Portal de Periódicos da CAPES

A population-based study of how children are exposed to saliva in KwaZulu-Natal Province, South Africa: implications for the spread of saliva-borne pathogens to children

2010; Wiley; Linguagem: Inglês

10.1111/j.1365-3156.2010.02474.x

1365-3156

Lisa Butler, Torsten B. Neilands, Anisa Mosam, Similo Mzolo, Jeffrey N. Martin,

Polyomavirus and related diseases

Objectives In sub-Saharan Africa, many viral infections, including Epstein–Barr virus, cytomegalovirus, Kaposi's sarcoma-associated herpesvirus and hepatitis B are acquired in childhood. While saliva is an important transmission conduit for these viruses, little is known about how saliva is passed to African children. We endeavoured to identify the range and determinants of acts by which African children are exposed to saliva. Methods To identify the range of acts by which African children are exposed to saliva, we conducted focus groups, semi-structured interviews and participant observations in an urban and a rural community in South Africa. To measure the prevalence and determinants of the identified acts, we administered a questionnaire to a population-based sample of caregivers. Results We identified 12 caregiving practices that expose a child's oral–respiratory mucosa, cutaneous surfaces or anal–rectal mucosa to saliva. Several acts were heretofore not described in the contemporary literature (e.g., caregiver inserting finger lubricated with saliva into child's rectum to relieve constipation). Among 896 participants in the population-based survey, many of the acts were commonly practised by all respondent types (mothers, fathers, grandmothers and siblings). The most common were premastication of food, sharing sweets and premastication of medicinal plants that are spit onto a child's body. Conclusions African children are exposed to saliva through a variety of acts, practised by a variety of caregivers, with no single predominant practice. This diversity poses challenges for epidemiologic work seeking to identify specific saliva-passing practices that transmit viruses. Most acts could be replaced by other actions and are theoretically preventable. Etude de population sur la façon dont les enfants sont exposés à la salive dans le KwaZulu-Natal, Afrique du Sud: conséquences dans la propagation de pathogènes transmis aux enfants par la salive Objectifs: En Afrique sub-saharienne, de nombreuses infections virales, comprenant le virus d'Epstein-Barr, le cytomégalovirus, le sarcome de Kaposi associéà l'herpèsvirus et l'hépatite B sont acquises dans l'enfance. Alors que la salive est un important moyen de transmission de ces virus, on sait peu sur la façon dont la salive est passée aux enfants africains. Nous nous sommes efforcés d'identifier l'a gamme et les déterminants des actes par lesquels les enfants africains sont exposés à la salive. Méthodes: Afin d'identifier la gamme des actes par lesquels les enfants africains sont exposés à la salive, nous avons organisé des discussions de groupes, des entretiens semi-structurés et des observations des participants dans une communauté urbaine et une communauté rurale en Afrique du Sud. Pour mesurer la prévalence et les déterminants des actes identifiés, nous avons administré un questionnaire à un échantillon de soignants de la population. Résultats: Nous avons identifié 12 pratiques de soins qui exposent la muqueuse oro-respiratoire, les surfaces cutanées ou de la muqueuse ano-rectale de l'enfant à la salive. Plusieurs actes n'étaient pas jusqu'ici décrites dans la littérature contemporaine (e.g., l'insertion de doigts lubrifiés avec de la salive dans le rectum de l'enfant à soulager la constipation). Chez les 896 participants à l'enquête de population, de nombreux actes étaient couramment pratiqués par tous les types de répondants (mères, pères, grands-mères, frères et sœurs). Les actes les plus courants étaient la premastication de la nourriture, le partage des bonbons et la prémastication de plantes médicinales crachées ensuite sur le corps d'un enfant. Conclusions: les enfants africains sont exposés à la salive par divers actes, pratiqués par une variété de soignants, sans aucune pratique unique dominante. Cette diversité pose des défis pour les travaux épidémiologiques qui cherchent à identifier des pratiques spécifiques passant la salive et transmetteuses de virus. La plupart des actes pourraient être remplacés par d'autres actions et sont théoriquement évitables. Estudio basado en la población sobre como los niños están expuestos a saliva en la provincia de KwaZulu-Natal, Sudáfrica: Implicaciones sobre la diseminación en niños de patógenos transmitidos a través de la saliva Objetivos: En África sub-Sahariana, muchas infecciones virales, incluidos el virus de Epstein-Barr, el citomegalovirus, el herpes virus asociado al sarcoma de Kaposi, y el virus de la hepatitis B son adquiridos durante la niñez. Mientras que la saliva es un medio importante de transmisión para estos virus, poco se