Condoms for HIV prevention do not lead to earlier sex, more sex, or more partners, meta-analysis concludes

Could increased condom use actually reduce sexual partners and frequency of sex?

The researchers found that interventions which provided condoms to study participants had no effect on the number of sexual partners, the total number of times participants had sex, and on whether participants who were previously abstinent became sexually active (p > 0.20 for all three). This suggests, they write, "that teaching individuals about condoms does not result in more acts of intercourse, more sexual partners, or earlier intercourse."

In fact, when they specifically examined interventions that had improved condom use, they found that these reduced numbers of sexual partners (r=0.22; p=0.035) and the number of times participants had sex (r=0.49; p=0.003). No link was found between interventions that improved condom use and sexual activity status.

"This finding," write the researchers, "should provide reassurance that increased numbers of sexual occasions, larger number of partners, and more likely sexual activity are not iatrogenic [unintentional] effects of providing condoms or training in condom use skills and interpersonal negotiation skills."

HIV prevention interventions that include information on condoms do not inadvertently encourage an earlier sexual debut, more frequent sexual activity or more sexual partners, according to a meta-analysis of 174 studies published in the March edition of the Journal of Acquired Immune Deficiency Syndromes. The study also concludes that including information on condoms alongside motivational and behavioural components is more likely to have a positive effect on reducing frequency of sexual activity than interventions which do not include all three components.

The influence on the Bush administration of Christian and right-wing organisations – who argue that condom-based prevention programmes encourage earlier and more frequent sexual activity – has led in recent years to a rapid increase in abstinence-only-based HIV prevention.

Nevertheless, condom use is clearly an important factor in HIV risk reduction. However, transmission risk can also be affected by what the authors of this meta-analysis term "sexual frequency indicators": number of sexual partners, total number of times participants have sex, and sexual activity status (i.e. sexually active or abstinent).

This meta-analysis, undertaken by researchers at the University of Connecticut and Syracuse University, with a grant from the United States (US) National Institutes of Health (NIH), is the first to address both the question of whether condom-based intervention affect the overall frequency of sexual behaviour and what makes frequency-reduction interventions most effective.

The meta-analysis covers 174 studies, investigating 206 separate interventions, published between January 1989 and May 2003. The interventions began with a total of 149,660 participants, of which 78% remained in the study, on average, leaving 116,735 participants (an equal amount of whom were men and women, and 54% of whom were black) for analysis. The majority (84%) of the studies were conducted in the US, with most (70%) conducted in medium to large cities.

Most of the interventions (74%) provided HIV counselling and testing, and 97% provided HIV education. Condom information was provided in 45%, 41% taught condom use skills, and 74% distributed condoms to participants. Interpersonal skills training, e.g. negotiation skills (66%), was included in the majority of studies, as was intrapersonal skills training, e.g. self-management to avoid risky sex (58%). The interventions averaged eight participants per session, and met for three session of 75 minutes each, for a total of approximately four hours.

What works best at reducing sexual frequency and partners?

The researchers found that interventions which included motivational components were more successful at reducing the number of times people in the studies had sex. "This finding suggests," they write, "that risk reduction interventions work, at least in part, by engaging intrinsic motivation for self-protection."

They also found that skills training was an important feature of interventions that were successful in reducing numbers of sexual partners. "This suggests", write the researchers, "that negotiating a safe sexual relationship with a partner requires skill. Individuals who lack the ability to skillfully decline a partner's advances may be at greater risk for unintended sexual advances, including unprotected sex."

Condoms alongside information, motivation and skills training may work best

The researchers say that overall they found that interventions that included information, motivation and skills training worked better than interventions that only included some of these components.

They conclude by arguing that "although condom use has been touted as the primary and most beneficial sexual risk reduction outcome", an approach which combines condom use with "key elements of behavioral theory" may be most successful at reducing sexual frequency and number of partners.