Access to safe drinking water, adequate sanitation, and appropriate hygiene practices (WASH) is a critical determinant of child health and well-being.1,2 Inadequate WASH access contributes to infectious diseases, diarrhea, and environmental enteropathy-conditions that impair nutrient absorption and child growth.3,4 Despite progress, nearly two billion people worldwide still consume contaminated drinking water, and 3.5 billion lack safely managed sanitation services.5,6
Globally, 42.8 million children under five were affected by wasting in 2024, representing approximately 6.6% of children under five years of age worldwide.7 In the Democratic Republic of Congo (DRC), the third Demographic and Health Survey (DHS III, 2023–2024) reported a national prevalence of wastingof 7.2% and WASH access remains critically low, with only 30% of households using improved water sources and 20% accessing improved sanitation.8 The peri-urban Mont Ngafula II health zone faces overcrowding, poor infrastructure,9 and frequent outbreaks such as cholera.
Most previous studies have relied on isolated WASH indicators to examine their association with child nutrition3,10–13 limiting understanding of cumulative exposure pathways. Composite indices provide a more integrated measurement. Evidence from Asia suggests that child-sensitive WASH scores may predict nutritional outcomes,14,15 but such approaches remain scarce in Sub-Saharan Africa.
Assessing the relationship between the child-sensitive WASH score and wasting is therefore crucial. This study aimed to examine the relationship between the WASH composite score and the prevalence of wasting in the Mont Ngafula II, providing context-specific evidence for targeted WASH and nutrition interventions.
METHODS
Study design and setting
A community-based cross-sectional survey was conducted from 25 November to 12 December 2024 in Mont Ngafula II, a peri-urban area with persistent poor child feeding practices (infant and young child feeding, IYCF), inadequate sanitation, acute food insecurity, and high prevalence of childhood illnesses.9 In July 2025, the zone experienced a cholera outbreak that affected 60 children, resulting in 16 deaths.
Study population and sampling
We surveyed households in Mont Ngafula II health zone that had at least one child aged 6–59 months. Inclusion criteria required that households had resided in Mont Ngafula II for at least six months and included at least one eligible child. The respondents were mothers or primary caregivers of these children.
The sample size of 415 children was determined using Cochran’s formula (n = z²pq/d²), where p represents the proportion of households using at least one basic drinking water service (43%, DHS 2023–2024), z = 1.96 for a 95% confidence level, and d = 5% margin of error. A 10% non-response rate et and a design effect of 1 were applied.
A three-stage probability sampling strategy was employed. In the first stage, five health areas (HAs) (Kimbwala, Matadi-Kibala, Matokama, Mitendi, Sans Fil) were randomly selected from the Mont Ngafula II sampling frame. In the second stage, within each HA, enumeration areas (EAs) were randomly selected, proportionally to their size. Finally, in the third stage, within each EA, target households were systematically identified and selected. A single eligible child aged 6 to 59 months was selected for nutritional status assessment.
Data collection and processing
A structured questionnaire, translated into Lingala and back-translated into French, captured sociodemographic characteristics (mother’s age, education level, occupation, marital status, child’s age and sex, household size, household wealth index, health area), WASH practices (based on JMP standards), and child anthropometry. Weight and height were measured using calibrated equipment following WHO guidelines.
Outcome variables
WASH indicators et composite score
WASH data were reported by mothers or primary caregivers, following the standard Joint Monitoring Programme (JMP) categorization.16 Eleven indicators were included: three related to water, five to sanitation, and three to hygiene (Table 1). Following the method described by Jubayer et al., we constructed a WASH composite score using 11 household-level variables on WASH practices.14
Receiver Operating Characteristic (ROC) curve analysis was performed to determine the optimal WASH score cut-off predicting wasting, maximizing both sensitivity and specificity. For the anthropometric indicator weight-for-height-Zscore (WHZ), a threshold of ≥6 was identified as appropriate for predicting nutritional status. Thus, households were categorized as having “improved WASH” (score ≥6) or “unimproved WASH” (score<6).
Nutritional status
Weight and height were measured for all eligible children. Weight was recorded without clothing or shoes using a digital SECA scale (Germany), with 100 g precision. For children under two years, the 2-in-1 function of the SECA scale allowed accurate measurement with maternal or caregiver support. Height (recumbent or standing, precise to 1 mm) was measured with a UNICEF-provided portable wooden stadiometer. Two field staff assisted: one took the measurement, while the other ensured correct child positioning. Weight-for-height-zscore (WHZ) values were generated using WHO Anthro software (version 3.3.2). WHZ values >+5 or <–5 SD were considered biologically implausible and excluded, as per WHO guidelines.17 Wasting was defined as WHZ < –2 SD.
Data analysis
Data exported from Kobo in Excel format were analyzed in STATA version 18. Descriptive statistics were calculated for household, mother, and child characteristics. Normality of continuous variables was assessed using the Shapiro–Wilk test. Bivariate associations between wasting and WASH practices, as well as between sociodemographic characteristics and the WASH composite score, were tested using Pearson’s chi-square. Variables with a p-value ≤ 0.10 in the bivariate analysis were retained for the multivariate modeling.
