Determinants of Relapse of Severe Acute Malnutrition Among Children Under Five in Mbujimayi Town, Democratic Republic of Congo

Introduction

Severe acute malnutrition (SAM), characterized by severe wasting or nutritional edema, remains a major public health challenge among children under five years of age in low- and middle-income countries.^1–4 Children affected by SAM face a markedly increased risk of mortality, approximately 9–12 times higher than that of their well-nourished peers.^5–10 In 2020, an estimated 45.4 million children under five worldwide were affected by acute malnutrition, including 31.8 million with moderate acute malnutrition (MAM) and 13.6 million with the server acute malnutrition.^1,3,5,7

In the Democratic Republic of Congo (DRC), the burden of child malnutrition remains high, particularly in provinces dependent on mining industries and conflict affected eastern regions.^10,11 The 2023–2024 Demographic and Health Survey (DHS) reported that 7% of children under five years of age were wasted, 45% were stunted, 25% were underweight, and 4% were overweight.¹² Similarly, the 2017–2018 Multiple Indicator Cluster Survey (MICS) found a prevalence of 6.5% for acute malnutrition and 49.6% for chronic malnutrition, ranking malnutrition among the leading causes of death in children under five.¹³ National nutrition surveys have further revealed that the prevalence of global acute malnutrition among children under five remains above the emergency threshold of 15%. While remaining at a very high threshold among children under five years of age, severe acute malnutrition dramatically increases the risk of mortality within this age group.¹⁴ Poor infant and young child feeding (IYCF) practices exacerbate the burden of malnutrition in the DRC. Exclusive breastfeeding rates remain low (53%), and dietary diversity is inadequate; 92% of children aged 6–23 months lack access to diverse foods. Limited access to food, in terms of frequency, quantity, and variety, impedes healthy growth and development.¹⁵

Relapse after discharge from treatment for SAM has emerged as an important programmatic and clinical issue. Studies from Ethiopia, Mali, Pakistan, Niger, and Burkina Faso have shown that relapse may be influenced by anthropometric status at discharge, previous malnutrition, infections, household vulnerability, seasonality, inadequate follow-up, and suboptimal quality of post-discharge care.^{1,2,7,16–20} These findings are consistent with theoretical work suggesting that relapse results from the interaction between residual biological vulnerability at discharge and the child’s post-treatment environment.

Emerging evidence indicates that some children recovering from SAM continue to experience poor health and nutritional outcomes after discharge from integrated management of acute malnutrition (IMAM) programs, with some relapsing into SAM after a period of recovery.^1,21,22 Recurrent episodes of SAM not only increase the risk of mortality and long-term developmental impairments but also contribute to persistently high global malnutrition rates. Relapses undermine the effectiveness of IMAM programs and strain-limited resources when the same child requires repeated treatment.^1,23,24

Despite decades of intervention, the prevalence of acute malnutrition in the DRC has not declined significantly. It currently stands at 6.5% from 10%. Severe forms affect approximately two million children aged 6–59 months, increasing their risk of death by four- to nine-fold. In Kasai-Oriental, the prevalence of chronic malnutrition prevalence is 43% and 13%, respectively.

Data on SAM management performance in the DRC remain scarce. Multiple aggravating factors, including poor dietary practices, acute food insecurity, a high prevalence of childhood illnesses (malaria, respiratory infections, and diarrheal diseases), inadequate hygiene and sanitation, limited access to safe drinking water, and insecurity leading to mass population displacement, contribute to repeated episodes of acute malnutrition. These challenges underscore the urgent need for tailored interventions to improve the quality of care.

The present study aimed to identify the determinants associated with the relapse of severe acute malnutrition among children aged 6–59 months in Mbujimayi, within the Muya health zone. By examining the factors contributing to relapse, this study sought to strengthen the quality of SAM management and reduce the recurrence of cases, thereby improving child survival and developmental outcomes in the DRC.

