Changes in male sexual behaviour in response to the AIDS epidemic: evidence from a cohort study in urban Tanzania

AIDS. 1996 Oct;10(12):1415-20. doi: 10.1097/00002030-199610000-00015.


Objective: To examine changes in sexual behaviour among men in urban Tanzania.

Design: An observational cohort study among factory workers during 1991-1994.

Methods: Data from five follow-up visits with structured questionnaire-guided interviews and biomedical data were analysed to examine trends in sexual behaviour and sexually transmitted disease/HIV among 752 men. In-depth interviews were conducted to evaluate the magnitude of reporting bias.

Results: During the 2 years of observation, the proportion of men with more than one sexual partner during the month preceding the interview declined from 22.3 to 12.2%. The proportion of men reporting casual sex partners during the last month was almost halved: from 9.8 to 5.2%. The decline in the reporting of extramarital partners was gradual and pronounced. There were only minor changes in reported condom use, notably an increase in use with casual partners, and no changes in coital frequency. Data from in-depth interviews confirmed that reduction in sexual partners was the predominant change.

Conclusions: This study documents that, in response to the AIDS epidemic, changes in male sexual behaviour are taking place in urban areas in Africa. The predominant change among these men, who are predominantly married and aged over 25 years, is a reduction of the number of sexual partners, although condom use remains low.

PIP: During October 1991 to April 1994, health workers in Tanzania interviewed and conducted a physical examination of 752 men who had made at least 4 follow-up visits to the clinic at a large urban textile factory in Mwanza to examine trends in sexual behavior and sexually transmitted disease (STD)/HIV of factory workers and their spouses. Each man was followed for about 2 years. Researchers aimed to determine whether the intervention to reduce HIV transmission at the clinic had an impact on sex behavior. The intervention included free and effective treatment of STDs, a syphilis test at each visit, voluntary HIV counseling services, and health education activities (workshops, peer educators). The proportion of married men increased from 83.4% to 88.4% during the study period. Men were less likely to have more than 1 sexual partner in the previous month after 5 visits than before the first visit (12.2% vs. 22.3%; p 0.01 for trend). They were also less likely to have had casual partners during the previous month (5.2% vs. 9.8%; p 0.001 for trend). Married men were less likely to have extramarital relations (6.5% vs. 20.1% for regular non-cohabiting partner; p 0.01 and 2.7% vs. 8; p 0.001). Condom use in the previous month did not change significantly (2.5% vs. 3.1%; p = 0.377). It did increase significantly with a casual partner, however (7.6% vs. 27.3%; p = 0.002 for trend). Frequency of intercourse did not change. The HIV-incidence rate decreased considerably (1.77 vs. 0.66/100 person-years of observation for a rate ratio of 0.37). Multiple sex partners in the previous month was more common among men 20-29 years old or who consumed moderate or excessive amounts of alcohol than among those over age 40 or who consumed no alcohol. These findings suggest that even though sexual activity did not decline, it became more commonly with 1 partner and within marriage, especially among men 40 and older who did not drink alcohol. Condom use remained low, except with casual partners. ¿

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Acquired Immunodeficiency Syndrome / epidemiology
  • Acquired Immunodeficiency Syndrome / psychology*
  • Adult
  • Cohort Studies
  • Coitus
  • Condoms / statistics & numerical data
  • Humans
  • Male
  • Marital Status
  • Middle Aged
  • Sexual Behavior*
  • Sexually Transmitted Diseases / epidemiology
  • Tanzania / epidemiology
  • Urban Population