Soz Pr/iventivmed 1992; 37:79 84 0303-8408/92/020079-06 $1.50 + 0.20/0 9 1992 Birkhfiuser VerlagBasel Predictors of high risk sexual behaviour in gay and bisexual men Jill Dawson 1, Ray Fitzpatrick 1, Mary Boulton 2, John McLean 2, Graham Hart 3 1 Department of Public Health and Primary Care, University of Oxford 2 Academic Department of Community Medicine, St. Mary's Hospital Medical School, London 3 Academic Department of Genito-Urinary Medicine, University College and Middlesex School of Medicine, London Gay and bisexual men are known to have made substantial adjustments to their sexual behaviour in response to AIDS t. Nevertheless studies continue to report a high frequency of homosexually active men engaging in potentially risky sexual activity, particularly, unprotected receptive anal sex 2. An understanding of such behaviour is required in order to design appropriate health educational interventions. The Health Belief Model 3 draws attention to a number of social-psychological factors that may be important influences upon health behaviour, and that may also suggest how to target health interven- tions. Thus one central variable of the HBM is the individual's perception of personal risk in relation to a disease such as AIDS in determining readiness to change behaviour 4. Related to perception of risk is the individual's level of knowledge about HIV transmission. The model also suggests that the greater the perceived seriousness or threat of a disease, the greater the likelihood of change 5. Another factor central to the HBM is the perception of costs and benefits of changing behaviour. In particular gay men may have strong attitudes regarding the costs and benefits associated with adopting safer sex 6. Individuals' sense of control over health, general self-confidence or self esteem have been suggested as elements additional to the HBM that may also be important influences upon behavioural change in this area 6. The HBM also indicates that a range of external cues may be stimuli to change. One factor that may be parti- cularly important is degree of contact with the AIDS epidemic v. This paper reports on aspects of the sexual behaviour of a sample of homosexually active men in England. Results from a longitudinal study are used to examine the extent to which high risk sexual behaviour may be predicted by the following vari- ables from the HBM: perceptions of risk, level of knowledge about HIV transmission, perceived severity and threat from the AIDS epidemic, per- ceived costs and benefits of adopting safer sex including views about condom use, personal sense of control over health generally and specifically in relation to HIV, self confidence and self esteem and contact with the AIDS epidemic. In addition the influence of mens' current partnerships is examined. Whilst not included within the normal range of variables of the HBM, relationships may be ex- pected to play a potential role in influencing high risk sexual behaviour. Methods and sample For inclusion in this study it was necessary for the subject to be a man who had had sexual contact with another man within the last 5 years. A sample of 502 men was recruited for interview from a diverse range of sources: 283 (56 %) from gay pubs, clubs and gay organisations; 96 (19 %) by referrals from those already interviewed ("snowball" sampling) and 123 (25%) from clinics of departments of genito-urinary medicine. Four main towns and cities were used: London 228 (45%), Manchester 145 (29 %), Oxford 65 (13 %) and Northampton 31 (6%). A further 33 (7%) of the sample were recruited from areas around these four centres. The interview gathered detailed information about cur- rent sexual behaviour, as well as variables from the HBM as described below. Three hundred and sixty- nine men (74%) completed a follow-up postal questionnaire 9 months later. The only significant differences between those who returned the follow up-questionnaire and the rest of the original sample were that those returning the follow-up ques- tionnaire were older (t = 2.28; p < 0.05), had spent longer in full-time education (chi square 6.35, p < 0.05) and were less likely to be unemployed (chi square 11.65, p < 0.001). In the interview, men were asked to assess their personal risk in relation to HIV/AIDS in a number of different ways. In one question they were asked to rate their position on a five-point scale in relation to the question, "How do you rate your chances of developing AIDS?". A second question asked "When you compare yourself to other gay men what would you say are your chances of developing AIDS?" In each case, a higher score denoted a higher perception of risk. A number of questions were asked about knowledge of AIDS, but the subjects were so well informed that there was very little variability in the responses. In this report answers to the following question are used to assess