Monitoring sexual behaviour in general populations: a synthesis of lessons of the past decade.

This supplement contains selected papers from a workshop on the measurement of sexual behaviour in the era of HIV/AIDS held at the London School of Hygiene and Tropical Medicine in September 2003. The focus was on low and middle income countries, where the majority of HIV infections occur. The motive for holding such a meeting is easy to discern. As the AIDS pandemic continues to spread and as prevention programmes are scaling up, the need to monitor trends in sexual risk behaviours becomes ever more pressing. Behavioural data are an essential complement to biological evidence of changes in HIV prevalence or incidence. Biological evidence, though indispensable, is by itself insufficient for policy and programme guidance. AIDS control programmes need to be based on monitoring of not only trends in infections but also of trends in those behaviours that underlie epidemic curtailment or further spread.

Where the action is: monitoring local trends in sexual behaviour.

OBJECTIVES: To assess the role in an overall monitoring strategy of the PLACE method of estimating local trends in sexual behaviour among individuals at social venues in areas at increased risk of HIV transmission. METHODS: Public venues identified by community informants as places where people meet new sexual partners were visited and characterised in Karaganda, Kazakhstan, in 2002 and 2003, and in a township in South Africa in 2000 and 2003. At a subset of venues, a representative sample of individuals socialising at the venue were interviewed about their sexual behaviour. The age distribution and partnerships rates of those socialising at these venues were compared by year and with national data from Kazakhstan and South Africa obtained via household surveys. RESULTS: Women and men socialising at venues were younger and had higher rates of new and concurrent partnerships than men and women interviewed in national household surveys. There was little apparent change in sexual behaviour between 2002 and 2003 in Karaganda. In the South African township, there was a reduction in sexual partnerships and an increase in condom use, possibly due to a local AIDS prevention programme. CONCLUSION: Findings from the PLACE method supplement national data on sexual behaviours with data from key populations in high transmission areas, inform local targeting of interventions, and, when subsequent rounds of PLACE are implemented, can evaluate change in target populations.

Community effects on the risk of HIV infection in rural Tanzania.

OBJECTIVES: To investigate the effect of community characteristics on HIV prevalence and incidence. METHODS: Data from an open cohort study with demographic surveillance, epidemiological surveys, and qualitative research were used to examine the association between individual and community risk factors with HIV prevalence in 1994-5 and incidence between 1994-5 and 1996-7 among men (n=2271) and women (n=2752) living in a rural area in northwest Tanzania. Using subvillages as the unit of analysis, community factors investigated were level of social and economic activity, ratio of bar workers per male population aged 18-59, level of community mobility, and distance to the nearest town. Logistic and Cox regression models were estimated to assess community effects, controlling for multiple individual factors. RESULTS: All four community factors had strong effects on HIV transmission. Men who lived in subvillages with the highest level of social and economic activity had an odds of being HIV positive that was about five times higher (OR=4.71, 95% CI=2.89 to 6.71) than those in places with low levels of activity; women in these subvillages had an odds that was twice as high (OR=1.92, 95% CI=1.27 to 2.92). After controlling for community effects, the effects of some individual factors on the risk of HIV-education, male circumcision, type of work, and number of household assets-changed notably. The association between HIV incidence and community factors was in the expected direction, but did not reach statistical significance (RR=2.07, p=0.10). CONCLUSIONS: Results suggest that community characteristics play an important part in the spread of HIV in rural Tanzania. Community effects need both to be considered in individual risk factor analyses and be given more attention in intervention programmes.

A pilot study of a rapid assessment method to identify places for AIDS prevention in Cape Town, South Africa.

A multidisciplinary approach appears promising for focusing interventions. The PLACE method (priorities for local AIDS control efforts) identifies where prevention programmes can access sexual networks containing individuals with high rates of new partner acquisition and provides information about whether sexually transmitted disease (STD) or AIDS prevention programmes are in place at those sites. The method provides useful information at relatively low cost, regardless of the phase of the HIV or STD epidemic. For example, in the growth phase of an epidemic, information on where people with high rates of new partner acquisition meet new sexual partners could be used to inform targeted surveillance or screening programmes to catch early cases.

