Characterizing Undernourished Children Under-Five Years Old with Diarrhoea in Mozambique: A Hospital Based Cross-Sectional Study, 2015–2019

Diarrhoea is associated with undernutrition and this association is related to increased morbidity and mortality in children under-five. In this analysis we aimed to assess the frequency and associated factors of undernutrition in children under-five with diarrhoea. A hospital-based cross-sectional study was conducted from January 2015 to December 2019 through a surveillance system in five sentinel hospitals in Mozambique. Sociodemographic and clinical information was collected, including anthropometry. A total of 963 children were analysed. The overall undernutrition frequency was 54.1% (95% CI: 50.9–57.2), with 32.5% (95% CI: 29.6–35.5) stunting, 26.6% (95% CI: 23.9–29.6) wasting and 24.7% (95% CI: 22.1–27.5) underweight. Children from Nampula province had 4.7 (p = 0.016) higher odds for stunting compared with children from Maputo. Children whose caregiver was illiterate had higher odds of being underweight 5.24 (p < 0.001), and the wet season was associated with higher odds 1.70 (p = 0.012) of being wasted. Children born under 2500 g of weight had 2.8 (p = 0.001), 2.7 (p < 0.001) and 2.6 (p = 0.010) higher odds for being underweighted, wasted and stunted, respectively. The HIV positive status of the children was associated with higher odds of being underweight 2.6 (p = 0.006), and stunted 3.4 (p = 0.004). The province, caregiver education level, wet season, child’s birthweight and HIV status were factors associated with undernutrition in children with diarrhoea. These findings emphasise the need for additional caregiver’s education on the child’s nutrition and associated infectious diseases. More studies are needed to better understand the social context in which a child with diarrhoea and undernutrition is inserted.


Introduction
Diarrhoea is one of the main causes of child morbidity and mortality in developing countries [1] and undernutrition is one of the most important risk factors for this illness

Study Design and Population
This study uses the data of children under-five enrolled in the National Surveillance of Diarrhoea (ViNaDia). The data was collected from January 2015 to December 2019 in five sentinel sites across the three regions of the country. In the Southern region, Hospital Geral de Mavalane (HGM), Hospital Geral José Macamo (HJM) and Hospital Central de Maputo (HCM) in Maputo city; in the Centre, Hospital Geral de Quelimane (HGQ), in the Zambézia province; and in the North, Hospital Central de Nampula (HCN), in the Nampula province.
Children from 0 to 59 months admitted as inpatients or outpatients with diarrhoea disease were invited to the surveillance through their caregiver. Diarrhoeal disease was defined as the passage of three or more loose or liquid stools in the last 24 h [22,25,26].

Case Report Form
For all eligible children whose legal guardian consented to their inclusion in the surveillance, a case report form (CRF) was filled with sociodemographic and clinical variables. The socio-demographical information of the children was obtained from the child's caregiver, and the clinical data (malaria, pneumonia, child's HIV status, and others) was collected from the patient report filled by the clinicians and collected from the child health card.

Anthropometric Measurements
Anthropometric measurements (weight and height/length) were performed on each child by trained nurses. Children's weight was taken naked or in light clothing. For children under two years of age, weight was measured while the child was lying down or by subtracting the weight of the mother from the weight of the mother with her child when the appropriate scale for babies was not available in the health facility. The length of children under two years or those unable to stand alone, was measured in a recumbent position lying down. The height of children aged two or older was measured while standing.

Sample Collection and Laboratory Procedures
A single faecal sample was collected per child for laboratory analysis to test the presence of intestinal parasites. The samples were collected from children to the sample collector receptacle, kept refrigerated in cooler boxes, and then sent to the parasitology reference laboratory at the Instituto Nacional de Saúde (INS) in Maputo. Faecal specimens from sentinel sites outside Maputo province were kept at −20 • C until shipment for laboratory analysis.
The presence of parasites was examined by light microscopy using a formol-ether concentration method (protozoans and helminths) and the Modified Ziehl-Neelsen stain was used for opportunistic parasites, namely, Cryptosporidium spp., Cyclospora cayetanensis and Cystoisospora belli [29]. A child was considered positive if at least one parasite was detected by either method. Microscopy results were validated after the slides were read by two independent technicians. In the case of disagreement, an additional reading was made by a third laboratory technician.