Methods

Study Design and Setting

This study employed an analytical cross-sectional design to investigate factors associated with relapse of severe acute malnutrition (SAM) among children aged 6–59 months in the Muya health zone. Data were collected simultaneously between May 2023 and April 2024. Within the study population, two groups were identified at the time of survey: children readmitted for SAM relapse after discharge from outpatient therapeutic nutrition units and children newly admitted for SAM without prior relapse. The design enabled the assessment of associations between demographic, social and programmatic variables and the occurrence of relapse, using appropriate statistical methods to estimate adjusted measures of association.

Study Population and Eligibility Criteria

The study population included children aged 6–59 months managed for SAM in all health facilities of the Muya Health Zone with outpatient therapeutic nutrition units. Two groups were considered at the time of data collection: children readmitted with SAM following discharge (relapse group) and children newly admitted with SAM during the same study period without documented relapse (new admission group). Children were excluded when age was outside the eligibility range, when the diagnosis of SAM could not be verified from clinical records, or when records lacked essential information on relapse status or key analytical variables.

Sampling and Group Ratio

Sampling was exhaustive and facility-based. All eligible relapse cases identified during the study period were included consecutively. New admissions were selected from the same facilities and time frame using a 1:2 ratio relative to relapse cases. This ratio was applied to enhance statistical precision while ensuring comparability of source populations.

Sample Size

The minimum sample size was estimated using Epi Info 7 (StatCalc) assuming a 95% confidence interval, 80% power, a 1:2 ratio between relapse and new admissions, expected relapse proportion of 2%, and an estimated odds ratio of 5. With a 5.5% non-response rate. The sample size was 267 children, comprising 89 relapse cases and 178 new admissions.

Variables and Conceptual Approach

The dependent variable was relapse of SAM. Independent variables were selected a priori from published literature and a conceptual framework on post-SAM relapse and grouped into: (1) child characteristics (age, sex, anthropometry, edema, clinical form of malnutrition, malaria, diarrhea, vaccination status, treatment duration, and outcome of previous care); (2) caregiver and household factors (age, education, occupation, household size, income, food access, water and hygiene conditions); and (3) program-related factors (availability and use of nutritional inputs, nutrition education, and community follow-up).

Data Collection

Data were collected from structured questionnaires administered to caregivers and from outpatient therapeutic unit records. Information covered sociodemographic characteristics, clinical history, nutritional status, treatment history, and household conditions. Data collection tools were reviewed before field use, and completed forms were checked for consistency and completeness. Where discrepancies existed between caregiver report and facility documentation, the clinical record was used as the primary source for case classification.

Data Management and Statistical Analysis

Qualitative variables were summarized as frequencies and percentages. Continuous variables were first described using medians with interquartile ranges or means with standard deviations, then categorized for bivariate and multivariable analyses using clinically meaningful thresholds or distribution-based groupings (for example, MUAC < 115 mm vs ≥ 115 mm, and caregiver age categories). Records with missing outcome information were excluded from the analysis; the final analytical dataset was based on complete available records for the variables retained in the model. Bivariate associations between independent variables and relapse were tested using Pearson’s chi-square test or Fisher’s exact test (when expected cell counts were < 5) and crude odds ratios (cOR) were calculated. Variables with p < 0.20 in bivariate analysis were entered into multivariable logistic regression to identify independent predictors of relapse. Adjusted odds ratios (aOR) with 95% confidence intervals (CI) are reported, and statistical significance was set at p < 0.05.

Ethical Consideration

Ethical authorization was sought from by the ethics committee of the Official University of Mbujimayi, registered in number 004/2024/CERUOM/NKL. Free and informed consent was used in both French and Tshiluba (the local national language) and was given/read to participants before the questionnaire was administered. For child participants, a parental authorization form was drawn up and made available to parents to give them access to the study. This consent form must respect the principles of all medical research, ie., the protection of human life, health, dignity, integrity, the right to self-determination and the confidentiality of the information of the persons involved in the research. The ethical regulations outlined in the Declaration of Helsinki were followed throughout the study. Data collection began only after the ethics committee had given its approval. The data were stored on a laptop computer with a password-protected and locked cabinet, and only the lead investigator and study supervisor had access to the data.

