Animals in child sleeping spaces and unimproved sanitation associated with diarrhea among young children in the Democratic Republic of the Congo: WASHmobile PICHA7 prospective cohort study

Abstract

Background

Exposure to animal and human feces in the household environment is associated with diarrheal diseases in young children. The objective of this study was to identify exposure pathways to fecal pathogens for young children that are significant contributors to diarrheal disease to allow for targeted water, sanitation, and hygiene (WASH) interventions on these pathways.

Methods

The WASHmobile Preventative Intervention for Cholera for 7 Days prospective cohort study was conducted in urban Bukavu, South Kivu province, Democratic Republic of the Congo. The cohort study of 794 children under 5 years of age included monthly diarrhea surveillance and unannounced spot checks to assess WASH behaviors and conditions over a 12-month period. Caregiver report of child mouthing of fomites was also obtained during monthly visits.

Results

The presence of animals in the child's sleeping space (odds ratio (OR): 1.87; 95% confidence interval (CI): 1.14, 3.08), unimproved sanitation (OR: 2.27; 95% CI 1.19, 4.33), and consumption of food outside the household (OR: 1.88; 95% CI 1.16, 3.06) were significantly associated with diarrhea during the subsequent month.Chickens [OR: 3.38; 95% CI: 2.05, 5.59) and cats [OR 3.9; 95% CI: 1.46, 10.46] in the child’s sleeping space was also associated with significantly higher odds of diarrhea in the subsequent month.

Conclusions

These results demonstrate the need for WASH interventions targeted at reducing child contact with animal feces in the indoor household environment and improved sanitation, to reduce exposure to fecal pathogens for susceptible pediatric populations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s41182-026-00901-1.

Background

Despite global improvements in water, sanitation and hygiene (WASH), diarrhea remains a significant contributor to morbidity and mortality, and is estimated to contribute to 54 million disability-adjusted life-years and over 1 million deaths each year [1, 2]. In the Democratic Republic of the Congo (DRC) alone, there are estimated to be 45 million diarrhea episodes annually among children under 5 years of age [3]. Chronic diarrhea episodes in young children contribute to undernutrition, increasing the risk of stunting, which is associated with mortality and impaired cognitive development [4–7]. Identifying exposures associated with diarrhea risk and interventions to intervene upon them is especially important in fragile contexts, such as eastern DRC, where ongoing conflict constrains delivery of WASH services.

Child contact with domestic animals and their feces has been found to be an important contributor to diarrheal disease and impaired growth in young children [8, 9]. In resource-constrained urban settings, domestic animals are often kept in close proximity to household living spaces. This close proximity can lead to exposure to animal feces, especially for young children, who play on the ground and mouth fomites, a known route of enteric pathogen exposure [10–12]. During structured observations in rural DRC, children observed putting soil in their mouth at baseline had significantly higher odds of diarrhea at follow-up [13]. In a recent meta-analysis reviewing animal exposure and diarrheal disease in low- and middle-income countries (LMICs), authors found a 1.95 higher odds of diarrhea with animal exposure [9]. While few studies have looked at specific animal types, Ecrumen et al. found higher diarrhea prevalence in households with more poultry compared to households with less poultry [14]. However, the association between animal ownership and diarrheal disease is complex, with the previous studies suggesting that diarrhea may decrease with animal ownership through higher socioeconomic status or improved nutrition [13, 15, 16]. Further evidence is needed in specific contexts to tailor WASH interventions to target exposure pathways to fecal pathogens.

Additional important risk factors for diarrheal disease include unimproved sanitation facilities and unimproved drinking water sources [17–19]. This is the case, for example, in rural Gambia and Kenya, and in urban Mali, where children whose source of drinking water or sanitation was lower quality had a higher risk of developing diarrhea [20]. Lack of handwashing with soap is also associated with increased diarrhea among children under 5 years of age [21]. Diarrheal disease risk is also impacted by severe weather events, with droughts, floods, and other severe weather damaging or overwhelming WASH infrastructure, thereby increasing water-borne pathogen exposure [22, 23]. Therefore, understanding context-specific diarrhea risk factors is important to design targeted intervention approaches.