Statistical Analysis
The CRF data and the laboratory results were introduced in Epi Info™ V3. 5 . The proportions of stunting, wasting and underweight were estimated through a 95% confidence interval (95% CI) based on the Wilson method, using the EpiTools [31]. Descriptive analyses for qualitative variables were summarized by absolute (n) and relative (%) frequencies. The asymmetric quantitative variables (e.g., the child's age) were described using median and interquartile intervals. Stunting, wasting and underweight were also used as dependent variables in multiple logistic regression models. The fitted models included explanatory variables (caregiver education level, mother's marital status, child's caregiver, caregiver age, type of house, piped water, treated water, season, number of family members, birthweight, someone with diarrhoea at home in prior seven days, malaria, pneumonia, parasitic infection, immunization, and child HIV) with a p-value below 0.20 in chi-square/Fisher's exact tests or other relevant epidemiological variables (child's age, child's sex, province, and study year). The Hosmer-Lemeshow test was used for goodness of fit for the logistic regression models. Cox and Snell R square coefficients were calculated to fitted multiple models. Adjusted Odds ratios, with a corresponding 95% CI, were obtained to quantify the magnitude of the associations. A p-value < 5% was fixed as the criteria for statistical significance.

Results
During the study period, 2256 children from 0 to 59 months old with diarrhoea were enrolled, from which 1010 were excluded due to a lack of data to calculate the zscore (sex, weight, or height) to identify their nutritional status. Based on the children's anthropometric measurements, the Anthro software by the WHO flagged a total of 283 zscore outliers, and they were also excluded from the analysis. The overall final sample size was 963 children (Figure 1). From the 963 children included, 59.6% (574/963) were male, and the median age was 12 months (IQR 9.0-19.0). The overall frequency of undernutrition was 54.1% ((521/963); 95% CI: 50.9-57.2).

Frequency of Undernutrition Status across Years and Sociodemographic Characteristics
Although no significant differences were found in the proportion of children with undernutrition by year, decreases of the percentage of underweighted  Among the provinces, significant differences were found regarding the percentage of underweight (p < 0.001) and stunted (p < 0.001) children ( Figure 3). Compared with Maputo (20.4%; 127/624), the Zambézia and Nampula provinces presented higher frequencies of underweighted children-35.9% (28/78) and 31.8% (83/261), respectivelyand of stunting-47.4% (37/78) and 49.0% (128/261), respectively. Frequencies of wasting were similar among the three provinces (p = 0.660) ( Figure 3). Table 1 gives a general description of the sociodemographic characteristics from the were similar among the three provinces (p = 0.660) ( Figure 3). Table 1 gives a general description of the sociodemographic characteristics from the children with diarrhoea per type of undernutrition; underweight, wasting, and stunting. Additionally, for the children's age (in months), a median of 13 months was obtained in all types of undernutrition (IQR 9.0-20.0 for underweight and for stunting), (IQR 8.0-20.0 for wasting). As presented in Table 1, underweight was more frequent in children whose mother was illiterate (48.4%, p < 0.001), and in children living in a house made of other types of material than reed, mud and brick (66.7%, p < 0.001). Underweight was also more frequent in children living in a household with no piped water (30.7%, p < 0.001).
Similarly, wasting was observed more frequently in children whose mother was illiterate (38.5%, p = 0.019), living in a house made of other types of material (55.6%, p = As presented in Table 1, underweight was more frequent in children whose caregiver was illiterate (48.4%, p < 0.001), and in children living in a house made of other types of material than reed, mud and brick (66.7%, p < 0.001). Underweight was also more frequent in children living in a household with no piped water (30.7%, p < 0.001).
Similarly, wasting was observed more frequently in children whose caregiver was illiterate (38.5%, p = 0.019), living in a house made of other types of material (55.6%, p = 0.045) different from reed, mud and brick. Wasting was also more frequent during the wet season (31.4%, p = 0.003).
Stunting was more frequent in children whose caregiver had no formal education (48.4%, p = 0.002), married/co-habitation mothers (35.6%, p = 0.005), and children living in a house made by mud (52.1%, p < 0.001). Stunting was also more frequent in children without access to piped water (40.8%, p < 0.001), and children living in a household with five or more family members (35.5%, p = 0.047) ( Table 1). Table 2 shows that underweight (38.4%, p < 0.001) and wasting (39.3%, p = 0.001), were more frequent in children born with low weight than children born with normal weight (≥2500 g). Underweight (52.2%, p < 0.001), wasting (47.8%, p < 0.001) and stunting (55.2%, p < 0.001) were more frequent in children HIV positive than in children HIV negative. Underweight was also more frequent in children with an enteric parasitic infection (35.5%, p = 0.028). Table 3 shows the selected models for each type of undernutrition. Cox and Snell R square values were 0.146 for underweight, 0.091 for wasting, and 0.151 for stunting.