Sociodemographic context of the AIDS epidemic in a rural area in Tanzania with a focus on people's mobility and marriage.

This analysis focuses on how sociocultural and economic characteristics of a poor semi-urban and rural population (Kisesa ward) in north west Tanzania may directly and indirectly affect the epidemiology of HIV and other sexually transmitted infections (STI). Poverty and sociocultural changes may contribute to the observed high levels of marital instability and high levels of short and long term migration in Kisesa, especially among younger adults. Marriage and migration patterns are important underlying factors affecting the spread of HIV. The most cost-effective intervention strategy may be to focus on the trading centre in which mobility is higher, bars were more common, and HIV prevalence and incidence were considerably higher than in the nearby rural villages. If resources suffice, additional work can be undertaken in the rural villages, although it is not clear to what extent the rural epidemic would be self sustaining if the interventions in the trading centre were effective.

The impact of HIV/AIDS on mortality and household mobility in rural Tanzania.

OBJECTIVE: To assess the impact of the AIDS epidemic on mortality and household mobility before and after death. DESIGN: Open community cohort study with a demographic surveillance system and two sero-epidemiological surveys. METHODS: Ten rounds of demographic surveillance were completed during 1994-1998 in the study area, which has a population of about 20 000 people in a rural ward in north-west Tanzania. Households with deaths were visited for a detailed interview, including a verbal autopsy. Data on HIV status were collected in two surveys of all residents aged 15-44 years. RESULTS: Mortality rates among HIV-infected adults were 15 times higher than those among HIV-negative adults and HIV/AIDS was associated with nearly half of deaths at ages 15-44 years. Verbal autopsies without HIV test results considerably underestimated the proportion of deaths associated with HIV/AIDS. The mortality probability between 15 and 60 years was 49% for men and 46% for women and life expectancy was 43 years for men and 44 years for women. By their second birthday nearly one-quarter of the newborns of HIV-infected mothers had died, which was 2.5 times higher than among children of HIV-negative mothers. Mobility of household members before and after death was high. In 44% of households in which the head died all members moved out of the household. CONCLUSIONS: In this rural population with HIV prevalence close to 7% among adults aged 15-44 years during the mid-1990s, HIV/AIDS is having substantial impact on adult mortality. A common response to death of a head of household in this community is household dissolution, which has implications for measurement of the demographic and socio-economic impact of AIDS.

Dynamics of male circumcision practices in northwest Tanzania.

BACKGROUND: Male circumcision status is considered an important cofactor in the spread of HIV and sexually transmitted disease. There is limited evidence that male circumcision practices in Africa may be changing. GOAL: To assess the determinants of male circumcision status in a traditionally noncircumcising ethnic group and to investigate the reasons for increasing acceptance of circumcision. METHODS: Data from a factory workers study and a rural cohort study in northwest Tanzania were used to analyze the levels and determinants of male circumcision status and assess the reliability of self-reported data. Qualitative data from focus group discussions and in-depth interviews were obtained to ascertain norms and values in relation to male circumcision. RESULTS: Male circumcision has become more popular in recent years, and 21% of 3,491 men reported themselves as circumcised. An increase in circumcision rates was observed in the rural cohort study during 1994 to 1997, though reporting inconsistencies are common. Circumcision rates were higher among men with higher levels of education and in Muslim men. Men are often circumcised in their late teens or twenties. The reasons for the increasing popularity of circumcision were investigated in group discussions and in-depth interviews. The most frequently mentioned reason was health-related; circumcision was thought to enhance penile hygiene, reduce sexually transmitted disease incidence, and improve sexually transmitted disease cure rates. CONCLUSION: Male circumcision is becoming more popular among a traditionally noncircumcising ethnic group in Tanzania, especially in urban areas and among boys who have attended secondary schools.

Adjusting ante-natal clinic data for improved estimates of HIV prevalence among women in sub-Saharan Africa.

