Abstract
The objective of this study was to assess the prevalence of diarrhea among under-five children in low-middle-income countries and identify the sociodemographic factors associated with it. Data of 36 countries in sub-Saharan Africa from demographic and Health Surveys (2006-2018) comprising 251 341 mother-child (singleton) dyads were analyzed to estimate the prevalence and various modifiable and non-modifiable risk factors of diarrhea. Occurrence of diarrhea during the last 2 weeks was the outcome variable which was measured by mothers’ observation of the condition. The overall prevalence of having diarrhea during last 2 weeks was 18.44% (19.12% among boys and 17.75% among girls). Boys had higher percentage of having diarrhea than girls in all countries except in Libya. The risk ratios of having diarrhea decreased progressively with higher wealth quintiles; the risks of were respectively 7% [RR = 0.93, 95% CI = 0.91; 0.97], 11% [RR = 0.89, 95% CI = 0.86; 0.92] and 18% [RR = 0.82, 95% CI = 0.78; 0.85] lower for households in the middle, richer and richest households. Rural residency was associated with lower risks [RR = 0.95, 95% CI = 0.93; 0.98] and not having access to improved water [RR = 1.05, 95% CI = 1.03; 1.08] and toilet facilities [RR = 0.04, 95% CI = 1.01; 1.07] were associated with higher risks of diarrhea. Regarding children’s characteristics, higher age groups, birth order were associated with higher risks and female sex with lower risks. Children with mothers in the higher age groups and with above secondary level education had lower risks, and primary education had higher risks of diarrhea. Meta-analysis of 36 countries revealed a significantly negative association between wealth quintile and diarrhea (Odds ratio = 0.72, 95% CI = 0.69; 0.74). Findings indicate the presence of a significant wealth gradient in the burden of diarrheal diseases among under-five children in sub-Saharan Africa, and underscores the need for paying special attention to the marginalized communities when designing intervention programs.
Diarrhea is a major public health burden in sub-Saharan Africa.
The prevalence of diarrhea varied widely across the countries, with Libya having the highest and Madagascar having the lowest burden.
There is a significant urban-rural and wealth gradient in the distribution of diarrhea.
The wealth disparity is more pronounced in rural areas compared with urban.
The authorities should pay more attention to the public health burden in rural areas than that in urban areas.
The target public health policy should be made so as to promote the health improvement of resident in sub-Saharan Africa, especially the under-five children.
Introduction
Diarrhea is a significant health issue among children in the low-middle income countries, especially those in sub-Saharan Africa.1,2 Diarrhea is influenced by various environmental and physiological factors, and is commonly caused by bacterial (Campylobacter, Escherichia coli, Salmonella, Shigella) or viral infections (norovirus and rotavirus) transmitted through the orofecal route. 3 Although it can be life-threatening condition, in many cases of diarrhea can be prevented by the adoption of simple measures such as improving hygiene and reducing exposure to contaminated food and water.4,5 Moreover, affordable interventions like oral rehydration therapy (ORT) and rotavirus immunization are very effective in managing diarrhea without requiring a hospital visit.6,7 In fact, a great majority (88%) of diarrhea-associated deaths are attributable to lack of access to WASH facilities, 8 which means that bulk of the disease burden could be avoided through providing better WASH facilities alone. However, lack of access to water, sanitation, and hygiene (WASH) facilities contributes to a significant proportion of diarrhea-related deaths in sub-Saharan Africa.7 -9
In sub-Saharan Africa, millions of children suffer from diarrhea which sometimes leads to increased bouts infection, malnutrition, and early death.7 -9 According to previous studies, unhealthy housing conditions, overcrowded living spaces, indoor air pollution, and lack of proper hygiene practices further exacerbate the vulnerability to diarrhea among children living in unhealthy settings.10 -12 As a disease of poverty, those in unhealthy living environments often receive insufficient attention in disease control priorities within African countries, despite their significant impact on child health. Currently, there is not enough evidence on the extent of diarrhea among under-five children in Africa. To address this gap, this study aims to measure the prevalence of acute diarrhea among children under the age of 5 and investigate the sociodemographic disparities in its occurrence. The analysis utilizes data from Demographic and Health Surveys conducted in 35 sub-Saharan African countries, which provide information on recent diarrhea episodes and relevant sociodemographic indicators. By highlighting variations in prevalence across countries, this report offers valuable insights into the geographical distribution of diarrhea and its association with household poverty. These findings have important implications for research, interventions, and policy actions aimed at preventing childhood diarrhea and promoting child health-related Sustainable Development Goals.