Factors Associated with Nutritional Status in Children with Diarrhoea
Children from the Nampula province had 4.7 (OR: 4.68, 95% CI: 1.63-13.41; p = 0.016) higher odds for stunting compared to children from Maputo.
Children whose caregiver had no formal education were 5.2 times (OR = 5.16, 95% CI: 2.39-11.13; p < 0.001) more likely to become underweighted when compared with those whose caregiver had a secondary or higher education. Children whose caregiver was younger (<21 years) were less likely to be wasted (OR = 0.45, 95% CI: 0.22-0.91; p = 0.044).

Discussion
In this analysis, we assessed the frequency and risk factors of undernutrition in children with diarrhoea living in three regions of Mozambique. The overall frequencies of underweight, wasting and stunting were 24.7%, 26.6% and 32.5%, respectively. These values are higher, compared with the prevalence of underweight and wasting in the same age group in other studies in children without diarrhoea, in other settings in Mozambique [10][11][12].
Although the frequency of stunting was lower than the prevalence observed in the same age group in Mozambique at the national level, (43.0% in 2011 and 42.5% in 2015) it was the higher frequency of undernutrition in children with diarrhoea described so far [10][11][12]32]. The results of the present analysis for the three types of undernutrition are aligned with the overall data from the country that shows that stunting is the most frequent type of undernutrition in children under-five [10][11][12]32].
Despite the fact that the prevalence of people living in poverty fell from 59% (2008) to 49% (2014) according to a report on the food crisis in 2019, Mozambique is still one of the 55 countries living with food insecurity [18]. Over the last 20 years, the country has been fustigated by cyclones, floods and droughts that lead to agricultural losses and thus food unavailability, and factors such as extreme weather and conflicts that are known to lead to food insecurity in low and middle income countries [18]. All these factors could help to explain the frequencies of undernutrition observed in the present analysis across the country.
Our results from the Nampula province (49%) for stunting, corroborates with the Demographic Health Survey (DHS) conducted in the country in 2011, where the province presented the highest prevalence (55.3%) of stunted children [32]. Conditions such as food insecurity, high population density, poor health indicators and the highest percentage of poverty (44%) may explain the higher prevalence observed in this province, compared with Maputo and Zambézia [33]. In Nampula, 61.9% of the population consume water from unsafe sources, 46% of the population do not have access to improved latrines and approximately 50% of the population takes 30 to 60 min to access basic services, such as a water source, public transportation, and food markets [34].
Children whose caregiver were illiterate were significantly associated with being underweighted, when compared to those whose caregiver had a secondary or higher education level. Studies conducted in Bangladesh, Nigeria, Kenya, Tanzania, Ethiopia and Mozambique presented the same relation between the mother's education level and a child's nutritional status, although those studies were conducted in children without diarrhoea [17,[35][36][37][38][39][40]. This association can be explained by the fact that, during formal education, a woman can gain knowledge on health subjects, including nutrition aspects that could help her to better address health issues in her family and community [35,40]. Additionally, formal education could allow a better understanding of how to implement correctly the health advice given by health workers [35,40].
The results of this analysis suggest that children whose caregiver was younger (<21 years) were less likely to be wasted, but this needs to be interpreted carefully as some studies found different associations between undernutrition and a caregiver's age [36,41]. Another result from the present analysis that must be carefully interpreted is the higher frequency of stunting observed in children whose mother was married/co-habiting, reminding us that in the present multivariate analysis no association was observed between a mother's marital status and undernutrition. For undernutrition and the mother's marital status, Fernandes et al. found that a child not living with their father was less likely to be undernourished while Amadu et al. found that children whose mother was married were more likely to be wasted and stunted [41,42]. Both associations, caregiver age and mother's marital status, should be evaluated using a specific study design that includes the analysis of cofounders to better understand the relationship.
The frequency of undernutrition in children with diarrhoea was higher in children without access to piped water, although no association was observed, but the higher frequency observed can be easily explained by the quality of water from non-piped sources. Non-piped water sources tend to be contaminated with pathogens that can cause diarrhoea. The relation between the quality of water and undernutrition is indirect but well known with poor water quality being associated with stunting and thinness [43].