A robust binary logistic regression model with standard errors [vce(robust)] was used to identify factors associated with an unimproved WASH composite score. Multicollinearity was assessed using the uncentered variance inflation factor (VIF); the mean VIF of 1.57 indicated no significant collinearity. Adjusted odds ratios (ORa) with 95% confidence intervals (95% CI) were reported. Overall model goodness of fit was assessed using the Hosmer-Lemeshow goodness-of-fit test, and its specification was verified by the linkage test. Model performance was evaluated using sensitivity, specificity, positive and negative predictive values, and the overall percentage of correct classifications. Finally, a linear regression model was used to assess the robustness and reliability of the association between wasting and the WASH composite score.
Ethical Considerations
Ethical approval was obtained from the Ethics Committee of the Kinshasa School of Public Health (authorization No. ESP/CE/37B/2024). Written consent was obtained from all participants.
RESULTS
Household, mother, primary caregiver, and child characteristics
A total of 406 households were analyzed after excluding missing or outlier values (n = 8). Households had a mean size of 6 persons; half were classified as poor. Mothers had a mean age of 31 years and were mostly married and educated to secondary level.Wasting prevalence was 11.6% (95% CI: 8.6-14.6). (Table 2)
Household WASH practices
As shown in Table 3, although 96.8% of households used improved water sources, only 52% had water available when needed, and 19% spent >30 min collecting water. Improved sanitation was reported by 63.6%, but most facilities were shared and located outside the home. While 83.7% had a handwashing facility, soap was available in only 71%.
Relationship between WASH practices and wasting
As shown in Table 4, wasting was significantly associated with unimproved latrines (p = 0.011), inadequate disposal of child feces (p = 0.049) and lack of water (p = 0.010) or soap (p = 0.021) in handwashing facilities. No water-related indicators were associated with wasting.
Association between socio-demographic characteristics overall according to WASH conditions
Most sociodemographic characteristics were not associated with households WASH conditions (p > 0.05), although poor households and non-working mothers were slightly more frequent in the “Unimproved” group (p ≈ 0.05). Only wasting was significantly associated among children in households with unimproved WASH facilities (p = 0.008) (Table 5).
Multivariate Analysis Between Composite WASH Score and Wasting
Adjusted analyses incorporating variables with p≤0.10, confirmed a significant association between household WASH conditions and wasting. Children living in households with an unimproved WASH were twice as likely to experience wasting (Adjusted OR = 2.01; 95% CI: 1.08–3.75; p = 0.028). Other covariates were not statistically significant in the fully adjusted model (Table 6).
Model Performance and Diagnostics
The classification table (cut off point 0.5) indicated high sensitivity (80.00%) but low specificity (34.87%), with overall correct classification rate of 66.22%. The Hosmer–Lemeshow goodness-of-fit test showed adequate model calibration (Hosmer–Lemeshow χ 2 (8) = 7.92; p = 0.326). The linktest yielded a significant _hat (p =0.016) and a non-significant _hatsq (p = 0.635), indicating that the model meaningful predictive information without evidence of specification error. Collinearity diagnostics (uncentered VIF) showed mean VIF of 1.57, confirming the absence of multicollinearity. Sensitivity analysis using the WASH score as a continuous variable showed that children in good nutritional status have on average a WASH score 0.175 points higher compared to children suffering from wasting (β = 0.175; 95% CI: 0.119–0.320).
DISCUSSION
This study conducted in the Mont Ngafula II health zone aimed to examine the association between household WASH condition, grouped into a child-sensitive WASH composite score, and the nutritional status of children under five in a peri-urban context in the Democratic Republic of Congo
This study provides evidence that inadequate sanitation and hygiene are associated with wasting in a peri-urban Congolese setting. One-third of households practiced inadequate WASH behaviors, a prevalence similar to national estimates,8 but remains lower than reports from rural Ethiopia13 likely reflecting differences in urbanization, access to infrastructure, and methodological criteria used to define WASH practices. These results highlight the urgent need for integrated WASH interventions in peri-urban DRC settings.
Wasting affected 11.6% (95% IC : 8.6-14.6) of children aged 6-59 months, slightly higher than national averages from DHS III. The elevated prevalence may be attributed to the rapid urbanization, limited access to safe water, poor sanitation, and food insecurity typical of peri-urban environments.
In this study, water-related indicators were not significantly associated with wasting. This absence of statistical association is highly consistent with findings from recent large randomized trials ,18which similarly reported no effect of water-focused or even comprehensive WASH interventions on linear growth or weight-for-height outcomes.3,19,20 One possible explanation is that the designation of an “improved” water source does not guarantee microbiologically safe water at the point of use, nor does it ensure sufficient availability to meet daily hygiene needs. Cumming and Cairncross emphasize that reductions in environmental fecal exposure cannot be achieved through water improvements alone, as multiple contamination pathways (soil, surfaces, objects, hands, flies) remain active. In the present context, the lack of continuous water availability (56.2%) and collection-related constraints may further limit households’ ability to maintain adequate hygiene practices despite accessing so-called “improved” sources.