OBJECTIVES: To find a simple and robust method for adjusting ante-natal clinic data on HIV prevalence to represent prevalence in the general female population in the same age range, allowing for fertility differences by HIV status. BACKGROUND: HIV prevalence comparisons for pregnant women and women in the general community show that prevalence in the latter is significantly higher than in the former. An adjustment procedure is needed that is specific for the demographic and epidemiological circumstances of a particular population, making maximum use of data that can easily be collected in ante-natal clinics or are widely available from secondary sources. METHODS: Birth interval length data are used to allow for subfertility among HIV-positive women. To allow for infertility, relative HIV prevalence ratios for fertile and infertile women obtained in community surveys in populations with similar levels of contraception use are applied to demographic survey data that describe the structure of the population not at risk of child-bearing. RESULTS: For populations with low contraception use, the procedure yields estimates of general female HIV prevalence of 35-65% higher than the observed ante-natal prevalence, depending on population structure. Results were verified using general population prevalence data collected in Kisesa (Tanzania) and Masaka (Uganda). For high contraception use populations, adjusted values range from 15% higher to 5% lower, but only limited verification has been possible so far. CONCLUSIONS: The procedure is suitable for estimating general female HIV prevalence in low contraception use populations, but the high contraception variant needs further testing before it can be applied widely.

Looking for change in response to the AIDS epidemic: trends in AIDS knowledge and sexual behavior in Zambia, 1990 through 1998.

This study investigates trends in AIDS knowledge and sexual behavior among men and women in urban Lusaka 1990 to 1998, and in all of Zambia, 1992 to 1998. Using data from representative surveys of urban Lusaka and of the country as a whole, population proportions were estimated to examine trends in knowledge and sexual risk behaviors. Differences in the estimated proportions between 1990 and 1998 were tested in Lusaka. In all Zambia, tests of difference were conducted between the earliest and latest years for which data were available for each indicator. A decline in premarital sexual activity was observed in urban Lusaka. In 1990, 50% of never married women reported no sexual experience, compared with 60% in 1998 (p =. 003); among men, the figures were 38% and 53%, respectively (p <. 001). Fewer women (1990, 8%; 1998, 2%; p <.001) and men (1990, 31%; 1998, 19%; p =.07) had extramarital partners. The bulk of change observed in urban Lusaka took place from 1990 to 1996; the changes in men's behavior observed between 1996 and 1998 were also observed in the national estimates for those years. National figures for other indicators from 1992 to 1998 were less encouraging. Apart from an increase in having ever used condoms, no change in women's sexual behavior was observed. Fewer men had premarital sex from 1996 to 1998 (1996, 64%; 1998, 46%; p <.001), but condom use with nonregular partners decreased among men (1996, 38%; 1998, 29%; p =.02). Prevention campaigns focused on education about AIDS and promoting safer sexual behavior appear to have made a difference in the early 1990s in Zambia. Findings from more recent years indicate that further change has stagnated. Renewed efforts are needed, particularly targeting condom use with nonregular partners.

HIV incidence and HIV-associated mortality in a cohort of factory workers and their spouses in Tanzania, 1991 through 1996.

OBJECTIVE: To describe HIV incidence and HIV-associated mortality in a cohort of factory workers and their spouses with access to adequate sexually transmitted diseases (STD) treatment services and moderate exposure to sexual health interventions. METHODS: Follow-up visits at 4-month intervals and home follow-up of those who failed to keep appointments were used to estimate HIV incidence and mortality in a cohort of factory workers and their spouses in Mwanza, Tanzania, during 1991 to 1996. RESULTS: HIV prevalence at intake was 10.5% and 15.3% among 1,594 men and 880 women, respectively. HIV incidence was just over 1/100 person-years (PY) in 1,427 men and 1.9/100 PY in 745 women. The main risk factors were related to sexual behavior and clinical evidence of an STD, although only one seventh of those with HIV seroconversions had had an STD during the period following the second-from-last follow-up visit. HIV incidence among discordant couples was 5.0 and 8.3/100 PY for 41 men and 37 women, respectively. Mortality rates were 9.0 and 7.8/1,000 PY for men and women, respectively, and 65% of male and 60% of female deaths were attributable to HIV infection. CONCLUSIONS: In this factory population with good access to and use of STD treatment and with a moderate level of exposure to HIV prevention education, HIV incidence among men and women is still over 1/100 PY, which suggests a relatively high level of program effort is needed to lower incidence. More than half of all adult deaths were attributable to HIV, but greater increases in HIV-associated mortality are likely.