sabe sobre como la saliva es transmitida a los niños Africanos. Nuestro objetivo era identificar el rango y los determinantes de los actos por medio de los cuales los niños son expuestos a la saliva. Métodos: Para identificar el rango de actos mediante los cuales los niños Africanos se exponen a la saliva, realizamos grupos de discusión focalizada, entrevistas semi-estructuradas, y observaciones participativas en una comunidad urbana y una comunidad rural en Sudáfrica. Para medir la prevalencia y los determinantes de los actos identificados, administramos un cuestionario a una muestra basada en la comunidad de cuidadores. Resultados: Hemos identificado 12 prácticas de cuidadores que exponen la mucosa oral-respiratoria de los niños, las superficies cutáneas, o la mucosa anal-rectal a la saliva. Algunos actos que aquí se reportan no habían sido previamente descritos en la literatura contemporánea (por ej. el cuidador insertando el dedo lubricado con saliva en el recto del niño para aliviar el estreñimiento). Entre 896 participantes del estudio basado en la población, muchos de los actos eran comúnmente practicados por todo tipo de respondedores (madres, padres, abuelos y hermanos). Los más comunes eran el masticado previo de los alimentos, el compartir caramelos, y el masticado previo de plantas medicinales posteriormente escupidas sobre el cuerpo del niño. Conclusiones: Los niños africanos están expuestos a la saliva por medio de una variedad de actos, practicados por una variedad de cuidadores, sin la existencia de una práctica individual predominante. Esta diversidad plantea retos en el trabajo epidemiológico a la hora de identificar prácticas específicas de transmisión de saliva que puedan resultar en la transmisión del virus. La mayoría de los actos podrían reemplazarse con otras acciones y son en teoría prevenibles. A variety of chronic viral infections are acquired in childhood in sub-Saharan Africa. While some organisms (e.g., herpes simplex virus 1) have only minor clinical consequences, others such as Epstein–Barr virus (EBV, causal agent of Burkitt's lymphoma)(De-The et al. 1978); cytomegalovirus (CMV, a causal agent of several end-organ diseases) (Gandhi & Khanna 2004); Kaposi's sarcoma-associated herpesvirus (KSHV, causal agent of Kaposi's sarcoma)(Chang et al. 1994); and hepatitis B virus (HBV, causal agent of hepatocellular carcinoma) (Beasley et al. 1981) are responsible for substantial morbidity and mortality. The importance of these infections, in particular CMV and KSHV, has been magnified by the HIV epidemic (Chokunonga et al. 2000; Wabinga et al. 2000; Heiden et al. 2007; Mosam et al. 2009). Horizontal, non-sexual transmission of these viruses is suggested by a direct relationship between age and prevalence of infection (Biggar et al. 1978; Krech & Tobin 1981; Whittle et al. 1983; Stroffolini et al. 1993; Martinson et al. 1998; Mayama et al. 1998; Gessain et al. 1999; Mbulaiteye et al. 2006). While most children in sub-Saharan Africa are infected with EBV and CMV in early childhood (Biggar et al. 1978; Krech & Tobin 1981; Stroffolini et al. 1993; Mbulaiteye et al. 2006), the majority of HBV and KSHV infections occur after 1 year of age but before puberty (Martinson et al. 1998; Mayama et al. 1998; Gessain et al. 1999). Common to EBV, CMV, KSHV and HBV is their presence in saliva (Bello & Whittle 1991; Zhevachevsky et al. 2000; Dedicoat et al. 2004; Mbulaiteye et al. 2006), a recognised conduit for transmission for these viruses and a likely explanation for a pattern of horizontal non-sexual spread among African children (Martinson et al. 1998; Dedicoat et al. 2004). However, little is known about the circumstances by which African children are exposed to saliva and whether such exposures are preventable. With respect to individual organisms, some have hypothesised that KSHV is spread via premastication of food (Plancoulaine et al. 2000; Mbulaiteye et al. 2003) or via exposure to saliva used to relieve the itch of an arthropod bite (Coluzzi et al. 2002). For HBV, one report has implicated acts such as the sharing of chewing gum or candy, dental cleaning materials and bath towels (Martinson et al. 1998). More broadly, an historical review of ethnographies from the early to mid-20th century revealed a range of practices in which saliva is used in medicinal, ritualistic or feeding practices (Wojcicki 2003). More recently, a qualitative study also described a variety of practices whereby adults spread saliva to children in Zambia (Wojcicki et al. 2007). While this prior work suggests a range of practices in which saliva is passed to children, it is not known how often these various practices occur in contemporary communities, which are most common, and which contacts of children most commonly perform the practices. Aside from immunisation, which currently only exists for HBV, it is only by population-level knowledge of how saliva is passed to children can progress be made in averting transmission of saliva-borne