Materials and Methods
Data Source
Data used for this analysis were open-access and obtained from Demographic and Health Surveys available in 36 countries in sub-Saharan Africa since 2006. List of countries with survey years are provided in the appendix (Table a). These surveys were implemented through a collaborative effort of the designated research institutions in each country along with financial and technical supports from international development partners such as United States Agency for International Development (USAID) and Inner City Fund (ICF) International, United States. 12 DHS is dedicated to providing quality data on crucial demographic and health indicators to facilitate evidence-based making in the respective countries.
For sampling design, DHS employs multistage cluster techniques that involve the selection of sampling strata and primary sampling units (PSUs) at the first step, which then provides the basis for choosing a certain number of households from each PSU. At the first step of the sampling method, a certain number of enumeration areas are selected in both urban and areas which form the basis for choosing the PSUs. At the second stage, households are selected systematically from the PSUs for interviews. From the selected households, eligible men (15-54 years), women (15-49 years), and children (0-59 months) are assigned for interview. These details were described in previous publications.13 -15
Description of Variables
The outcome variable was occurrence of diarrhea during the 2 weeks (Yes/No), which was measured by asking the respondent (mother) whether or not the child had diarrhea during the last 2 weeks. 15 Based on the literature review, the analysis included 10 household, child, and maternal variables—Residency: Urban/Rural; household wealth quintile: Poorest/Poorer/Middle/Richer/Richest 16 ; Water source: unimproved/Improved 15 ;Toilet facility: unimproved/Improved 15 ; Child’s age: 1 to 12/13 to 24/25 to 36/37 to 48/49 to 59 months; Sex of child: Male/ Female; Birth order; Mother’s age: 15 to 19/20 to 24/25 to 29/30 to 34/35 to 39/40 to 44/45 to 49 years; Educational level: No education/Primary/Secondary/Higher.
Statistical Analysis
All data analyses were performed using Stata version 16. As the DHS surveys employ cluster sampling methods for sample selection, we used the svy command to account for the clustering effect for all analyses. Prevalence of children with diarrhea during last 2 weeks was shown as percentages in the form of bar charts in 3 separate columns: pooled, boys and girls. The risk ratios (RRs with 95% Confidence Intervals) of having diarrhea were across the sociodemographic factors were measured using multivariate regression models. Variable importance charts were created to plot the percentage contribution of the explanatory factors to the variance in diarrhea occurrence. Finally, we performed a country-level meta-analysis of the odds ratios of having diarrhea by wealth quintile while controlling for the other explanatory factors. For all associations P-value of <.05 was considered statistically significant.
Results
The prevalence of diarrhea in the overall sample was 18.4% (95% CI = 18.3; 18.6) (Table 1). Chi-squared bivariate tests showed significant (P < .05) group differences with all the household level variables except for type of toilet facility.
Background Characteristics of the Sample Population.
Note. Cell values represent column percentages with 95% confidence interval (Prevalence of diarrhea).
The overall prevalence of having diarrhea during the last 2 weeks was 19.12% among boys and 17.75% among girls (Figure 1). Figure 2 illustrates that boys had a higher percentage of having diarrhea than girls in all countries except for Libya. Of the 36 countries, Libya had the highest percentage of diarrhea where more than one-quarter of the boys and girls had diarrhea during 2 weeks preceding the survey with the prevalence being marginally higher among girls than boys (Boys 27.2% vs Girls 27.5%). Burundi was the only other country where more than a quarter of the children had diarrhea during last 2 weeks. Boys had higher percentage of having diarrhea than girls in all countries except in Libya. Countries where more than a fifth of the children had diarrhea included Libya, Burundi, Uganda, Tchad, Malawi, Senegal, Cameroon, Gambia, Ivory Coast, Congo, Mali, Namibia, Congo DR, and Comoros. The lowest prevalence was observed in Madagascar (10.8%) and South Africa (11.5%).
