Variations in the amount of water contact made by individuals and in the amount of water contact made at different sites may have significant impacts on patterns of human schistosome infection. Previous studies have reported variations in the rate of water contact and differences in the sites used between age/sex classes, but there is limited information on variations in individual water contact behaviour. In this paper we report and analyse observations of essentially all water contacts made over a two week period by all individuals in a rural community in eastern Zimbabwe. The mean rate of water contact was 0.43 contacts/person/day. These data were over-dispersed, ranging from zero to 3.3 contacts/person/day; 90% of contacts were made by only 37% of the population. Contact rates were related to age (highest in 8 to 10-year-olds) but not sex, with substantial variation unaccounted for by these variables. Age and sex classes differed in types of water-related activities and the time of day of contact. A greater diversity of sites was used by children than by adults and by males than by females. Individual contact rates were correlated with intensities of infection, although the risk of infection per contact was estimated to be highest in 2 to 4-year-old children and higher for males than females. Five contact sites were used during the study period, with more than 50% of contacts occurring at just 2 sites. Different age and sex classes used different sites and there were additional site-related differences in types of activity and the time of day of use. The implications of these water contact patterns for schistosome epidemiology are discussed. In particular the results provide strong quantitative support for control programmes aimed at heavily used sites (e.g. focal mollusciciding) or at the minority of individuals making most water contacts (e.g. targeted chemotherapy).

A descriptive analysis of observed water contact activities in seven Kenyan (Akamba) communities is presented. The patterns of contact with time of day, month of year, type of activity, degree of immersion, use of soap, use of ‘kithima’ and day of week are all considered, with particular attention given to how these vary with age and sex. It is noted that (a) patterns of contact vary dramatically between these culturally rather similar communities, (b) contact usually peaks in the second decade of life, (c) generally females, especially young women, spend more time at the water than males and (d) simple (unweighted) total observed duration of contact gives a relatively inflated estimate of exposure in adults, especially young women. The methodology of observation and data handling is described in some detail.

Behavioural, parasitological and immunological data were obtained from 48 children up to 6 years old, resident in a Schistosoma haematobium endemic area in Zimbabwe. The children averaged more than 1 contact with infective water bodies every 3 days and all showed immunological evidence of exposure (an anti-cercarial and/or anti-egg antibody response). IgM was the dominant isotype and appeared in the youngest children, followed by IgA, IgE and IgG3. However, only 38 children showed evidence of infection (an anti-egg response or eggs in urine) and only 14 were excreting eggs. The best estimates from these data are that less than 1 in 100 contacts result in infection and less than 1 in 1000 result in egg output. This suggests that there may be substantial attrition of invading cercaria even in naïve individuals.

In a fishing community on Lake Albert in Uganda the pattern of intensity of Schistosoma mansoni infection 6 months after treatment with praziquantel was found to be very similar to reinfection patterns seen in previously studied endemic communities: the profile peaks sharply at around the age of 10 years falling away rapidly to much lower levels in adults. This is in stark contrast to the patterns of water contact, which differ greatly between fishing and non-fishing communities. On Lake Albert, adults appear to be more heavily exposed than children. From these observations we conclude that adults are physiologically (perhaps immunologically) more resistant to infection after treatment than children.

We review the theoretical framework for exploring the impact of individual and spatial heterogeneities in patterns of exposure and contamination and on the basic reproduction number, R0, for human schistosomes. Analysis of water contact data for 5 communities in Zimbabwe and Mali suggests that the impact is substantial, increasing R0 by factors of up to 6·5, mostly due to highly overdispersed distributions of contact rates among individuals. Several practical conclusions emerge: concentration of contacts at a single site should be avoided; the impact of control targeted at certain sites cannot be predicted without knowledge of how individuals' contacts are distributed among sites; control programmes targeted at individuals or sites contributing most to transmission can be very efficient but, conversely, will be ineffective if any of these individuals or sites are missed.

The relocation of several thousand members of the Kamba tribe from the Kyulu Hills to the Thange valley near Masongaleni in Kenya provides an excellent opportunity to study the development of the immune response to schistosomiasis mansoni in a population with little or no previous experience of the infection. An adjacent, well-established Kamba community with similar patterns of water contact provides a suitable endemic control population. The immigrants were, uniquely, examined shortly after their arrival in the endemic area, while the prevalence of infection was still low. At this time faecal egg counts peaked atypically around 30 years of age. Over the next 12–18 months infection increased rapidly, especially among teenagers, producing a pattern of infection more typical of endemic communities. This substantially narrows estimates of the time required to develop the important determinants of the age–intensity profile, supporting the notion that changes related to age per se, rather than duration of infection, dominate. Age-dependent factors might include behaviour or physiology, including immune response. This paper provides the background for continuing longitudinal studies on the development of immunological responses to this parasite.

Infection with Schistosoma mansoni was studied in 5 troops of olive baboons (Papio cynocephalus anubis) in Gombe Stream National Park, Tanzania. Three troops were infected with S. mansoni. An aggregated distribution of parasites was observed among hosts. Troop membership was found to be the most significant factor influencing parasite prevalence. Age and reproductive status had no significant effect, but there was a trend for males to acquire higher levels of infection. However, age–prevalence curves showed a high infection in young baboons declining in the older baboons. Behavioural components of exposure – as measured in water-contact pattern – may be related to parasite burden. A ‘peak shift’ between infection in different age-classes in the different troops was observed: troops with higher schistosome prevalences displayed an earlier peak in prevalence of infection. The baboon troop with the most contact with people showed highest prevalence of infection possibly due to longer exposure to the parasite than the other troops and/or higher host density.