Spread of HIV infection in a rural area of Tanzania.

OBJECTIVE: To assess the spread of HIV into rural areas. METHODS: Since 1994 a demographic surveillance system (with 5-monthly rounds) and open adult cohort study have been established in a rural ward in Tanzania. Two sero-surveys of all resident adults aged 15-44 and 15-46 years were conducted in 1994 1995 and 1996-1997 respectively. Qualitative data were collected on mobility, bars and commercial sex. RESULTS: Attendance of the two rounds of survey was 5820 (78%) and 6413 (80%) in 1994/1995 and 1996/1997 respectively. HIV prevalence increased from 5.8% to 6.6%. HIV incidence was 0.73 and 0.84 per 100 person years among men and women respectively. HIV incidence under the age of 20 years was low among both sexes. Striking differences in HIV prevalence and incidence were observed within the small geographic area studied: HIV prevalence in the trading center was twice that in the area surrounding the trading center (within 2 km) and three to four times that in the rural villages (within 8 km of the trading center). Aggregate level data showed significant differences between the trading center and nearby rural villages in terms of sexual behavior, commercial sex workers, mobility of the population, and alcohol use. CONCLUSION: This study documents the existence of very substantial HIV prevalence and incidence differences within a small geographic rural area. The rapid decrease in HIV prevalence within a small rural area emphasizes the importance of concentrating HIV prevention efforts on high transmission areas, such as trading centers, especially in resource-poor settings. Furthermore, this has considerable implications for monitoring the spread of HIV through sentinel sites, as such sites are typically located in high transmission areas.

Mortality impact of the AIDS epidemic: evidence from community studies in less developed countries.

BACKGROUND: The AIDS epidemic is now more than a decade old and direct evidence of mortality impact has become measurable, as indicated by an increasing number of publications presenting empirical data from less developed countries. METHODS: This review focuses on the evidence of mortality impact among adults and children in community studies. The majority of these studies are located in Africa, particularly eastern Africa, where the AIDS epidemic is conjectured to be older than in other less developed countries. RESULTS: Community studies show a two- to threefold increase in total adult mortality with an even larger increase in mortality among young adults in communities with adult HIV prevalence levels below 10%. Mortality amongst HIV-infected adults ranges from 5 to 11% per year, and more than half of all adult deaths can be attributed to HIV. HIV-infected women die at an earlier age than men and thereby lose significantly more productive years of life. Follow-up studies of incident cases are few, but population-based data indicate that the median survival time is substantially longer than originally thought on the basis of mortality amongst HIV-infected commercial sex workers. Tuberculosis incidence is on the increase, but evidence of additional impact on mortality is hitherto limited. Infant and early child mortality among children of HIV-infected mothers is two to five times higher than among children of HIV-negative mothers in follow-up studies of maternity-based and community samples. CONCLUSION: There is now empirical evidence of the mortality impact of HIV/AIDS from several community studies. The large increase in adult mortality and moderate increase in child mortality lead to dramatic falls in life expectancy. For instance, in a rural area of Uganda, which has an HIV prevalence of 8%, life expectancy has dropped from just under 60 years to 42.5 years.