In an effort to reduce diarrheal diseases at our study site in the DRC, our research group developed WASHmobile, a WASH mobile health program targeting diarrhea patient households and those in high-risk areas for diarrhea outbreaks. Together with the DRC Ministry of Health, we developed a country-specific version of WASHmobile, the Preventative Intervention for Cholera for 7 Days (PICHA7) program, to create a targeted WASH program promoting drinking water treatment, safe drinking water storage, and handwashing with soap at food and stool-related events [24]. The WASHmobile PICHA7 program was designed through 24 months of community-centered formative research guided by health behavior theory [25]. The WASHmobile PICHA7 program is healthcare facility initiated with a health promoter delivering a pictorial WASH module bedside to a diarrhea patient and their accompanying household members. A hygiene kit is also provided with a handwashing station and quarterly home visits are conducted by a health promoter. Finally, a mobile health program is delivered, where a doctor at a local hospital sends weekly automated voice and text messages to diarrhea patient households over a 12-month period. Our recent RCT of the WASHmobile PICHA7 program found that the program was effective in reducing diarrhea and stunting for children under 5 years and increasing handwashing with soap, water treatment practices, and stored drinking water quality. [24]

This prospective cohort study, nested within the broader WASHmobile PICHA7 program, was conducted to investigate household-, environmental-, and individual-level risk factors for diarrhea among young children in Bukavu in South Kivu province of the DRC. Risk factor analyses are important to effectively tailor WASH programs and target exposure pathways.

Methods

Study design and population

This study was conducted in urban Bukavu in South Kivu province, DRC in the health zones of Kadutu, Ibanda, Bagira, and Nyantende from December 2021 to December 2023 (NCT05166850). The target population for this study was diarrhea patients and their household members. Daily diarrhea patient surveillance and recruitment was conducted at 115 public and private health facilities from December 2021 to December 2022.

Diarrhea patients were enrolled based on the following criteria: 1) diarrhea patient admitted to a health facility with three or more loose stools over a 24-h period; 2) provided a blood and a stool sample within 24 h of enrollment; 3) had a child under 5 years living in the household; 4) planned to reside in Bukavu for at least 12 months; 5) resided at home for the three nights before hospitalization; 6) no functioning tap to collect water inside the home (mostly informal settlements); and 7) at least one working mobile phone in the home. If the diarrhea patient was eligible and willing to participate in the WASHmobile PICHA7 program, both them and their household members were enrolled as a household (cluster level). Household members of the diarrhea patient were eligible for the trial if: 1) they shared the same cooking pot and resided in the same home with the diarrhea patient for the last 3 days and 2) planned to reside with the diarrhea patient for the next 12 months. Participants were followed for 12 months. The detailed protocol for this trial was published previously [24]. Participants were enrolled in one of the two study arms: the standard message arm, where participants received the standard discharge recommendations for diarrhea patients in the DRC on oral rehydration solution (ORS) use for rehydration, or the WASHmobile PICHA7 arm, where participants received the standard arm plus the WASHmobile PICHA7 mHealth program.

Data collection

Starting at baseline, research officers administered questionnaires monthly to study participants. Questionnaires included information on caregiver-reported diarrhea (defined as 3 or more loose stools in a 24-h period in the past 2 weeks), food consumption and water source use outside the home in the previous 2 weeks, and child mouthing of soil (geophagy) and other objects in the past week. In addition, at baseline caregivers were administered a questionnaire on sociodemographic factors, including household roof type, refrigerator ownership, household literacy, household income, and number of sleeping rooms.