viruses. To comprehensively understand how children are exposed to saliva in sub-Saharan Africa, we first conducted a series of qualitative studies in South Africa to identify the range and purposes of practices by which saliva is passed to children. This qualitative work informed the construction of a quantitative questionnaire to assess the prevalence, frequency and determinants of the identified acts, which was subsequently administered to a population-based sample of individuals who reside with children. The study was conducted in two phases in 2003 (Phase 1: October; Phase 2: November–December), in two settings within KwaZulu-Natal Province, South Africa: Cato Manor, an urban community near Durban, and KwaXimba, a rural community. To identify the range of acts by which children are exposed to saliva, we conducted focus groups, semi-structured interviews and participant observations in KwaXimba and Cato Manor. Focus groups comprised individuals who commonly interact with children ≤6 years old, including mothers, fathers, grandmothers, grandfathers, older siblings and traditional healers. Participants were purposively recruited with the assistance of local traditional leaders and community health workers. Each focus group consisted of 7–10 people, lasted 1.5–2 h, and was conducted in the local vernacular (isiZulu). Topics included which individuals are directly responsible for the everyday care of children, common ailments and treatments for children involving the eyes, nose, mouth, ears, skin and rectum, and how saliva plays a role, if at all, during caregiving or play. Next, semi-structured interviews were conducted with mothers, fathers, grandparents and traditional healers. Finally, participant observations were conducted in homes and community settings where young children interacted with older children and/or adults (e.g., church, market, crèche). During the observations, the author (L. M. Butler) lived and worked in close contact with families. Based on information collected during the qualitative phase, we developed two structured quantitative questionnaires regarding the practise of specific acts by which saliva is passed to children. One questionnaire was designed for adults (mothers, fathers or grandmothers) who reside with children ≤6 years old, and another was designed for older siblings (7–18 years of age) of children ≤6 years old. Grandfathers were excluded given that their role in the day-to-day care of young children was relatively limited. The questionnaires were administered to individuals residing in a previously described population-based sample of homes in either Cato Manor or KwaXimba in which one or more children ≤6 years old resided (Butler et al. 2009a). In brief, we examined aerial photographs and, using roads as landmarks, circumscribed the two communities and divided each into sections. Households from each section were then randomly sampled. At each approached household, the team determined whether a child ≤6 years old resided within. Within each household deemed eligible for inclusion, a household census was taken and one eligible respondent (i.e., mother, father, grandmother, brother or sister of child/children <6 years old) was randomly selected. Upon provision of verbal informed consent, each respondent was administered a structured questionnaire regarding acts involving the passage of saliva to a child and socio-demographic characteristics. Participants were asked if they had ever engaged in each act with a child ≤6 years of age in the household and about the frequency of their practise of each act over the past 6 months. Logistic regression was used to estimate independent associations between lifetime practise of each saliva-passing act and a priori selected caregiver-related socio-demographic characteristics (i.e., age, sex, residence, ethnicity, education) and household density. To adjust for possible confounding, all potential explanatory variables were retained in these models. We also assessed, by use of interaction terms, whether associations between respondent's gender and each act were modified by respondent's age or place of residence. Institutional review board approval was obtained from the University of California, San Francisco Committee for Human Research and the University of KwaZulu-Natal Biomedical Research Ethics Committee. We conducted 15 focus groups with 106 participants [27 mothers (three groups), 25 fathers (three groups), 24 grandmothers (three groups), 20 siblings (two groups with girls, two groups with boys) and 10 traditional healers (one group with women, one group with men)]. Semi-structured interviews were conducted with 10 mothers, 10 fathers, five grandmothers, five grandfathers and five traditional healers. Participant observations were made over 14 non-consecutive days in the homes of children ≤6 years of age and in community settings (e.g., church, marketplace, ceremonial gatherings). These qualitative methods revealed 12 practices by which children