associated with maternal underweight. Mothers who ate less than three times daily were more likely to be underweight compared with those who consumed three or more meals. This finding aligns with studies from Moyale and Raya, Alamata,16 32 highlighting that lactating mothers require frequent meals to meet their increased nutritional needs during lactation.

Socioeconomic status, as indicated by wealth index, was also a significant determinant. Mothers from poor and medium wealth index households were more likely to be underweight compared with those from wealthier backgrounds. This result is consistent with studies from Uganda, Tanzania and Ethiopia,1 14 33 suggesting that economic constraints limit access to nutrient-dense foods, thereby increasing the risk of underweight. This could be due to the fact that women with low economic status unable to buy different foods which are dense in nutrient contents that prevent lactating mothers from being underweight.

This study has several notable strengths, primarily its focus on socially stigmatised and often overlooked populations, specifically returnee lactating mothers. By zeroing in on this vulnerable group, the research provides crucial insights into their unique challenges and nutritional needs, which are frequently neglected in broader nutritional studies. Returnee lactating mothers, who have faced displacement and resettlement, represent a group with specific difficulties that are not typically addressed in general maternal health research. This targeted approach enables a deeper understanding of the factors contributing to maternal underweight in this context.

Additionally, the study’s design incorporated robust methodologies for data collection and analysis, including precise anthropometric measurements and comprehensive questionnaires on dietary diversity and food security. The use of multivariable logistic regression allowed for the identification of significant determinants of maternal underweight, adjusting for potential confounders. This methodological rigour enhances the validity of the findings and provides a strong foundation for formulating evidence-based interventions.

The focus on returnee lactating mothers also offers valuable information for policy-makers and health planners. The study’s results highlight critical areas that need attention, such as birth spacing, dietary diversity and socioeconomic factors. By addressing these issues, stakeholders can develop targeted interventions aimed at improving the nutritional status and overall health of this underserved population.

Despite its strengths, the study has several limitations. The primary limitation is the reliance on BMI as the sole measure of nutritional status. While BMI is a useful screening tool, it does not capture all aspects of nutritional health, such as micronutrient deficiencies or body composition changes. Incorporating additional measures, such as biochemical assessments or detailed dietary surveys, could provide a more comprehensive view of maternal nutritional status.

Another limitation is the potential recall bias associated with the 24-hour dietary recall method. Despite efforts to minimise this bias, the accuracy of dietary data can be affected by the participant’s memory and reporting accuracy. To mitigate this, the study employed probing questions and cross-checked data, but recall bias remains a concern. A notable limitation is the relatively small sample size for cases, with only 88 participants, which may limit the generalisability and statistical power of the results. Future studies should aim to include larger sample sizes and use additional methods to comprehensively assess nutritional status.

Additionally, the study’s cross-sectional design limits the ability to infer causality. While associations between various factors and maternal underweight were identified, establishing a causal relationship requires longitudinal studies. Future research with longitudinal designs could provide more definitive evidence on the causal pathways leading to maternal underweight.

Lastly, using unmatched case–control study design might raise a limitation. In the absence of matched cases and controls, there may be significant differences that might affect the results. It may be more difficult to identify real association when there is a lack of matching since it increases data variability More individuals may also be needed to achieve the same statistical power compared with matched studies, especially if there is higher heterogeneity among controls. Matching can enhance precision in estimating the association; unmatched studies may lack this advantage, leading to less reliable estimates. Overall, while unmatched case–control studies can be useful, their design necessitates careful consideration of potential biases and confounding factors. Taking these limitations into consideration, we advise that future research use matched control studies in order to provide statistically significant results.

The findings from this study underscore the urgent need for targeted interventions to address the factors contributing to maternal underweight among returnee lactating mothers in Gedeb District. The significant associations between underweight and factors such as short birth intervals, unplanned pregnancies, inadequate dietary diversity and household food insecurity highlight areas where intervention can make a meaningful impact.

To address these issues, it is essential for Ethiopian policy-makers and stakeholders to implement comprehensive strategies. These strategies should focus on improving birth spacing through FP education and services, promoting dietary diversity by enhancing access to varied and nutritious foods and addressing food insecurity by supporting economic and food access programmes. Additionally, providing support for planned pregnancies through education and healthcare access can help reduce the incidence of underweight among lactating mothers.

Special attention should be given to returnee lactating mothers, considering their unique postdisplacement challenges. Programmes should be designed to provide adequate nutritional support, healthcare services and socioeconomic assistance to this population. Collaboration between government agencies, non-governmental organisations and community-based organisations can facilitate the development and implementation of effective interventions.

Overall, a multifaceted approach that includes better healthcare access, nutritional support and socioeconomic development is essential for improving maternal health outcomes. By addressing the identified determinants of underweight, stakeholders can help enhance the well-being of returnee lactating mothers and contribute to broader maternal and child health goals.