Article Text


Epidemiology oral session 9: Use of observational data and mathematical modelling for impact evaluation
O1-S09.05 Decline in HIV prevalence among young people in the general population of Cotonou, Benin, 1998–2008
  1. L Behanzin1,
  2. A Buve2,
  3. C M Lowndes3,
  4. D M Zannou4,
  5. I Minani5,
  6. S Anagonou6,
  7. M C Boily7,
  8. A C Labbe8,
  9. R Bitera9,
  10. M Alary1
  1. 1Centre hospitalier affilié universitaire de Québec, Quebec, Canada
  2. 2Institute of Tropical Medicine, Antwerp, Belgium
  3. 3Health Protection Agency, London, UK
  4. 4Faculté des Sciences de la Santé de Cotonou, Université d'Abomey-Calavi, Cotonou, Bénin Benin
  5. 5Dispensaire IST, Cotonou, Bénin
  6. 6Faculté des Sciences de la Santé, Université d'Abomey-Calavi, Cotonou, Bénin
  7. 7Imperial College, London, UK
  8. 8Hôpital Maisonneuve-Rosemont, Montréal, Canada
  9. 9Centre hospitalier affilié universitaire de Québec, Quebec, Canada


Objective Comparative study of the prevalence of HIV and sexually transmitted infections (STI), and associated behaviours in the general population of Cotonou between 1998 and 2008.

Methods In Cotonou, two studies employing similar methods were carried out in 1998 and 2008 respectively. In these studies, the census areas (clusters) were sampled with probability proportional to size. After enumeration of all households in the selected clusters, a certain number of households were randomly sampled from each selected census area (950 in 1998, 1070 in 2008). Consenting adults, aged 15–49 years (but 15–64 years for the men in 2008) were interviewed and screened for HIV, syphilis, and HSV-2 (serologic detection of antibodies for the latter infections), Neisseria gonorrhoeae and Chlamydia trachomatis (nucleic acid amplification assays on uro-genital samples). The Roa-Scott χ2 was used to consider the cluster effect in the univariate comparison of proportions. Logistic regression (taking into account the cluster effect) was used for multivariate analysis, adjusting for socio-demographic variables.

Results The global HIV prevalence was stable (3.4% in 1998 vs 3.1% in 2008). There was however a trend towards decreasing among men (Abstract O1-S09.05 table 1). The decrease was highly significant among men aged less than 30 (3.0% in 1998 vs 0.5% in 2008, p<0.0001). A trend towards decreasing prevalence was also observed among women aged less than 20 (2.4% in 1998 vs 0.5% in 2008, p=0.102). On the other hand, an upward trend was observed among women aged 20+ (3.8 in 1998 vs 4.8% in 2008, p=0.346). Syphilis prevalence also decreased significantly, but this decline was more pronounced among women (Abstract O1-S09.05 table 1). The prevalence of gonorrhoea trended lower among men while prevalence of HSV-2 increased among both men and women (Abstract O1-S09.05 table 1). The proportion of adults who reported condom use during their last extramarital sexual intercourse increased (23.0% in 1998 vs 40.1% in 2008, p<0.0001).

Abstract O1-S09.05 Table 1

Multivariate comparison of HIV/STI prevalence between 1998 and 2008 among men and women of the general population of Cotonou aged 15–49

Discussion The decrease in HIV prevalence among young people could be explained by the increase in condom use and may also be related to the impact of intensive interventions targeting the prostitution milieu during the same period. The upward trend among older women could be related to a large increase in access to antiretroviral therapy that occurred from 2004 onwards.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.