are exposed to saliva, which fell into three categories: those that (i) expose oral–respiratory mucosa to saliva, (ii) expose cutaneous surfaces to saliva and (iii) expose anal–rectal mucosa to saliva (Table 1). To determine the population-level prevalence and determinants of the 12 identified acts by which African children ≤6 years of age are exposed to saliva, we enrolled 258 mothers, 198 fathers, 204 grandmothers and 236 siblings (97 brothers and 139 sisters) who reside with children ≤6 years old (n = 896) (Table 2). There were no refusals. The majority of participants were Zulu, and approximately half (51%) lived in crowded conditions (≥1.5 persons per room). While most of the parents had a secondary level of education or more, the majority of grandmothers had a primary level education or less. About 10% of grandmothers, 6% of fathers and 2% of mothers reported practising as a traditional healer. Of all 896 participants in the population-based survey, 93% reported engaging in one or more of the 12 identified acts involving passage of saliva to children ≤6 years old residing in their household at some point during the child/children's lifetime. These practices were widespread across respondent types (Table 3). The majority of the 660 adult respondents (68%) also reported engaging in one or more of the 12 identified acts with children younger than 2 years old. More than 90% of all participants reported having engaged in at least one of the eight identified acts that expose a child's oral–respiratory mucosa to saliva (Table 3). Among all participants, the most commonly reported acts were sharing sweets (68%), premastication of food (66%) and premastication of medicinal plants and spitting or rubbing the mixture onto the child's face, head or body (64%). Fewer than half of participants reported having used their tongue, or finger or a cloth moistened with saliva, to clean a child's face or eyes (27%) or hands or fingers (29%) and blowing mpuphu (powdered muthi – a traditional Zulu medicine) into a child's nostrils with their mouth or through a reed or pipe (29%). Somewhat fewer reported sharing a toothbrush/toothstick (16%) or sucking mucus from a child's nostrils (16%). One-fifth (20%) of all participants reported engaging in one or both of the acts that expose a child's cutaneous surfaces to saliva (Table 3). Among all participants, 16% reported using their tongue, or finger or a cloth moistened with saliva, to soothe the itch from an arthropod bite, and 13% reported using their tongue, or finger or a cloth moistened with saliva, to clean a cut or scrape. Approximately one-quarter (27%) of adult respondents (n = 660) reported engaging in one or both of the acts that expose a child's anal–rectal mucosa to saliva (Table 3). Among adult participants, 16% reported having inserted their finger lubricated with saliva into the rectum of a child for the purpose of relieving constipation, and 19% reported blowing water or mpuphu through a small reed or pipe into the rectum of a child for healing purposes. To estimate the frequency of engagement in each of the 12 identified acts, participants who stated that they had ever practised each act were asked how frequently they had engaged in the act in the prior 6 months. Acts that were most commonly reported – premastication of food, sharing sweets and premastication of medicinal plants that are spit and rubbed onto a child's head, face and/or body – were reported to be practised more frequently in the past 6 months (Table 4). For these acts, between 85% and 88% of all participants who reported performing an act in the prior 6 months performed it more than once. For those acts that were the least commonly reported (i.e., practised by fewer than 20% of all respondents), the reported frequency of performance was also lower. Specifically, between 69% and 82% (depending upon the act) of all participants who reported performing these less common acts performed it more than once. For most of the 12 acts, mothers and grandmothers engaged in these acts more commonly than fathers, even after adjustment for age, ethnicity, place of residence, education and household density (Table 5). In only one act (blowing mpuphu into a child's nostrils through reed/pipe) was there strong evidence of a difference between mothers and grandmothers. Other than caregiver type, we also found that older age, residence in the rural community and female gender were independently associated with the practise of most of the acts (Tables 6 and 7). Specifically, for 9 of the 12 acts, the odds of engaging in that act were significantly increased with advancing age. For 11 of the 12 acts, the odds of engaging in that act were two to three times higher for participants living in the rural than in the urban community. For 9 of the 12 acts, the odds of engaging in that act were 2.3–27 times higher for female than male participants. Household density of the participants was also influential, albeit in just 4 of the 12 acts, with