University of California, San Francisco Logo

University of California, San Francisco | About UCSF | Search UCSF | UCSF Medical Center

Risk compensation is not associated with male circumcision in Kisumu, Kenya: A multi-faceted assessment of men enrolled in a randomized controlled trial
Global Health Sciences Literature Digest
Published May 11, 2009
Journal Article

Mattson CL, Campbell RT, Bailey RC, et al. Risk compensation is not associated with male circumcision in Kisumu, Kenya: A multi-faceted assessment of men enrolled in a randomized controlled trial. PLoS ONE 2008;3(6):e2443.doi:10.1371/journal.pone.0002443.

In Context

Three randomized controlled trials have demonstrated the efficacy of male circumcision (MC) to decrease the risk of acquiring HIV infection.(1, 2, 3) There are continuing concerns about risk compensation (i.e., increased risk behavior among circumcised men because of the known lower risk associated with circumcision) and its potential to mitigate the protective effect of circumcision. In the clinical trials, evidence of increased sexual risk among circumcised men compared to non-circumcised men was not definitive.(1, 2, 3) Uncertainty regarding risk compensation may be due to insufficiently detailed information about risk among study participants or to difficulty in analyzing multiple, highly correlated variables.


To measure the effect of circumcision for HIV prevention on risk compensation using a detailed propensity scale measuring sexual risk behavior


Kisumu District, Kenya

Study Design and Participants

Cohort of a sample of participants from a randomized controlled trial (RCT) of MC


HIV risk and incident Neisseria gonorrhoeae, Chlamydia trachomatis, and Trichomonas vaginalis infections


The RCT-enrolled men who were sexually active in the past 12 months, HIV uninfected, uncircumcised, aged 18-24 years, and resident in Kisumu District. All men enrolled in the RCT between March 2004 and September 2005 were systematically recruited into this study. All men received HIV counseling and testing at one, three, six, and 12 months and were informed that although some evidence showed decreased HIV risk associated with circumcision, the evidence was not conclusive. Participants consented to have interviews at six and 12 months and access to data from the RCT.

The Timeline Follow-back approach was adapted and used to obtain sexual histories regarding every sexual relationship in the previous six months for up to 12 partners at baseline and at six- and 12-month follow-up. Information was collected for each partner separately. Concurrency was defined as having two or more partners with whom the start and stop dates of the relationship overlapped. Men were asked if they believed that circumcision reduced the risk of HIV acquisition.

An 18-item propensity scale was developed to capture specific risk behaviors and practices reported for each partner. Count variables were created to summarize behaviors across 0-12 partnerships. The validity of the score was determined by measuring the association between scores on the risk scale and incident sexually transmitted diseases (STDs).


Of the 1,780 men enrolled in the RCT between March 2004 and September 2005, 1,319 (74%) participated in this sub-study, 10 of whom had missing data and were excluded from analysis. Follow-up was 76% at six months and 77% at 12 months. Loss to follow-up was the same between intervention and control groups. Risk behaviors between circumcised and uncircumcised men did not differ. Both groups reduced risk behaviors at follow-up. At baseline, 57% of circumcised and 56% of uncircumcised men reported believing that circumcision reduces HIV risk; by 12 months, 75% of circumcised and 76% of uncircumcised reported this belief.

In both groups, STDs declined between six and 12 months, and the incidence did not differ between the groups at 12 months. At baseline, circumcised men were more likely to have an STD than uncircumcised men (OR 1.6, P=0.02) and a new STD at six months (OR 1.8, P=0.05).

Median risk scores declined for men in both groups and there was no difference between the risk scores in the two groups. Specifically, there was no statistical interaction between risk scores and time for the circumcised men.

The decline in risk scores at 6 and 12 months persisted after adjustment for age, marital status, education, income, and belief that circumcision reduces risk of HIV.


In the context of an RCT, circumcision does not appear to increase sexual risk or incident HIV.

Quality Rating

This is a high quality study with clear ascertainment of exposure and outcomes. Groups were comparable. Follow-up was adequate. Sample size was sufficient.

Programmatic Implications

This study has important implications considering that MC has been identified as having a strong protective effect against HIV acquisition. A concern with MC as an HIV prevention intervention has been the possible risk of increased risk behavior due to a belief that circumcision would reduce HIV risk, thereby allowing the circumcised man to engage in high risk behaviors. Such risk compensation, however, was not observed and risk-reduction counseling appears to have resulted in declines in risk behavior among both groups. The findings from this study are consistent with those from the RCT but used more detailed measures of risk. Other studies also have failed to find increases in risk among circumcised men.(2, 3, 4) It is important to include MC as one of several HIV prevention methods. In addition, risk-reduction counseling should be provided along with MC to enhance decreases in risk behaviors.


  1. Auvert B, Taljaar D, Lagarde E, et al. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 trial. PLoS Med 2005;2:e298.
  2. Bailey R, Moses S, Parker CB, et al. Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. Lancet 2007;369:643-56.
  3. Gray RH, Kigozi G, Serwadda D, et al. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet 2007;369:657-66.
  4. Agot KE, Kiarie JN, Nguyen HQ, et al. Male circumcision in Siaya and Bondo Districts, Kenya: Prospective cohort study to assess behavioral disinhibition following circumcision. JAIDS 2007;44:66-70.