Drinking unimproved water was not significantly associated with any type of undernutrition. This could be explained by the fact that in some families even with treated water available, the members of the family would rather drink the untreated water because some treatment methods can change the taste of the water [44]. Another plausible justification would be the incorrect implementation of known methods to treat drinking water [16]. The results of the present analysis were similar to those found in studies conducted in Ghana and Tigray [16,45]. On the contrary, a study conducted in children aged 5-24 months in Cameroon showed that unimproved drinking water was associated with underweight, wasting and stunting [36]. Another study also revealed an association with stunting in the Democratic Republic of Congo [46]. The higher risk of undernutrition due to drinking unsafe water is related to water borne diseases by exposure to enteric pathogens [47].
A significative difference was observed between the wet and dry season with wasting. Children aged under-five with diarrhoea had higher odds of being wasted in wet compared with dry seasons. During the wet season, diarrhoea diseases are more frequently observed which is already known as a risk factor for undernutrition [16] due to food insecurity, food access and availability [48].
The number of family members has been reported as a risk factor for the occurrence of undernutrition in children under-five with diarrhoea [16]. Although in the present analysis no association was observed, the association between these two factors is justified through aspects related to hygiene, sanitation, family income, food access and availability [49].
Although the results of this study were not statistically significant for the association between malaria and undernutrition in children under-five, malaria is commonly reported to be a predictor for the occurrence of undernutrition [50].
There was a positive association between all type of undernutrition and low birthweight in the present analysis. A child born with a low weight has indeed a higher risk of developing wasting, underweight and stunting, than a child born with a normal weight [21,51]. For example, using Malawian DHS data, it was shown that a child born with a low weight is at risk of remaining undernourished during the first years of their life [51]. Additionally, low birthweight is described as a factor associated with an increased vulnerability to diarrhoeal infection [21].
In the public health literature, HIV is one of the factors identified as being associated with undernutrition [14,52,53]. Thus, many studies have showed that the frequency of underweight, wasting and stunting is higher in HIV positive children compared to HIV negative children. HIV positive children have a higher risk of developing undernutrition, and especially, underweight and stunting [52,53]. Studies have showed that the association between undernutrition and HIV in children under-five years of age, results from a reduction of food intake and poor absorption of nutrients, that is common in children HIV positive with diarrhoea [20,52,53].
Since the data analysed were from a hospital-based surveillance, and the health workers collected all the data during the routine treatments, there are two identified limitations. The first limitation is the missing information related to the independent and dependent variables that reduced the sample size, from an initial 2256 to 963 for the analysis. The second limitation is inherent to cross-sectional studies that does not allow to establish a direct cause-effect relation between the collected variables and undernutrition in these children with diarrhoeal disease. Despite the limitations, due to the lack of data on the subject in our country, this analysis brings important data on factors associated with undernutrition in children with diarrhoea that can guide future studies, design actions for prevention and can improve childcare.

Conclusions
The frequency of stunting varied across the provinces, being higher in the Nampula province. The mother's education level (illiterate) was associated with being underweight. The wet season was associated with being wasted. Children born with a low birthweight were related to the three types of undernutrition, and a HIV positive status was related to underweight and stunting.
The findings of this analysis emphasise the need for additional caregiver education activities on children's health with a focus on diarrhoea and undernutrition and infectious diseases frequently associated with diarrhoea. Our results highlight the need to address the efforts during the antenatal care and children's health visits to disseminate and promote actions that can help prevent the development of any type of undernutrition. The findings of this analysis also highlight the importance of the integrated management of childhood illness, which can help reduce the morbi-mortality in children due to health conditions associated with undernutrition.
Studies to understand better the social context of children with diarrhoea and undernutrition are also recommended to evaluate the strength and weakness of the health system to implement an integrated management of diseases in children.