higher density being associated with more common practise of the acts. For each of the 12 identified acts, we assessed whether associations between respondent's gender and each act were modified by respondent's age or place of residence. For only one act – premasticating food – the effect of respondent's gender differed by respondent's age (pinteraction < 0.05)(Table 6). For this act, there was no evidence of a difference between males and females when limited to individuals <15 years old, but women performed the act more commonly in the older age groups. The effect of respondent's gender differed by respondent's place of residence (i.e., urban vs. rural) for three acts – sharing sweets (pinteraction < 0.01), inserting a finger lubricated with saliva into the rectum (pinteraction < 0.01) and blowing water or mpuphu through a reed or pipe into the rectum (pinteraction < 0.01)(Tables 6 and 7). The prevalence of EBV, CMV, HBV and KSHV infection is high among children in sub-Saharan Africa, and because it commonly harbours these viruses, saliva is likely the chief conduit in their transmission. However, other than for HBV, any ability to thwart transmission of these viruses has been halted by the scant knowledge of how African children are exposed to saliva. While some acts involving the use of saliva for healing or other traditional purposes has been described in recent qualitative work (Wojcicki 2003; Wojcicki et al. 2007), little is known about the quantitative population-level frequency of such acts in contemporary African communities. In what we believe is the first population-based investigation, we identified specific caregiving acts that expose oral–respiratory and anal–rectal mucosa as well as cutaneous surfaces of children to saliva. While the diversity of acts and practitioners of these acts poses substantial difficulty for epidemiologic work seeking to identify the independent role of each act for transmission of specific viruses, most of the acts could be replaced by other actions and are thus theoretically preventable. While several of the acts identified in this study have been described previously (e.g., premastication of food, application of saliva to cuts or wounds, sharing of sweets/chewing gum and sharing of dental cleaning materials) (Martinson et al. 1998; Wojcicki 2003), we identified several acts that have not been heretofore recognised in the ethnographic or biomedical literature. For example, in the realm of acts which expose a child's oral–respiratory mucosa to saliva, we identified the practise of blowing mpuphu into the child's nostrils for healing purposes. For acts that expose a child's cutaneous surfaces to saliva, the practise of using saliva to soothe blood-sucking arthropod bites has been previously theoretically described (Coluzzi et al. 2002), but we now have quantitated its prevalence. Finally, we have uncovered two acts that expose a child's anal–rectal mucosa to saliva, namely the insertion of a caregiver's finger lubricated with saliva into the rectum of a child to relieve constipation and a caregiver's blowing water or mpuphu from his/her mouth into a child's rectum to relieve constipation or other healing purposes. While exposure of anal–rectal mucosa to saliva has not been widely studied, recent data have demonstrated how homosexual men in the United States commonly use saliva as a lubricant in their anal sexual practices (Butler et al. 2009b). This behaviour has been hypothesised to account for high KSHV prevalence in homosexual men, which is otherwise uncommon in the United States. That children in Africa – an area endemic for KSHV – also have their anal–rectal mucosa exposed to saliva gives special impetus for the investigation of practices that expose anal–rectal mucosa to saliva for their transmission of KSHV. In summary, there are many practices that expose children's oral–respiratory, anal–rectal mucosa or cutaneous surfaces to saliva, and thus may serve as the route by which various pathogens, such as HBV, KSHV, CMV and EBV, are spread. In addition to the diversity in the types of acts that expose children to saliva, we also observed substantial variability across acts in their frequency. While it might be reasonable to assume that those acts practised most commonly are the most relevant culprits in transmitting saliva-borne pathogens, this may not necessarily be the case. Indeed, it may be the case for some pathogens – like HBV and KSHV – that saliva-passing acts that are only performed by a minority of the population are responsible for why some persons, but not most, become infected. We found that the frequency of engaging in most acts depended on the relationship of the respondent to the child. For example, while we observed little difference between mothers and grandmothers for acts related to the direct care of children (e.g., acts performed for the purpose of feeding or cleaning a child or for curative purposes), significantly fewer fathers reported engaging in most