Overall prevalence of diarrhea among under-five children in SSA.
Prevalence of diarrhea among under-five children by country.
Results of the variable importance plot shwed in figure 3, as well as the results of multivariate regression calculating the risk ratios of having diarrhea were shown in Table 2. In the pooled sample, the risk ratios of having diarrhea decreased gradually with higher wealth quintiles. Compared with the lowest wealth quintile households, the risks of having diarrhea were respectively 7% [RR = 0.93, 95% CI = 0.91; 0.97], 11% [RR = 0.89, 95% CI = 0.86; 0.92] and 18% [RR = 0.82, 95% CI = 0.78; 0.85] lower for households in the middle, richer and richest households. After stratifying the analysis by type of residency, the association remained significant for rural samples only. Rural residency was associated with lower risk of diarrhea [RR = 0.95, 95% CI = 0.93; 0.98]. Children in households not having access to improved water [RR = 1.05, 95% CI = 1.03; 1.08] and toilet facilities [RR = 0.04, 95% CI = 1.01; 1.07] had higher risks of diarrhea as well. Regarding children’s characteristics, higher age groups, birth order [RR = 1.04, 95% CI = 1.03; 1.04] were associated with higher risks and female sex [RR = 0.93, 95% CI = 0.91; 0.95] with lower risks. Children with mothers in the higher age groups and with above secondary level education [RR = 0.74, 95% CI = 0.68; 0.80] had lower risks, and primary education [RR = 1.06, 95% CI = 1.03; 1.09] had higher risks of diarrhea.
Variable importance plot.
Risk Ratios of Diarrhea Among Under-Five Children in Sub-Saharan Africa.
Note. All models adjusted for country. Risk ratios with 95% confidence. Level of significance: *P < .05. **P < .01. ***P < .001.
Discussions
Our analysis showed that less than a fifth of the children had diarrhea during last 2 weeks, with the overall prevalence being slightly higher among boys compared with girls. In fact, boys had higher percentage of having diarrhea than girls in all countries except in Libya. Overall, girls also had higher percentage of receiving any medical treatment. The sex differences in the pathophysiology of diarrhea is not well-understood in the current literature. In a meta-analysis of 16 demographic and health surveys, Wamani et al found that boys are more stunted than girls in Sub-Saharan Africa. 17 Although no clear explanation is possible to make from these findings regarding the sex-differentials, this still contradicts the notion that boys are better off than girls bias in terms of health and healthcare access.18,19 It is possible that the patterns and magnitude in the sex difference in terms of health outcomes vary across age, as different age groups share varying degrees of susceptibilities to illness conditions. These findings call for more in-depth investigation to explore the factors that underlie the sex-specific outcomes.
Considerable disparities were observed across the countries in the prevalence of diarrhea. Of the 36 countries, Libya had the highest percentage of diarrhea with more than one-quarter of the boys and girls had diarrhea during 2 weeks preceding the survey. Burundi was the only other country where more than a quarter of the children had diarrhea during last 2 weeks. Countries in the west Africa (Tchad, Senegal, Cameroon, Gambia, Ivory Coast, Mali) had the relatively higher prevalence compared with other regions with more than a fifth of the children being affected by diarrhea.
Maternal health and general literacy play a vital role in her own and children’s health status, including the chance of survival. 20 Investing in women’s education, especially promoting health literacy regarding maternal and child health through community education programs may reduce the burden of diarrhea in low-middle-income countries 21 In addition, ensuring access to improved sources of water and sanitation facilities should also be considered as an integral part of diarrhea control programs. Our analysis showed that households not having access to improved water and toilet facilities had higher risks of diarrhea which is consistent with previous findings.15,22 Regarding non-modifiable risk factors of children, male sex, higher birth order and lower age groups were associated with higher risks. The link between sex and birth order could be linked to their association with malnutrition. As shown in previous studies, male children 23 and those of higher birth order 24 are more likely to suffer from malnutrition which in turn can increase the vulnerability to infectious diseases through weakening the immune system and chances of recovery. Further research is necessary to understand the mechanism through which these non-modifiable risk factors affect the risk of diarrhea. Diarrhea prevention programs in sub-Saharan Africa should make targeted interventions through addressing these sociodemographic determinants in order to reduce the burden of diarrhea among under-five children.