Results

Participant Characteristics

A total of 267 children were included, comprising 89 relapse cases (33.3%) and 178 new admissions (66.7%). The median age of the children was 24 months (IQR: 15–40), and 56.6% were aged 24–59 months. Boys represented 52.8% of participants. Most children had a birth order between the third and fifth child in the household (67.0%). A documented history of acute malnutrition was reported for 195 children (73.0%). Among these previously affected children, 32.8% had correctly completed prior treatment and 28.2% had a history of initial treatment abandonment. Measles vaccination coverage was 53.6%. Anthropometric indicators showed that 75.3% had MUAC < 115 mm and 55.1% had a weight-for-height Z-score < - 3 SD. Marasmus was the predominant clinical presentation (88.4%), whereas malaria and diarrhea were identified in 39.0% and 7.5% of children, respectively (Table 1).

Table 1 Sociodemographic and Clinical Characteristics of Children

Caregivers had a median age of 31 years (IQR: 27–37), and 53.2% had primary level education only. Most households were large (> 6 persons, 73.8%), had low monthly income (≤ 100,000 CDF, 85.4%), and reported only one meal per day (62.5%). Although 86.9% reported access to safe drinking water, 64.8% did not use water purification methods and 80.9% lacked a handwashing facility (Table 2).

Table 2 Sociodemographic Characteristics of Caregivers/Households

Factors Associated with Relapse of SAM

In bivariate analysis, parental occupation, previous history of malnutrition, measles vaccination status, MUAC at admission, and diarrhea at admission were significantly associated with relapse. Variables meeting the threshold for multivariable modeling were household size, parental occupation, history of malnutrition, measles vaccination, MUAC, edema, malaria, diarrhea at admission, community follow-up, and clinical outcome. After adjustment, the strongest predictor of relapse was a previous history of SAM (aOR: 9.847; 95% CI: 3.619–26.792; p < 0.001). Children without measles vaccination had threefold higher odds of relapse (aOR: 3.120; 95% CI: 1.657–5.873; p < 0.001). Malaria (aOR: 1.910; 95% CI: 1.012–3.605; p = 0.046) and diarrhea at admission (aOR: 3.690; 95% CI: 1.143–11.921; p=0.029) were also independent predictors. In addition, an unfavorable clinical outcome after the previous episode was associated with higher odds of relapse (aOR: 3.660; 95% CI: 1.215–11.026; p=0.021). These results suggest that relapse is shaped by both baseline vulnerability and insufficiently stabilized recovery after treatment (Table 3).

Table 3 Simplified Analysis of Factors Associated with Relapse of SAM

Discussion

This study aimed to improve the quality of care for severe acute malnutrition (SAM) by reducing the number of relapse cases and identified several determinants of relapse of severe acute malnutrition among children under five in Mbujimayi, Democratic Republic of Congo.

Relapse after treatment for SAM remains a documented concern across diverse contexts, with reported frequencies ranging from 0% to 37% depending on setting and follow-up duration.^{1,3,16–20,25–29} Evidence from Nepal,¹⁷ Mali,¹⁶ Pakistan,¹⁹ Niger,²⁰ Burkina Faso,¹⁸ Ethiopia,^1,2,28,30,31 and DRC³ highlights variability in relapse incidence, while surveys in observations reinforce the understanding that relapse in driven by residual biological vulnerability, ongoing exposure to infection, and the adequacy of post-discharge support.

In this study, relapse of SAM was associated with multiple demographic, familial and clinical factors. The predominance of children aged 24–59 months, with relapse more frequent among males, aligns with evidence from Lubumbashi and other African contexts, while contrasting report highlight variability across settings.^{1,3,17,26,30,31} Family characteristics such as intermediate birth order, prior episodes of malnutrition, and incomplete adherence to treatment underscore the role of household dynamics and continuity of care.

