Meta-analysis on risky sexual behaviour of men: Consistent findings from different parts of the world YifruBerhan*andAsresBerhan College of Medicine and Health Sciences, Hawassa University, Hawassa, Ethiopia (Received 6 January 2012; final version received 26 April 2012) The aim of this analysis was to determine the consistency of higher-risk sex practice among educated and/or wealthymenindifferentpartsoftheworld.Meta-analysiswasdoneonriskysexualbehaviourofmenusingthe recentDemographicandHealthSurveys(DHS20032009)datafrom26countriesinandoutsideAfrica.DHS data were accessed through electronic databases. In this analysis, since there was significant heterogeneity (I 2 50%) among surveys findings, random effects analytic model was applied. Mantel-Haenszel statistical methodwasusedtocalculatethepooledoddsratiosacrosscountries.Outof79,736menaged1549yearswho hadsexualintercoursein12monthsprecedingtherespectivesurvey,35.7%reportedtohavehigher-risksex.The proportionofhigher-risksexwasfoundpositivelycorrelatedwithincreasedwealthindex.In24countries,higher- risksexwasfoundtohavehighlystatisticallysignificantassociationwithmenlivinginurbanareas,educatedto secondaryandabove,andownedmiddletohighestwealthindex.Theoverallcondomuseduringthelasthigher- risksexualencounterwas47%butcondomusewasbetterpracticedbyeducatedmen.Nearlyintwo-thirdsof countries reported HIV-prevalence, the proportion of HIV infection was highest among better educated. In conclusion, this meta-analysis has shown that risk taking sexual behaviour is invariably associated with high educational attainment, urban residence and better wealth index regardless of geographic location of men participatedinthesurveys. Keywords: education;higher-risksex;residence;wealth;condom;HIV-prevalence Introduction HIV infection is still one of the world’s challenges because of its complex nature from biomedical and behaviouralperspective.Globally,thenumberofnew HIV infections has been reported as falling but the proportion of HIV-prevalence in the age range of 1549 years showed no change between 2001 and 2009 (UNAIDS, 2010). Risky sexual behaviour is known to predispose for acquiring and transmitting HIVinfection(Sebastian&Hofmann,2006). However, previous research findings were incon- sistentregardingwhichgroupofpeoplewereatrisk for HIV infection and propagation. Some studies showedthatHIVinfectionisassociatedwithpoverty (Fenton, 2004; Madise, Zulu, & Ciera, 2007; Zulu, Dodoo & Ezeh, 2002). Other studies questioned whether being poor or wealthy is at the root of HIV.Evensomeauthorsarguedthathighprevalence of risky sexual behaviours among the wealthier is likelytoincreasetheirvulnerabilitytoHIVinfection (Butler, 2000; Shelton, Cassell, & Adetunji, 2005). Studies that analysed one or two countries’ demo- graphic and health surveys (DHS) concluded that there was strong association between wealth and sexual risk taking behaviour, which was ascribed as a driving factor for HIV epidemics (Awusabo-Asari & Annim, 2008; Eugene, Charles, Robinson, Philip, &Luc,2006). Definitely, the prevalence and impact of HIV infection in the poorest nations in particular Sub- Saharan African countries was reported to be the highest: out of the total HIV infected population globally, about three-fifths were from Sub-Saharan Africancountries(UNAIDS,2006).Thequestionis, evenincountrieswithlowresources,whichgroupof people are at risk of HIV infection. Some evidences hadchallengedthepreviousassumptionsthatthereis strong linkage between poverty and HIV infection (O’Farell,2001;White,Cleland,&Carael,2000). An overview of 15 quantitative studies (20042007) examining the socio-economic status and risk of HIV infection concluded that poor individualsandhouseholdsarelikelytobehitharder bythedownstreamimpactofAIDS,buttheirchances of being exposed to HIV in the first place are not necessarily greater than wealthier individuals. In short,‘‘AIDScan’taccuratelybetermedasadisease of poverty’’ (Gillespie, Kadiyala, & Greener, 2007). AuthorswhoanalysedeightDHSfromSub-Saharan Africa concluded that HIV-prevalence does not *Correspondingauthor.Email:yifrub@yahoo.com AIDS Care Vol.25,No.2,February2013,151159 ISSN0954-0121print/ISSN1360-0451online # 2013Taylor&Francis http://dx.doi.org/10.1080/09540121.2012.689812 http://www.tandfonline.com