Unannounced spot checks were performed in a random subset of 100 households per study arm (200 households total) at 7 days and 1, 3, 6, 9, and 12 months after enrollment. Information was collected on stored drinking water covering status, animal presence on the compound and in the sleeping space, animal ownership, and feces presence on the household compound. At the Day 7 spot check, research officers assessed the type of water source and sanitation options present. The sanitation options and water sources were defined using the WHO/UNICEF Joint Monitoring Program (JMP) for Water Supply, Sanitation and Hygiene definitions [28, 29]. Improved water sources included piped water, protected springs, rainwater, packaged or delivered water, boreholes, and tube wells. Unimproved water sources included unprotected dug wells, springs, lakes, ponds, rivers, dams, streams, canals, or irrigation canals. Basic water service was defined as drinking water from an improved source, where the collection time did not exceed 30 min (including queuing time). Limited water service was defined as drinking water from an improved source, where the collection time exceeded 30 min (including queuing time). Improved sanitation included septic tanks, pit latrines with slabs, composting toilets, and flush/pour toilets connected to a piped sewer system. Basic sanitation service was the use of improved sanitation facilities that were not shared with other households. Limited sanitation service was defined as the use of improved sanitation facilities that were shared with two or more households. Unimproved sanitation included a hanging latrine, bucket latrine, and a pit latrine without a slab or platform. Open defecation was defined as no sanitation facility used. The sample size for variables differed slightly because of the “do not know” responses and the subset of households sampled for spot check variables (also for water source type, sanitation option type).

Statistical analysis

To investigate the risk factors for diarrhea among young children, we compared individual- and household-level characteristics among children under 5 years of age enrolled in the WASHmobile PICHA7 RCT. The outcome for this study was caregiver-reported diarrhea for children under 5 years of age (defined as 3 or more loose stools in a 24-h period in the past 2 weeks). Logistic regression was performed to assess the associations between individual and household risk factors (predictors) and diarrhea in the past 2 weeks during the subsequent month (outcome). Generalized estimating equations were used for all regressions to account for clustering at the household and individual levels with exchangeable correlation structure. Analyses were adjusted for baseline age, gender, study arm, roof type, household literacy, ownership of a refrigerator, and number of bedrooms and household income. We did not include soap or handwashing station presence or water quality indicators in this risk factor analysis as these variables were found to be associated with WASHmobile delivery in our recent RCT. [30]

Ethical approval

This study was approved and validated by the ethics committee of the Bloomberg School of Public Health (Johns Hopkins University) (9848) and the Catholic University of Bukavu (UCB) (7107). All study participants (parent/guardian for children) provided written informed consent.

Results

In total, 794 children under 5 years of age were enrolled in this study (Table 1). The mean age for children at baseline was 24.9 months (standard deviation: 17.4, range: 0–60), and 51% of children were female (401/794). Five percent (41/758) of children resided in a household with a concrete roof, and 95% (717/758) of the houses were covered with other materials, mostly sheet metal. Three percent (22/758) of children resided in a household with a refrigerator. Ninety-eight percent (742/757) of children had at least one household member who could read and write, 6% (45/754) of children lived in households with no income, and 84% (637/758) had homes with two or more sleeping rooms. Forty-three percent (341/787) of children had caregiver-reported diarrhea in the past 2 weeks at baseline.

Table 1.

WASHmobile PICHA7 program baseline participant characteristics

	%	n	N
Households			352
Children < 5 years of age			794
Individuals living in household
Mean ± SD (min–max)		7.6 ± 2.6 (1–16)
Enrolled children per household
Mean ± SD (min–max)		2.3 ± 0.9 (1–6)
Baseline age (months)
Mean ± SD (min–max)		24.9 ± 17.4 (0–60)
0–12	33%	262	794
12–24	20%	155	794
24–36	17%	137	794
36–48	17%	132	794
48–60	14%	108	794
Female	51%	401	794
Household roof type
Concrete	5%	41	758
Other	95%	717	758
Refrigerator ownership	3%	22	758
At least one household member who can read and write	98%	742	757
Household income
Has income	94%	709	754
No income	6%	45	754
Sleeping rooms
Two or more	84%	637	758
Less than two	16%	121	758
Diarrhea in previous 2 weeks*	43%	341	787

SD = standard deviation

Diarrhea defined as caregiver-reported diarrhea (3 or more loose stools in a 24-h period in the past two weeks)

At baseline, 79% (574/725) of children had caregiver reports of consuming food outside the home during the past 2 weeks (Table 2). Fifty-five percent (278/509) of children resided in households with unimproved sanitation, 18% (93/509) with a limited sanitation option, 18% (90/509) with a basic sanitation option, 9% (48/509) with a safely managed sanitation option, and none resided in households reporting open defecation. Twenty-nine percent (149/515) of children resided in households that reported animal ownership and 37% (192/515) of children resided in households with animals observed on the household compound during spot checks at baseline. Animals were present in the children's sleeping area for 29% (57/194) of children based on observation during spot checks. Presence by animal type is reported in Supplemental Table 1. The most common animal presence were chickens at 18% (94/515), cats at 11% (56/515), and mice at 10% (50/515).