Socioeconomic vulnerability was evident, with most caregivers engaged in non-remunerated activities, low household income, limited food security, and inadequate hygiene practices. These determinants, consistent with findings from Ethiopia and Pakistan, reinforce the importance of addressing structural poverty and strengthening community-based support systems.^1,3,25,31

Clinical presentation was dominated by marasmus, with malaria and diarrhea as frequent comorbidities. Admission and discharge criteria were largely respected, yet gaps in post discharge follow-up and nutrition education were reported. The associations identified parental occupation, history of SAM, measles vaccination, MUAC at admission, and diarrhea highlight the multifactorial nature of relapse and the need for integrated interventions combining medical, nutritional, and social support.

The analysis identified several key factors associated with relapse of SAM among children. A prior history of SAM emerged as a strong predictor, indicating that children who had previously experienced the condition remained particularly vulnerable to recurrence. Lack of measles vaccination was also significantly associated with relapse, underscoring the importance of preventive health measures in sustaining nutritional recovery. Malaria and diarrhea at admission were independent predictors, reflecting the close interplay between infectious disease and nutritional status. Notably, children discharged as clinically cured were more likely to relapse, suggesting that current discharge criteria may not fully capture residual risk and that closer post-discharge monitoring may be warranted.

These associations are biologically and programmatically plausible. Children with repeated episodes may remain nutritionally fragile, live in persistently high-risk households, or present underlying vulnerabilities insufficiently corrected during prior treatment. Comparable findings have been reported in Ethiopia and Nepal, where poor nutritional status at discharge and prior episodes of acute malnutrition increased relapse probability.^1,17 The association with measles vaccination underscores the role of preventive child health services in sustaining recovery, as vaccination not only provides immunological protection but also reflects caregiver engagement with routine health services.^16,27

Malaria and diarrhea emerged as independent predictors, reinforcing the bidirectional relationship between infection and malnutrition. Infectious episodes increase metabolic demand, reduce appetite, exacerbate nutrient losses, and accelerate return to wasting, particularly in children whose recovery remains incomplete.^18,26 This interpretation is consistent with theoretical frameworks emphasizing post-discharge morbidity as a proximal driver of relapse.²⁷

From a programmatic perspective, these findings highlight the need to strengthen verification of discharge readiness,^16,17 ensure complete vaccination coverage, systematically screen and treat malaria and diarrhea,^1,26 reinforce caregiver counseling on infant and young child feeding (IYCF) practices,¹ and establish structured post-discharge follow-up through community health workers.^16,21 In urban contexts such as Mbujimayi, relapse prevention should be integrated into broader child survival services including vaccination, infection control, and food security rather than being addressed as a nutrition-only issue.^20,28,31

Study Limitations

This study has several limitations. First, the analytical cross-sectional design identifies associations but cannot establish temporal causality, as exposures and outcomes were measured at the same time. Future longitudinal or cohort studies would be better suited to estimate incidence and the timing of relapse. Second, some variables were based on caregiver recall and routine health records, which may have introduced misclassification or incomplete reporting. Third, the study was conducted in a single health zone, which limits the generalizability of the findings to other urban or rural contexts. Finally, household indicators were measured in broad categories, which may not fully capture the complexity of socioeconomic vulnerability. These limitations were taken into account when interpreting the results.

Conclusion

Relapse of SAM among children under five in Mbujimayi was frequent and was independently associated with previous SAM, lack of measles vaccination, malaria, diarrhea at admission and unfavorable clinical outcome after a prior episode. The findings suggest that relapse prevention requires more than facility based nutritional treatment alone. Strengthening discharge quality, immunization, infection prevention and management, caregiver counseling, and organized post-discharge follow-up could reduce repeated admissions and improve child survival in this setting.