Both bivariate and multivariate analyses revealed a strong association between the use of unimproved latrines, inadequate disposal of child feces, lack of water or soap at handwashing stations, and the prevalence of wasting. Children living in households with an unimproved WASH were twice as likely to experience acute malnutrition compared with those in improved households (adjusted OR = 2.01; 95% CI: 1.08-3.75). This association aligns with previous research showing that poor sanitation conditions expose children to a higher risk of enteric infections, recurrent diarrhea, and environmental enteric dysfunction (EED), all recognized contributors to the onset or worsening of wasting.2,11
These findings also corroborate evidence from various low-resource settings where inadequate sanitation practices have been consistently identified as major determinants of both wasting and stunting.10,13,21 They further reflect national trends in the DRC, where peri-urban and rural areas are characterized by particularly low levels of hygienic excreta management and high burdens of diarrheal disease.8
Although poor households and mothers without formal employment were overrepresented among those with unimproved WASH, these variables lost significance after adjustment. This suggests that WASH may function as a key mediator between socioeconomic status and child nutrition, as observed in previous analyses.22,23 The persistence of malnutrition across different socioeconomic groups highlights that environmental contamination affects all children, regardless of household economic status, in contexts where community-level infrastructure and services are insufficient.
However, the interpretation of these findings requires caution. First, the cross-sectional design prevents establishing temporal relationships, raising the possibility of reverse causality, such as households improving hygiene behaviors when a child becomes malnourished. Second, although the composite score integrates multiple WASH dimensions, its predictive validity was modest (AUC = 0.62), indicating limited discriminative power. As such, the cut-off ≥6, while operationally practical, remains somewhat arbitrary. Third, equal weighting of indicators may oversimplify complex interactions between sanitation, hygiene and water pathways. Additionally, unmeasured confounders, such as child morbidity, feeding practices, and household food security, may partially explain the association.
Nonetheless, the sensitivity analysis using the continuous score strengthens confidence in the observed relationship. The study contributes valuable evidence from a context with limited research on WASH-nutrition interactions in peri-urban Africa.
STRENGTHS AND LIMITATIONS
This study offers several strengths, including the use for the first time in the DRC of a child-friendly WASH composite score to capture the combined effects of water, sanitation, and hygiene practices on child nutrition. Its contextual focus on a vulnerable peri-urban setting in the DRC enhances the relevance of findings for targeted interventions. The methodological rigor clear inclusion criteria, representative sampling, standardized anthropometric measurements with validated tools (WHO), use of the UNICEF JMP questionnaire for WASH practice and sensitivity analysis strengthening robustness of findings.
However, the cross-sectional design limits causal interpretation. Self-reported WASH data may be subject to bias, and key nutritional determinants such as dietary intake and morbidity history were not assessed. Moderate predictive performance of the composite score (AUC = 0.62), limiting its diagnostic value. Additionally, the binary classification of the WASH score may oversimplify household realities, and findings may not be generalizable beyond the study area. These limitations justify cautious interpretation and highlight the need for longitudinal studies and improved WASH measurement tools in similar settings.
CONCLUSIONS
This study conducted in the Mont Ngafula II health zone in the DRC highlights the determining role of WASH conditions in nutritional status in children aged 6 to 59 months. In Mont Ngafula II, inadequate sanitation and hygiene conditions were strongly associated with wasting among children under five. Efforts to reduce wasting must integrate WASH improvements with nutrition-specific and community-level interventions. Strengthening local sanitation infrastructure, promoting safe disposal of child feces, and ensuring availability of water and soap for handwashing could reduce environmental contamination and improve child nutritional outcomes. Although the composite WASH score had limited predictive capacity, it remains a practical tool for identifying vulnerable households in resource-constrained settings
Acknowledgements
We are grateful to the parents and children in Mont Ngafula II HZ for sharing their experiences with us. We also thank Michel Muvudi for his contribution to our training at the Kinshasa School of Public Health
Ethics statement
Ethical approval was obtained from the Ethics Committee of the Kinshasa School of Public Health (authorization No. ESP/CE/37B/2024). Written consent was obtained from all participants.
Data availability
The datasets generated and/or analyzed during this study are available from the corresponding author upon reasonable request.
Funding
This research received no external funding. The article publication charge (APC) was not funded by any organization.
Authorship contributions
Disuemi Emérite is the principal author who conceptualized and drafted the main research protocol, which was validated by the facilitators Botomba Steve and Mutombo Paulin. Waby Tannie, Makaba Richard, Mbiki Lisbeth, Lipemba Jean, Poba Junior, Bukasa Gloria, Izanga Arsène, Batuli Din-ar, Baboko Christian, Jobalo Yannick and Mopasola Eric conducted the data collection and analysis. Disuemi Emérite, Waby Tannie, Botomba Steve and Mutombo Paulin provided statistical oversight, critical revisions, and final approval. All authors reviewed and approved the final version in accordance with ICMJE authorship criteria.
Disclosure of interest
The authors completed the ICMJE Disclosure of Interest Form (available upon request from the corresponding author) and disclose no relevant interests.
CORRESPONDING AUTHOR
Disuemi Emérite, emeritedisuemi@gmail.com
Telephone : +243822846713