Behavioural disinhibition

One great concern in relation to circumcision – or indeed of any novel HIV-prevention method – is that it might simply cause people (in this case, men) to abandon previous risk-management behaviour such as partner reduction or condom use.

There were some signs of behavioural disinhibition amongst circumcised men in two of the three RCTs.

In the Orange Farm study, circumcised men reported more sexual contacts in the months preceding study visits than uncircumcised men.

The average number of sexual partners in the previous three months was 5.9 versus 5.0 (p= <0.001) in circumcised versus uncircumcised men at visits in the first year of the study. During the second year of the study, the number of partners reported in the previous nine months was 7.5 in circumcised men and 6.4 in uncircumcised (p=0.0015). However, adjusting for these potential confounders did not alter the overall efficacy observed, so it’s possible these increases, while statistically significant, were not so in terms of transmission risk.

In the Rakai study, in contrast, there were no significant differences in risk behaviour between arms except that the uncircumcised men were more likely to use alcohol during sex (57 versus 51%, p=0.02). Of note, the mean age in the Rakai trial was older. This may possibly indicate that younger men in these settings need more intensive support to maintain safer sex.

In the Kisumu study, at 24 months statistically significant differences were observed between circumcised and uncircumcised men reporting any unprotected intercourse in the last six months (51% circumcised versus 46% uncircumcised, p=0.0349) and reporting consistent condom use (36 versus 41%, p = 0.0326). The researchers say that ‘notably greater’ numbers of circumcised men reported other risky behaviours, but that differences were non-significant.

However, a later substudy of the Kisumu RCT¹ found that there were virtually no differences in behaviour or in STI infections between circumcised and uncircumcised men when more detailed questions were asked regarding risky sexual behaviour. The new analysis controlled for whether men had sex with casual or regular partners or sex workers, whether they or their partner were in a concurrent relationship, whether they perceived their partner to have HIV, and several other variables.

The primary finding of the study was that there were no statistically significant differences between the proportion of circumcised and uncircumcised men who engaged in any of the 18 risk behaviours they were asked about. Circumcised men were also no more likely than uncircumcised to believe that their operation reduced the risk of HIV infection at any time point. There was no relationship between believing that circumcision reduced the risk of acquiring HIV and any sexual-risk factor.

Circumcised men were more likely to be diagnosed with a new STI at the six-month time point (6 versus 3%), but this difference disappeared at 12 months (2 versus 3%). The authors speculate that this might be due to reinfection from partners, who were not treated at baseline.

The 18 risk variables were composited into a risk score on a scale from zero (no sex at all since the last visit) to nine (highest risk on all variables). This risk score fell substantially for both circumcised and uncircumcised men during the study, from 3.55 in uncircumcised men and 3.25 in circumcised men at baseline to 2.5 for both groups at twelve months. In practical terms this meant that at the twelve-month follow up men were 59% less likely to report not having any sex, and of those who did, there was a 16% decrease in risk scores. These were statistically significant declines.

STI incidence fell for both groups during the study, and the 45% decline between months six and twelve was statistically significant. Men with a prevalent STI at baseline were three times more likely to catch another one during the study (which may confirm the reinfection theory).

This substudy included about half the total number of men enrolled in the RCT, and its findings are probably representative of behavioural change amongst other participants.

The authors comment that: “these are important results in the face of reluctance on the part of some in the international community to endorse male circumcision…[they] provide evidence that risk compensation is likely to be minimal or absent among circumcised men and, therefore, it should not…be considered a barrier to the promotion of male circumcision for HIV and STI prevention.”

However, they also point out that participation in the RCT involved repeated HIV testing and individually tailored risk-reduction counselling.

They comment: “Conditions under which [male circumcision] is provided widely are likely to be different”, adding that, “it will be necessary to further evaluate the possibility that men increase their HIV risk behaviour after circumcision is offered in more naturalistic public health and medical settings”.

A separate study in the Kisumu region² was conducted between 2002 and 2004 and involved 324 men undergoing elective (i.e., non-randomised) circumcision who were then demographically matched with equal numbers of men remaining uncircumcised. Follow up took place throughout the year following circumcision.

The investigators found that men who chose to be circumcised were significantly more likely than men who chose to remain uncircumcised to have had unprotected sex with a non-regular partner in the three months before circumcision (p = 0.03).

In the month following circumcision, men undergoing the procedure were, as one would expect, 87% less likely than uncircumcised men to report unprotected risky sex.

However, in the year following circumcision, there ceased to be any difference in the amount of risky and unprotected risky sex reported by circumcised and uncircumcised men.

There was a linear decrease across visits throughout the study in the proportion of men in the control group reporting two or more partners in the previous six months, whereas the proportion reporting the same behaviour in the circumcision group fell from month zero to month six, but did not fall thereafter. This was statistically significant (p=~0.03).

However, the investigators stress, “at no point during this year was there any appreciable reported excess of risky sex or unprotected risky sex among circumcised men.”