PIP: A review is presented of the available empirical evidence which now exists from several community studies on the impact of HIV/AIDS upon mortality. Community studies, largely conducted in Africa, indicate a two- to three-fold increase in total adult mortality, with an even larger increase in mortality among young adults in communities with adult HIV prevalence levels under 10%. Mortality among HIV-infected adults is 5-11% per year, with more than half of all adult deaths attributable to HIV. HIV-infected women die younger than men, losing more productive years of life. The available population-based follow-up data suggest that the mean survival time with HIV/AIDS is considerably longer than originally thought given findings upon the mortality trends of HIV-infected prostitutes. Although the incidence of tuberculosis is rising, there is only limited evidence of any additional impact upon mortality. Infant and early child mortality among children of HIV-infected mothers is 2-5 times higher than among children of HIV-negative mothers in follow-up studies of maternity-based and community samples. Despite the considerable excess mortality caused by HIV/AIDS, no negative population growth has been documented due to the HIV/AIDS epidemic.

[Data on birthweight in developing countries: are surveys useful?]. / Datos sobre el peso al nacer en países en desarrollo: son útiles las encuestas?

The main source of data on birth weight in developing countries is statistics from health facilities, although most developing countries do not produce annual estimates of the incidence of low birth weight from these data. Such estimates would be subject to selection bias as the data are usually limited to babies born within health facilities, and therefore are representative of a subgroup that is markedly different from the overall population of neonates. Since 1990 the Demographic and Health Surveys programme has included questions on recalled birth weight and relative size at birth in 15 national surveys. In this article, we show that these cross-sectional surveys can provide a useful data source for making national estimates of mean birth weight and the incidence of low birth weight. The extent of misclassification of birth weight is, however, too large to use the data on relative size as birth as an indicator of low birth weight at the individual level.

Datos sobre el peso al nacer en países en desarrollo: son útiles las encuestas? / Data on birth weight in developing countries: can surveys help?

Las estadísticas de centros de atención de salud son la fuente principal de datos sobre el peso al nacer en los países en desarrollo, si bien en la mayor parte de esos países no se producen estimaciones anuales de la incidencia de bajo peso al nacer a partir de esos datos. Si se produjeran, las estimaciones serían propensas al sesgo de selección ya que los datos están limitados en general a los niños nacidos en centros de salud y, por lo tanto, representan un subgrupo decididamente diferente de la población general de neonatos. Desde 1990, el programa de Encuestas Demográficas y de Salud ha realizado 15 encuestas nacionales en las que se incluyeron preguntas sobre el peso y el tamaño relativo al nacer tal como los recordaban las madres. Este artículo muestra que las encuestas transversales pueden constituir una fuente útil de datos para estimar la media nacional de peso al nacer y la incidencia de bajo peso. Sin embargo, la clasificación errónea del peso al nacer es demasiado extensa para emplear los datos de tamaño relativo al nacer como indicador de bajo peso a escala individual

The professionals and patients involved in dental examinations are at risk for infection by various disease-causing bacteria, viruses, and fungi, such as those responsible for hepatitis, tuberculosis, herpes, and AIDS. It is known that aerosols and spatter containing pathogenic microorganisms can spread during an examination. Nevertheless, some dental clinics are designed to have multiple examination areas in the same room, with no physical barriers between them. The objective of this study was to verify the reach of spatter resulting from the use of a triple syringe and high-rotation turbine during five simulated exams in a collective clinic, bearing in mind that spattercan contain the patient's saliva and blood. To facilitate tracking of the spatter, aniline dye (pink, blue, yellow, green, and brown) was added to the water in the appropriate receptacle in each of the five units. The room, the equipment, and the patient's and operator's clothing were covered with white paper. A high concentration of spatter was observed on the chair, the operator, and the floor of each unit, and it also appeared on the chairs and trays of the surrounding units. The maximum distance reached by spatter was 1.82 m from a point on the chair corresponding to the position of the patient's mouth. During real simultaneous examinations, the surrounding chairs and their patients and operators, as well as the trays containing sterilized instruments, are within spatter range. Therefore, there is a real possibility of crossinfection, and physical barriers should be placed between the units. This study also confirmed the need for protection of the operator's face, body, hair, and arms, since these regions were heavily affected by spatter.

Consequences of adult HIV infection for outpatient morbidity and treatment costs: a prospective study in a factory clinic in Tanzania.