Table 2.

Participant-level baseline risk factors for diarrhea (total participants = 794)

Risk factor	%	n	N
Consumed food outside home—past 2 weeks	79%	574	725
Consumed water source outside home—past 2 weeks	52%	409	786
Caregiver-reported geophagy—past week	87%	287	331
Caregiver-reported mouthing of dirty object—past week	87%	297	342
Primary Water Source Type
Basic Water Service	72%	335	466
Limited Water Service	19%	90	466
Unimproved Water Source	9%	41	466
Primary Sanitation Option Type
Safely Managed Service	9%	48	509
Basic Sanitation Service	18%	90	509
Limited Sanitation Service	18%	93	509
Unimproved Sanitation	55%	278	509
Open Defecation	0%	0	509
Household animal ownership- caregiver reported	29%	149	515
Animal presence on the household compound— spot check observation	37%	192	515
Animal presence in the child's sleeping space— spot check observation	29%	57	194
Feces present on household compound	27%	138	515

^*The sanitation options and water sources were defined using the World Health Organization (WHO)/ UNICEF Joint Monitoring Program (JMP) definitions. An improved water sources included piped water, protected springs, rainwater, packaged or delivered water, boreholes, and tubewells. Unimproved water sources included unprotected dugwells, springs, lakes, ponds, rivers, dams, streams, canals, or irrigation canal. Basic water service was defined as drinking water from an improved source, where the collection time did not exceed 30 min (including queuing time). Limited water service was defined as drinking water from an improved source, where the collection time exceeded 30 min (including queuing time). Improved sanitation included septic tanks, pit latrines with slabs, composting toilets, and flush/pour flush toilets connected to a piped sewer system. Unimproved sanitation included a hanging latrine, bucket latrine, and a pit latrine without a slab or platform. Basic sanitation service was the use of improved sanitation facilities that were not shared with other households. Limited sanitation service was the use of improved sanitation facilities that were shared with two or more households. Open defecation was where no sanitation facility was used

n = number of participants with characteristic

N = number of participants with information on characteristic

N not equal to 794 for all factors due to "do not know" responses for questions, and information collected on a subset of households for spot check factors (including sanitation and water source type)

Feces presence includes any feces (animal or human)

Children residing in households where animals were observed in their sleeping space during spot checks had a 1.87 times higher odds of diarrhea in the subsequent month compared to children in households where no animals were observed in the sleeping area [95% confidence interval (CI) 1.14, 3.08] (Table 3). Analyses by animal type are reported in Supplemental Table 2.  Chickens in a child's sleeping space was associated with a 3.38 higher odds of diarrhea in the subsequent month (95% CI: 2.05, 5.59), and cats in a child's sleeping space was associated with a 3.9 times higher odds of diarrhea in the subsequent month (95% CI: 1.46, 10.46).  Unimproved sanitation was associated with a 2.27 times higher odds of diarrhea in the subsequent month compared with basic sanitation (95% CI 1.19, 4.33), and consuming food outside the home was associated with 1.88 times higher odds of diarrhea in the subsequent month (95% CI 1.16, 3.06).

Table 3.

Associations between individual- and household-level risk factors and diarrhea in the past 2 weeks at subsequent visit for children under 5 years (total participants = 631)

	Risk Factor			Diarrhea in the past 2 weeks
Risk factor	%	n	N	OR (95% CI)
Consumed food outside home past 2 weeks	93%	2695	2893	1.88 (1.16, 3.06)
Consumed water source outside home past 2 weeks	59%	1700	2883	1.26 (0.99, 1.60)
Caregiver-reported geophagy—past week	96%	1201	1252	2.12 (0.94, 4.77)
Caregiver-reported mouthing of dirty object—past week	93%	1322	1428	0.81 (0.50, 1.34)
Primary Water Source Type
Basic Water Service	74%	643	870	(reference)
Limited Water Service	20%	172	870	0.85 (0.50, 1.43)
Unimproved Water Source	6%	55	870	1.10 (0.51, 2.38)
Primary Sanitation Option Type
Safely Managed Service	7%	74	998	1.30 (0.53, 3.18)
Basic Sanitation Service	17%	171	998	(reference)
Limited Sanitation Service	18%	175	998	1.79 (0.87, 3.67)
Unimproved Sanitation	58%	578	998	2.27 (1.19, 4.33)
Open Defecation	0%	0	998	–
Household animal ownership - caregiver reported	28%	283	993	0.84 (0.56, 1.27)
Animal presence on the household compound—spot check observation	45%	449	993	1.41 (0.99, 2.00)
Animal presence in the child's sleep space—spot check observation	33%	149	457	1.87 (1.14, 3.08)
Feces present on household compound	33%	335	1023	1.04 (0.74, 1.45)