of those acts. The practical implication of these caregiver differences on transmission of saliva-borne pathogens cannot be determined with our data. Specifically, even caregivers who practise the acts less commonly could still be very relevant in pathogen transmission if the acts are efficient in biological infectivity. Thus, our findings that all acts were performed by at least some of the various caregiver types serve to highlight the potential role of non-maternal sources in the spread of saliva-borne pathogens. There are several limitations of our work. First is the self-reported nature of the data. However, given that there are currently no community prevention messages pertaining to the acts described here, we do not believe that caregivers underreported or over reported these acts to provide socially desirable responses. Second, we did not collect data on individuals' beliefs or attitudes about engaging in acts involving the passage of saliva to children. This information would be useful to optimise a culturally appropriate educational intervention aimed at reducing practise of these acts, but this was outside of the scope of this study. Third, we do not have data on practices of caregivers or other children who live outside of the household. Such persons could have equally high prevalence of some acts, such as sharing sweets, but lower prevalence of most caregiving acts where saliva is passed. We concede that the existence of these interactions that occur outside of the household makes epidemiologic work, as described later, even more challenging. There are several implications of these findings with respect to epidemiologic research. First, it will be difficult to establish an independent relationship between exposure to saliva via any specific act and infection with saliva-borne viruses. The presence of so many closely correlated acts practised by different types of caregivers indicates that only detailed prospective measurement will be able to distinguish the independent relationships. Second, to understand transmission of viruses such as HBV, KSHV, CMV and EBV within a community, we must identify and understand cultural and other practices in a given setting. While this study has established a number of acts involving saliva passage to children in two communities in South Africa, it is possible that there are different practices in other communities in Africa. Most of the practices we identified may be amenable to modification. Even in the absence of epidemiologic data that shows an independent relationship between of any of these acts and viral transmission, we could consider age- and culturally appropriate health education that aims to modify the practise of these acts. For example, a rubber bulb syringe may be used to relieve constipation, rather than using one's finger lubricated with saliva for the same purpose. While we recognise that some of these acts are important culturally (e.g., application of premasticated plant material on children's bodies to ward off 'bad spirits'), we believe that consideration is needed as to their value relative to potential health implications. This is particularly the case for some of the acts more than others. For example, premastication of food has been associated with transmission of group A streptococci (Steinkuller et al. 1992), Helicobacter pylori (Taylor & Blaser 1991) and, most recently, HIV (Gaur et al. 2009). At a minimum, education is needed within the community to explain the potential risks. Finally, because there will likely be persistent tension between the potential risks and benefits of these acts, further research is required to determine which acts contribute the most to the transmission of saliva-borne pathogens to children in sub-Saharan Africa. Finally, to optimise educational programs that aim to explain and reduce the risks of exposing children to saliva, further research is also required to better understand individuals' motivations for engaging in such practices. We are grateful to the communities of KwaXimba and Cato Manor in South Africa for welcoming our study team and to the study participants and their families for their cooperation. We thank the field staff for their efforts: S.Maphumulo, T.I.Mbumbe, B.Mkhize, M.I. Mlaba, M.I. Mngadi, D. Ndlovu, E.S.B. Ndlovu, J.R. Ngcobo, S. Nzuza, M. Peta, and N.N. Pheta. We are grateful to Kate Steiner Scott and Xuan Li for data management. Finally, we acknowledge the following colleagues for their helpful review of this work: Arthur Reingold, Alan Hubbard and Gertrude Buehring. This work was supported by National Institutes of Health individual grants, the University of California, San Francisco-Gladstone Institute of Virology and Immunology Center for AIDS Research and the Center for AIDS Prevention Studies; it was presented in part at the 11th International Conference on Malignancies in AIDS and Other Immunodeficiencies: Basic, Epidemiologic and Clinical Research. Bethesda, Maryland, October 6–7 2008.