Most studies of the medical costs of HIV infection focus on the terminal stage of this chronic illness when the patients have developed AIDS or severe HIV disease and in-patient care dominates. Data are also needed on the medical costs during the prolonged phase of HIV infection preceding severe terminal illness and the effects it may have on the provision of outpatient care. The study population was derived from a cohort study of factory workers and their spouses in Tanzania. Morbidity and outpatient health services utilization are estimated for 1832 adults who on average had been enrolled for two years and utilized the study clinic. Among those who had been enrolled at least 2 years, 50 cases (HIV+ since enrollment) and 150 control (HIV- until last visit) were selected, matched by age, sex and income level to estimate expenditure on drugs by HIV status. There was an increase in morbidity during HIV infection: the incidence of clinical diagnoses was 30% higher among HIV-positive than among HIV-negative adults (p < 0.001). HIV-infected adults also made more frequent use of the outpatient services (23% higher utilization). Estimates of essential drug costs among the subsample showed a 15% increase for HIV infected adults compared to HIV-negative adults, caused by higher use of antibiotics and other antimicrobial drugs. The overall increase in morbidity, outpatient care services utilization and essential drug use due to HIV infection was limited, as HIV prevalence in this adult population was 11%. For example, the net proportion of all illness episodes attributable to HIV infection was 3.2%. Possible biases are discussed and suggest that our findings are a minimum estimate of the effect of adult HIV infection on outpatient care costs. There is a need for more studies in different settings to assess the impact of HIV infection on outpatient care in developing countries.

PIP: Morbidity and outpatient health services utilization are estimated for 1832 adult factory workers and their spouses in Tanzania who had on average been enrolled at the study clinic for at least 2 years and used the clinic. A subsample of 50 cases HIV-positive since enrollment and 150 controls HIV-negative up to the most recent visit was selected, matched by age, sex, and income level to estimate expenditure on drugs by HIV status. An increase in morbidity was observed during HIV infection, with the incidence of clinical diagnoses 30% higher among HIV-positive than among HIV-negative adults. HIV-infected adults also used outpatient services 23% more often than did controls. Estimates of essential drug costs among the subsample showed a 15% increase for HIV-infected adults compared to HIV-negative adults. However, the net proportion of all illness episodes attributable to HIV infection was only 3.2%. These findings likely represent a minimum estimate of the effect of adult HIV infection on outpatient care costs.

HIV infection and sexual behaviour among women with infertility in Tanzania: a hospital-based study.

BACKGROUND: Infertility is common in Africa, but virtually no data exist on HIV prevalence among infertile women. Mainly anthropological studies in Africa have shown that infertile women have higher risks of marital instability and possibly also have more sexual partners than fertile women. METHOD: This study was conducted in a hospital in northwest Tanzania during 1994 and 1995. Women presenting themselves with infertility problems to the outpatient clinic were interviewed, examined and blood was drawn. Women who came to deliver in the hospital, excluding primiparae, were taken as a control group. The analysis was limited to women > or = 24 years. In total 154 infertile and 259 fertile women were included in the study. RESULTS: HIV prevalence was markedly higher among infertile women than among fertile women: 18.2% and 6.6% respectively (adjusted odds ratio [OR] for age, residence and occupation 2.7; 95%-confidence interval [CI]: 1.4-5.3). Data on past sexual behaviour showed that infertile women had more marital breakdowns, more lifetime sexual partners and a higher level of exposure to sexually transmitted diseases (STD). CONCLUSION: Women with fertility problems appear to have higher HIV prevalence, which justifies more attention for such women in the context of AIDS programmes. In addition, caution is needed when using sentinel surveillance data from antenatal clinics to monitor HIV prevalence.

PIP: Infertility is common in Africa. Anthropological studies conducted on the continent have found infertile women to have higher risks of marital instability and possibly more sex partners than fertile women. Findings are reported from a study conducted during 1994 and 1995 in a hospital in northwest Tanzania to determine the prevalence of HIV infection among infertile women. Women presenting with infertility problems to the outpatient clinic were interviewed, examined, and blood was drawn. Women who came to deliver in the hospital, excluding primiparae, comprised the control group. A total of 154 infertile and 259 fertile women were included in the study, all age 24 years and older. 18.2% of infertile women and 6.6% of fertile women were infected with HIV. Data on past sex behavior indicated that infertile women had more marital breakdowns, more lifetime sex partners, and a higher level of exposure to sexually transmitted diseases.