CI = confidence interval. OR = odds ratio. % indicates percent of visits with diarrhea. Total differs as some assessments were conducted in a random selection of households, and some households/participants did not respond to all questions. Adjusted for baseline age (months), gender, study arm, household roof type, household literacy, refrigerator ownership, number of sleeping rooms, and household income

^*The sanitation options and water sources were defined using the World Health Organization (WHO)/UNICEF Joint Monitoring Program (JMP) definitions. An improved water sources included piped water, protected springs, rainwater, packaged or delivered water, boreholes, and tubewells. Unimproved water sources included unprotected dugwells, springs, lakes, ponds, rivers, dams, streams, canals, or irrigation canal. Basic water service was defined as drinking water from an improved source, where the collection time did not exceed 30 min (including queuing time). Limited water service was defined as drinking water from an improved source, where the collection time exceeded 30 min (including queuing time). Improved sanitation included septic tanks, pit latrines with slabs, composting toilets, and flush/pour flush toilets connected to a piped sewer system. Unimproved sanitation included a hanging latrine, bucket latrine, and a pit latrine without a slab or platform. Basic sanitation service was the use of improved sanitation facilities that were not shared with other households. Limited sanitation service was the use of improved sanitation facilities that were shared with two or more households. Open defecation was where no sanitation facility was used

n = number of participant visits with characteristic

N = number of participant visits with information on characteristic

Feces presence includes any feces (animal or human)

Discussion

In the WASHmobile PICHA7 prospective cohort study, we investigated risk factors for diarrhea in children under 5 years of age. Animals in the child’s sleeping space particularly chickens and cats, unimproved sanitation, and consuming food outside of the home were associated with a significantly higher odds of diarrhea in the subsequent month. These findings emphasize the urgent need for interventions targeting the reduction of child contact with domestic animals in the indoor household environment and improved sanitation to reduce diarrhea among pediatric populations in urban DRC.

Children living in households with animals present in the child’s sleeping area had higher odds of diarrhea compared to those with no animals present. This finding is consistent with previous studies showing that close contact between animals and children is associated with diarrhea, with a previous study in Ethiopia finding a relationship between animal presence in the home and increased risk of diarrhea in children [31]. Notably, while animal presence in the sleeping space was significant, neither animal presence on the household compound nor animal ownership were associated with increased odds of diarrhea. This critical finding emphasizes the importance of the proximity of animals to the child’s indoor environment as an important risk factor for diarrhea rather than solely animal presence on the household compound. Previous studies have found strong associations between animal ownership and improved nutrition, child growth, and other household factors [32, 33]. A nuanced understanding of animal exposure risk pathways is needed to reduce diarrheal disease risk while supporting nutrition improvements.

In this study, the presence of chickens and cats in child sleeping spaces were found to be significantly associated with a higher risk of diarrhea. Our previous work in Bangladesh and rural DRC found associations between animal contact and impaired linear growth, as well as feces in the child’s sleeping space and impaired linear growth in the DRC study [13, 34]. When looking at specific animal types in the DRC Reducing Enteropathy, Undernutrition, and Contamination in the Environment (REDUCE) study, caregiver-reported contact with guinea pigs and with rabbits were most significant. To address child exposure to animals in the household, we designed the REDUCE Baby WASH models in rural DRC, where we promoted the construction of animal hutches for rabbits and guinea pigs, and the construction of a compost pit for animal feces and household food waste [35]. Designing such interventions tailored to specific risk factors is important to addressing the risks of animal exposures while supporting improved nutritional benefits of livestock ownership.