Our analysis also showed a significant association between child’s age, sex and birth order with the risk of diarrhea. Findings indicate that higher age groups were associated with an increased risk of diarrhea. This finding suggests that as children grow older, they may become more exposed to various risk factors or behaviors that contribute to diarrhea incidence. We also found that girls are less likely to have diarrhea than boys, which adds an important insight regarding the sex difference in the risk of diarrhea among under-five children in the continent. The sex difference in the risk of diarrhea is perhaps due to variations in susceptibility, exposure, or response to risk factors between boys and girls. Understanding the underlying factors contributing to these differences is vital for designing appropriate intervention strategies. Further research should explore the biological and socio-cultural influences to better understand the mechanisms behind the observed sex differences. Additionally, our analysis revealed that higher birth order was associated with higher risk of diarrhea. This finding is in line with previous research, indicating that children with a higher birth order may experience increased exposure to infectious diseases. 25 Factors such as overcrowding and reduced attention to hygiene practices could contribute to the higher risk of diarrhea among children with higher birth orders.
Based on the results of this study, several policy recommendations can be made to prevent diarrhea. Firstly, given that the risk of diarrhea decreases with higher wealth quintiles, policies that aim to improve household wealth could have a significant impact on reducing diarrhea prevalence. Secondly, interventions to increase access to improved water and sanitation facilities in households and communities could be effective in preventing diarrhea as well. Thirdly, given the association between rural residency and lower risk of diarrhea, policies targeting rural areas could be particularly effective. Fourthly, policies could be implemented to target vulnerable populations, such as children in higher age groups and with lower birth order, providing age-appropriate education on hygiene practices and addressing their specific needs. Finally, policies that aim to increase maternal education could help prevent diarrhea in children, such as providing education and training programs for mothers on hygiene practices and childcare.
The current analysis is the largest of this kind to report the among- and within-country disparities in the prevalence of diarrhea among under-five children in sub-Saharan Africa. The datasets were nationally representative and therefore the findings are of good external validity. Household poverty was assessed by a widely applied method based on assess index, instead of direct income which is more prone to reporting bias, and thereby increasing the robustness of the indicator. Inclusion of data from all the available countries provide a contrasting scenario and clearer understanding of the distribution of the disease across the region. There are several limitations of this study that need to report as well. First of all, diarrhea was assessed by mothers’ observation which could be biased due to an incorrect understanding of the symptomology. Secondly, the data were secondary and thus we had no control over the choice of predictor variables that are relevant when assessing the risk factors of diarrhea. Dietary, self-efficacy and personal hygiene-related factors are also associated with diarrhea which were not available in the surveys. The generalizability of the findings can also be low beyond the selected countries. Last but not least, the data were cross-sectional and therefore no causality of the associations can be inferred. Further research is necessary on the effectiveness of interventions to reduce the burden of diarrheal diseases in sub-Saharan Africa.
Conclusions
In conclusion, our analysis revealed that the prevalence of diarrhea among under-five children in sub-Saharan Africa was higher among boys compared to girls, except in Libya. Being a female child and of higher age were identified as protective factors, while being of higher birth order was associated with an increased risk of diarrhea. Household-level risk factors such as lower wealth status, urban residence, and lack of access to improved water and sanitation facilities were significantly associated with higher risks of diarrhea. Based on these findings, it is crucial to address sociodemographic determinants and implement targeted interventions. Recommendations include improving access to clean water and sanitation facilities, promoting maternal health and education, and targeting interventions toward marginalized communities. By addressing these factors, we can effectively reduce the burden of diarrheal diseases and improve the health outcomes of under-five children in sub-Saharan Africa.