Male circumcision and susceptibility to HIV infection among men in Tanzania.

BACKGROUND: Evidence from ecological studies and from studies of and sexually transmitted disease (STD) patients in sub-Saharan Africa suggests that there is a protective effect of male circumcision against HIV infection. There are, however, few population-based studies that have controlled adequately for potential confounding factors. METHODS: Data from the five population-based studies in north-western Tanzania were used to investigate the association between male circumcision and the risk of HIV infection and STD. The effects of circumcision on HIV prevalence, syphilis (positive Treponema pallidum haemagglutination; TPHA) and self-reported STD were analysed, controlling for a range of demographic and sociocultural variables, and indicators of sexual behaviour. RESULTS: In north-western Tanzania, circumcision was previously restricted to Muslims and specific ethnic groups, but is now more widespread, particularly in urban ares and among more educated men. Assessment of the reliability and validity of self-reported circumcision status showed that these data could be considered fairly accurate, although there was some tendency for circumcision to be over-reported. On univariate analysis, circumcision status was unrelated to HIV prevalence in most studies. After controlling for confounding variables, however, there was a modest but significant reduction of the HIV prevalence among circumcised men [odds ratio (OR), 0.62; 95% confidence interval (CI), 0.48-.81]. This effect appeared stronger in urban areas (OR, 0.46; 95% CI, 0.32-0.68) and roadside villages (OR, 0.65; 95% CI, 0.42-1.01) than in rural areas and islands (OR, 1.00 and 1.01 respectively). There was no association between circumcision status and syphilis serology (TPHA), but there was a positive association between circumcision and self-reported STD, although this was not significant after adjustment for confounding variables. CONCLUSION: Male circumcision has a protective effect against HIV infection in this population, which may be stronger in urban areas and roadside settlements than in the rural areas. Ethnic group and religious denomination are no longer the sole determinants of male circumcision.

Male circumcision and susceptibility to HIV infection among men in Tanzania.

BACKGROUND: Evidence from ecological studies and from studies of and sexually transmitted disease (STD) patients in sub-Saharan Africa suggests that there is a protective effect of male circumcision against HIV infection. There are, however, few population-based studies that have controlled adequately for potential confounding factors. METHODS: Data from five population-based studies in north-western Tanzania were used to investigate the association between male circumcision and the risk of HIV infection and STD. The effects of circumcision on HIV prevalence, syphilis (positive Treponema pallidum haemagglutination; TPHA) and self-reported STD were analysed, controlling for a range of demographic and sociocultural variables, and indicators of sexual behaviour. RESULTS: In north-western Tanzania, circumcision was previously restricted to Muslims and specific ethnic groups, but is now more widespread, particularly in urban areas and among more educated men. Assessment of the reliability and validity of self-reported circumcision status showed that these data could be considered fairly accurate, although there was some tendency for circumcision to be over-reported. On univariate analysis, circumcision status was unrelated to HIV prevalence in most studies. After controlling for confounding variables, however, there was a modest but significant reduction of the HIV prevalence among circumcised men [odds ratio (OR), 0.62; 95% confidence interval (CI), 0.48-0.81]. This effect appeared stronger in urban areas (OR, 0.46; 95% CI, 0.32-0.68) and roadside villages (OR, 0.65; 95% CI, 0.42-1.01) than in rural areas and islands (OR, 1.00 and 1.01 respectively). There was no association between circumcision status and syphilis serology (TPHA), but there was a positive association between circumcision and self-reported STD, although this was not significant after adjustment for confounding variables. CONCLUSION: Male circumcision has a protective effect against HIV infection in this population, which may be stronger in urban areas and roadside settlements than in the rural areas. Ethnic group and religious denomination are no longer the sole determinants of male circumcision.