Diarrhea and health inequity among Indigenous children in Brazil: results from the First National Survey of Indigenous People’s Health and Nutrition

In 2006, the prevalence of diarrhea among Brazilian children < 5 years, as reported by caregivers for a 15-day recall period, was 9.4% [24]. Contrasting this figure with the seven-day prevalence of the present study highlights the importance of diarrhea in the health profiles of Indigenous peoples in Brazil. According to results of the First National Survey of Indigenous People’s Health and Nutrition, almost a quarter of Indigenous children throughout the country had diarrhea during the previous week. This rate reached 38.1% in the North region. The degree of health inequity between Indigenous and non-Indigenous children is even more apparent when viewed diachronically. The prevalence of diarrhea among Indigenous children in 2008–2009 was still vastly higher than it was for the national population over a decade before (14.8% in 1996) [8]. In addition to this elevated frequency, analyses show that a wide range of demographic, socioeconomic, and heath variables were associated with the occurrence of diarrhea among Indigenous children.

The most vulnerable age group with respect to diarrhea was children between 6 and 23 months, who presented 101% greater chance of having had diarrhea during the previous week than infants < 6 months. This chance fell to 40% in the group of children between 24 and 35 months. The relatively lower chance of diarrhea among children < 6 months may be closely related to the protective effect of breastfeeding for infectious diseases in general and diarrhea in particular [25]. In culturally differentiated populations, such as Indigenous peoples, breastfeeding practices may differ from the national populations to which they pertain.

There are no epidemiological or systematic anthropological studies of breastfeeding practices among the Indigenous peoples of Brazil. However, among the diverse ethnographic accounts of these societies are indications that the practice is ubiquitous [26-28], which supports the National Survey results reported here. Nevertheless, although the practice appears to be nearly universal, several of its specificities as reported in the literature should be considered. Among these are observations that breastfeeding in many Indigenous societies in Brazil does not necessarily follow international health agency recommendations, including exclusive breastfeeding during the first six months of life, to reduce risk of disease and death due to diarrhea [25]. Many reports suggest that Indigenous children in Brazil may, during the first months of life, when they are also breastfeeding, receive solid foods, often premasticated, such as nuts, roasted corn, and pieces of meat, fruit, or bread. Among the Xavante, for example, Maybury-Lewis [29] noted that crying infants “…may be fed slops from the mouth of one of the women present. Grandmothers are particularly fond of feeding their tiny grandchildren in this way…” (p. 67). Similarly, Crocker [30] reported that among the Canela, “…depending on a mother’s attitude, she feeds her [2 to 4 month] baby mashes of foods, such as bananas, manioc, sweet potatoes, rice, and even brown sugar” (p. 158). Murphy and Murphy [31] observed similar baby feeding behavior among the Munduruku: “After the age of about four months, the child is fed gruels of tapioca and yam and somewhat later he is given meat, premasticated by the mother” (p. 194). The very act of breastfeeding also may be distinctive, since among many ethnic groups children may be breastfed by other lactating women in the family, such as their aunts and grandmothers [29,30].

Another practice that substantially interferes in child feeding is the use of bottles. The National Survey showed bottle feeding to be frequent and to occur early. Among Indigenous children < 6 months, 29.9% were bottle fed independently of whether or not they also breastfed at the same time. This finding is worrying since it occurs precisely during the most critical period of life – the first six months, when exclusive breastfeeding is indicated. The National Survey results do not permit identification of the liquids they receive by bottle, whether cow’s milk, baby formula, juice, or plain water. However, from the point of diarrhea transmission, administrating other foods during this critical phase of life not only compromises the child’s nutritional status and immune system development, but also favors the transmission of infectious agents by means of contaminated foods or liquids [32,33].

Diverse socioeconomic factors may be associated with risk of disease and death from diarrhea. In the case of the National Survey, maternal education and socioeconomic status remained statistically significant in the multivariable model, which finds ample support in several studies in Brazil and other countries showing these to be indicators of material resources availability and child care knowledge, especially with regard to diarrhea [34-37].

The results of the National Survey show a significant inverse association between the occurrence of diarrhea and maternal age, indicating a protective effect, particularly from the age of 40. Similar findings, widely reported in epidemiological studies in Brazil and elsewhere, are interpreted as reflecting the greater experience of older women in child care [38,39].

With respect to socioeconomic status, it is notable that diverse variables indicative of sanitation conditions and widely recognized as important in diarrhea control [36,40-43] were associated in the descriptive analyses with the occurrence of the disease among Indigenous children but were not retained in the final multivariable model. For example, children in households with a faucet, a toilet connected to a sewage system, drinking water that was filtered or derived from a public supply, or garbage collection service, were showed significantly less likely to have diarrhea. These variables were excluded from the model because they are highly correlated with the household goods socioeconomic index. Therefore, after controlling for socioeconomic condition, no association was observed for sanitation variables.

Undernutrition is among the most important health challenges for Indigenous children in Brazil, as amply demonstrated in several recent studies of specific communities and ethnic groups, as well as by the National Survey [20,44-49]. In the present study, results indicate that children with low weight-for-age showed over 20% higher risk of having diarrhea in the previous week. This finding is consistent with other research in which diarrhea figures among the infectious diseases that most affect growth and development of children < 5 years [50].