Footnotes
Appendix
List of Countries and Respective Sample Size.
| Country | Survey year | n | % |
|---|---|---|---|
| Angola | 2015 | 4259 | 1.69 |
| Angola | 2016 | 4326 | 1.72 |
| Benin | 2017 | 4211 | 1.68 |
| Benin | 2018 | 4337 | 1.73 |
| Burkina Faso | 2010 | 9782 | 3.89 |
| Burundi | 2016 | 4894 | 1.95 |
| Burundi | 2017 | 3470 | 1.38 |
| Cameroon | 2011 | 7072 | 2.81 |
| Comoros | 2012 | 1932 | 0.77 |
| Congo | 2011 | 5281 | 2.10 |
| Congo | 2012 | 900 | 0.36 |
| Congo DR | 2013 | 8129 | 3.23 |
| Congo DR | 2014 | 2495 | 0.99 |
| Eswatini | 2006 | 1470 | 0.58 |
| Eswatini | 2007 | 413 | 0.16 |
| Ethiopia | 2008 | 6866 | 2.73 |
| Gabon | 2012 | 3845 | 1.53 |
| Gambia | 2013 | 5198 | 2.07 |
| Ghana | 2014 | 4141 | 1.65 |
| Guinea | 2018 | 5165 | 2.05 |
| Ivory Coast | 2011 | 476 | 0.19 |
| Ivory Coast | 2012 | 4546 | 1.81 |
| Kenya | 2014 | 14 428 | 5.74 |
| Lesotho | 2014 | 2389 | 0.95 |
| Libia | 2013 | 4971 | 1.98 |
| Madagascar | 2008 | 1656 | 0.66 |
| Madagascar | 2009 | 6513 | 2.59 |
| Malawi | 2015 | 8975 | 3.57 |
| Malawi | 2016 | 3939 | 1.57 |
| Mali | 2018 | 6049 | 2.41 |
| Mozambique | 2011 | 7136 | 2.84 |
| Namibia | 2013 | 3593 | 1.43 |
| Niger | 2012 | 7295 | 2.90 |
| Nigeria | 2018 | 20 374 | 8.11 |
| Rwanda | 2014 | 2083 | 0.83 |
| Rwanda | 2015 | 3687 | 1.47 |
| Sao Tome | 2008 | 1334 | 0.53 |
| Sao Tome | 2009 | 55 | 0.02 |
| Senegal | 2017 | 8021 | 3.19 |
| Sierra Leone | 2013 | 7635 | 3.04 |
| South Africa | 2016 | 2810 | 1.12 |
| Tanzania | 2015 | 4961 | 1.97 |
| Tanzania | 2016 | 1740 | 0.69 |
| Tchad | 2014 | 3355 | 1.33 |
| Tchad | 2015 | 7041 | 2.80 |
| Togo | 2013 | 1790 | 0.71 |
| Togo | 2014 | 3000 | 1.19 |
| Uganda | 2016 | 9736 | 3.87 |
| Zambia | 2013 | 5421 | 2.16 |
| Zambia | 2014 | 3521 | 1.40 |
| Zimbabwe | 2015 | 4625 | 1.84 |
| N = 251 341 | 100.00 |
Acknowledgements
The authors appreciated the support of Southwest University of Political Science & Law, University of Ottawa and Chongqing Health statistical Information Center, as well as the Chinese National Social Science Project.
Authors’ Contributions
ZFH contributed to the conception and design of the study. BG performed the primary data capture and collection. ZFH and GB performed the data analysis. ZFH, GB and ZHC contributed the final review.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Chinese National Social Science Project (No. 20XGL027). Chongqing education Commission Based Project of Humanities and Social Science (No. 21SKJD012). All the authors were independent in the design of this study, data collection and analysis, interpretation of results, writing and publishing of the manuscript.
Ethical Approval
The protocol of DHS surveys was approved by the Ethics Committee of ORC Macro Inc., Calverton, MD, USA. The study was based on analysis of anonymized secondary data available in the public domain of DHS, therefore no additional approval was necessary.