In this study, we also found that children living in households with unimproved sanitation had an increased odds of diarrhea compared to children living in households with basic sanitation. However, limited sanitation did not have a significant difference in diarrhea compared to basic sanitation. We found no households with open defecation. In the DRC, the percentage of households with unimproved sanitation has increased between 2000 and 2024 (42–54%), accompanied by decreases in limited and safely managed sanitation and stable rates of open defecation, according to data from the WHO/UNICEF JMP for Water Supply, Sanitation and Hygiene [36]. This increase is likely related to the decades of conflict, particularly in eastern DRC. Given this context, interventions are of critical importance to increase improved sanitation coverage in the DRC.

Finally, we found that consumption of food outside the home in the previous 2 weeks was associated with diarrhea in the subsequent month. Across LMIC settings, eating food prepared outside the household is consistently associated with increased diarrhea risk, including cholera [40–42]. Microbiological surveys from Bukavu report heavy contamination of street-vended foods, finding widespread contamination of E. coli and Salmonella in food samples [43]. Food contamination was related to food handling practices, including frequency of  cooking food and whether food was heated before serving customers in both formal restaurants and street-vended food. This highlights the need for interventions targeting food vendors outside of the home environment. In addition, household WASH interventions can recommend avoidance of consuming food outside of the home during outbreak periods, and consumption of restaurant or street-vended food that is served hot.

This study had several strengths, including a prospective cohort design. The prospective cohort design is important, because many studies of diarrheal disease risk factors use a cross-sectional model, not allowing for a clear cause and effect relationship. We also included several risk factors measured through direct observation during unannounced spot checks instead of relying solely on participant reported data, which may be subject to reporting bias. However, this study also has limitations. For some variables, including the occurrence of diarrhea, we relied on caregiver report. We also only enrolled households in an urban setting; however, most previous studies were conducted in rural settings so this study provides needed evidence in an urban context. In addition, for our spot check analysis of the presence of animals by type the number of participants in some categories was low therefore caution should be applied when interpreting these findings.  Finally, participants were enrolled during an ongoing WASH RCT, while although we adjusted for study arm in our analysis this may still limit the generalizability of our findings to other settings.

Conclusion

In this study, we prospectively examined risk factors associated with diarrhea among children under the age of five in urban eastern DRC. The presence of animals in the child’s sleeping area, unimproved sanitation, and consumption of food outside the home were associated with diarrhea among young children. These results demonstrate the need for WASH interventions targeted at reducing child contact with animal feces in the household environment, improved sanitation, and food hygiene to reduce exposure to fecal pathogens for susceptible pediatric populations.

Author contributions

Conceptualization, Christine Marie George; formal analysis, Kelly Endres, Presence Sanvura and Jean-Claude Rusanga; investigation, Christine Marie George, Jean-Claude Rusanga and Kelly Endres; methodology, Christine Marie George, Jean-Claude Rusanga Kelly Endres and Presence Sanvura; project administration, Christine Marie George, Cikomola Cirhuza, Maheshe Ghislain, Lucien Bisimwa and Alain Mwishingo; supervision, Christine Marie George, Pacifique Mwene-Batu and Maheshe Ghislain. writing and editing; Jean-Claude Rusanga, Maheshe Ghislain, Pacifique Mwene-Batu, Justin Bengehya, Presence Sanvura, Kelly Endres, Jean-Claude Kulondwa, Cikomola Justin, Lucien Bisimwa, Alain Mwishingo, Jamie Perin, Camille Williams, Elizabeth Thomas, and Christine Marie George.

Funding

This work was made possible by UK aid funding from the Foreign and Commonwealth Development Office and by the Reckitt Global Hygiene Institute, funding provided to Christine Marie George of Johns Hopkins University, a School of Public Health. The funding organizations were not involved in study design, data collection, data analysis or data interpretation.

Data availability

Anonymized data can be made available upon request. Requests should be addressed to Christine Marie George ([cmgeorge@jhu.edu] (mailto:cmgeorge@jhu.edu)).

Declarations

Informed consent

Informed consent was obtained from all subjects involved in this study

Competing interests

The authors declare no competing interest.