The interaction between diarrhea and undernutrition also severely compromises child immune functions, not only favoring the occurrence of new episodes of the illness, but also contributing to increased risk of other infectious diseases, such as pneumonia [50,51]. According to the National Survey, the risk of an Indigenous child having diarrhea in the previous week was 60% higher if he or she was also reported to have had pneumonia. There are several pathways that may explain the amplification of risks arising from the interaction between undernutrition, diarrhea, and acute respiratory disease. For example, a child may start with diarrhea and evolve towards acute malnutrition, finally contracting pneumonia, or, conversely, begin with a respiratory infection that worsens with an acute episode of diarrhea [51-55]. Whether occurring simultaneously (i.e., comorbidity) or as sequential infections, the interaction between diarrhea and pneumonia, aggravated by malnutrition, seriously endangers the lives of young children and contributes to the maintenance of high infant mortality rates in regions of the world marked by poverty, lack of access to basic public services, and food insecurity.

Another important aspect of diarrhea determination among Indigenous children was the association with number of household residents. Indigenous children living in households with ≥ 9 residents had 27% higher risk of contracting diarrhea than those in households with ≤ 4 inhabitants. At the same time, the risk of a child having diarrhea during the previous week increased 66% if another child in the house was also reported to have the disease during the same period. Although there is no consensus in the literature regarding the association between households with large numbers of inhabitants (overcrowding) and increased risk of diarrhea and pneumonia in children under five years, there is strong evidence pointing in this direction based on studies conducted in different context [56-58]. This potential association may derive from cosleeping practices among Indigenous peoples, which often differ from larger national populations.

Among the Bororo and Terena of Central Brazil, for example, practically all children < 2 years and about 80% of those under 10 cosleep with other family members [59,60]. According to Cohen [61], among the Kayapó-Xikrín, “…in an ideal arrangement, a toddler sleeps with her mother, while her brothers up to about 10 years sleep with their grandparents; older children sleep alone” (p. 69). Although it is not possible to generalize about the internal organization of Indigenous households nationally, child cosleeping is widespread. The anthropological literature on Indigenous peoples in Brazil is replete with references to houses lacking internal divisions and children sleeping together and with adults on the same bed or mat, as well as sharing hammocks [62,63]. Sleeping together in the same room or sharing the sleeping surface is an important cultural practice among numerous ethnic groups and is consistent with the associations observed in this study between number of residents and diarrhea risk. This point is based on the idea that pathogenic microorganisms are more easily transmitted to others in the same house, especially children, if living in close proximity to one another and under conditions of poor ventilation, dampness, and precarious sanitation installations.

In Brazil, diarrhea prevention and care for children are available through the public primary healthcare network and involve administration of anti-rotavirus vaccine and dispensation of oral rehydration therapy. Because the rotavirus vaccine was first made available through the Brazilian public health system in 2006, only children < 24 months in our sample were expected to have received the vaccine. In the age group 6 to 24 months, only 37.8% of Indigenous children nationally had received two doses of the vaccine. In the North, this proportion fell to 21.6%. The National Survey data show that at the time of fieldwork the proportion of Indigenous children who had received two doses of vaccine, as prescribed by the Health Ministry, was low and no protective effect was observed. Considering the importance of rotavirus in the determination of diarrhea among Indigenous children in Brazil [64-66] and positive evaluations of the vaccine reducing the incidence of diarrhea in the general Brazilian population [67,68], additional studies are necessary to evaluate its impact on diarrhea morbidity and mortality among Indigenous children.

The Brazilian Health Ministry adopted oral rehydration therapy for diarrhea management in 1982. Since then, its availability through the public health system has been a major factor responsible for the continued decline in infant mortality observed in recent decades [5,69]. The National Survey revealed that among Indigenous children reported as having had diarrhea in the previous week, 55.1% made use of oral rehydration therapy, whether homemade or industrialized. In the Northeast region, the frequency of use of oral rehydration therapy was lower (48.6%) than in the rest of the country, despite being the region in which Indigenous mothers had the highest number of years of schooling [10].

This study does not explain the low use of oral rehydration therapy in the management of diarrhea among Indigenous children. In addition to known deficiencies in healthcare services available to the Indigenous population in Brazil, including low coverage and quality [9,10,70], the potential influence of sociocultural factors in child diarrhea treatment also require attention. As described in several anthropological studies, the therapeutic program for any disease usually starts in the home and directly depends on interpretation of severity by family members, based on the patient’s exhibited signs and symptoms. The choice to initiate a child’s diarrhea treatment with homemade remedies or to seek rehydration salts through local health services depends on multiple factors, the analysis of which is beyond the scope of this study but certainly deserves to be investigated further [71-75].

This study has potential limitations that should be discussed. Analyses were based on data obtained through interviews with mothers, fathers, or other caregivers of Indigenous children under five years. Therefore, responses may have been influenced by local concepts and perceptions about diarrhea signs and symptoms, which vary among the country’s diverse ethnic groups [73,75,76]. Additionally, some non-parent relatives and other respondents may have lacked direct knowledge of information requested in interview questions. Although recall bias may have influenced responses, methodological studies have shown the quality of information on child diarrhea occurrence reported for a seven-day recall period to be adequate [19,77,78]. Also, the study design included only Indigenous peoples residing in federally recognized Indigenous reserves. This limitation deserves special note because many Indigenous peoples not contemplated, such as those living along roadsides, in makeshift camps, and in urban outskirts and slums, may live under even more precarious sanitation conditions. Finally, the cross-sectional design of this study does not allow conclusions about causality.