The study protocol was reviewed and approved by the ethical boards of all participating centers. The study was conducted during January 2004–December 2007 in clinical centers that offered primary healthcare to illegal migrants in northern (Brescia), central (Rome), and southern (Palermo) Italy. All adult migrants from a non–European Union country who registered for a visit at each center were asked to participate in the study. Participants were divided into 4 groups according to their risk for HIV infection: 1) commercial sex workers, 2) persons reporting unsafe sex (occasional not-for-money homosexual or heterosexual contacts with persons other than their regular partner), 3) persons with other risk factors (intravenous drug use or blood transfusion in their country of origin, and 4) persons with no risk factors identified.

Participants were tested for antibodies against HIV types 1 and 2 and for p24 antigen by using a commercial microparticle enzyme immunoassay (AxSYM HIV Ag/Ab Combo; Abbott Laboratories, Abbott Park, IL, USA). HIV-positive serum samples were further tested for HIV antibody avidity by using the AxSYM HIV 1/2gO assay (Abbott Laboratories) as reported (9), to ascertain likely time of infection. An avidity index <0.80 indicated that infection was acquired recently (within the past 6 months); a higher index indicated that infection was acquired earlier (>6 months ago). A cutoff value of 0.80 was used and validated as having 93.0% sensitivity, 98.5% specificity (10), and >90.0% reproducibility. Avidity results were cross-checked with reported time of migration to assess likely place of exposure.

A total of 4,078 persons were invited to participate in the study. Of 3,976 (97.5%) who agreed to participate, 3,003 (73.6%) underwent HIV testing (2,815 in Brescia, 48 in Rome, and 140 in Palermo). In terms of HIV risk for 2,853 respondents, 191 were commercial sex workers, 1,246 practiced unsafe sex, 47 reported other risks, and 1,494 reported no risk factors (total = 2,978 because multiple risk factors were reported by some persons). Demographic characteristics for the participants are shown in Table 1.

HIV-1 infection was detected for 29 (0.97%) of 3,003 participants (95% confidence interval [CI] 0.90%–1.2%); no participants were infected with HIV-2. Avidity index results were obtained for 27 of the 29 HIV-positive participants. Univariate analysis showed that sociodemographic and behavior factors associated with HIV infection were Christian religion (p = 0.029, odds ratio [OR] 3.07, 95% CI 1.06–8.83), migration from sub-Saharan Africa (p = 0.001, OR 11.94, 95% CI 1.61–88.81), commercial sex (p = 0.0001, OR 18.2, 95% CI 6.25–52.97), and unsafe sex (p = 0.016, OR 3.43, 95% CI 1.22–9.66). Multiple logistic regression showed that factors independently associated with increased risk for HIV infection were migration from sub-Saharan Africa (p = 0.0001, OR 3.7, 95% CI 2.2–9.4), commercial sex (p = 0.025, OR 18.4, 95% CI 4.9–48.5), and unsafe sex (p = 0.01, OR 2.3, 95% CI 1.7–8.6).

Place of infection could not be determined for 17 (63.0%) of 27 persons; 6 (22.2%) of 27 were presumably recently infected in Italy, and 4 (14.8%) of 27 presumably acquired their infection in their country of origin before emigration. Results of avidity index determinations are shown in the Figure and Table 2. Sociodemographic characteristics of our population did not differ from those reported nationwide (1).

Our data confirm that many illegal migrants practice unsafe sex (low rate of condom use). These findings are worrisome if one considers poor knowledge of HIV transmission reported in our study population (11). Consequently, migrants are particularly vulnerable to STIs, as shown by the high prevalence rate (0.97%) for the adult migrant population, which is higher than the estimated 0.4% prevalence rate for the national population in Italy (12). The higher HIV prevalence rate for persons from an area (sub-Saharan Africa) in which HIV is highly endemic might reflect exposure in the country of origin or new infections in the host country. A total of 6 (22.2%) of the 27 HIV infections for which avidity index data were available were probably acquired in Italy by migrants from sub-Saharan Africa (n = 3), eastern Europe (n = 2), and Latin America (n = 1). Conversely, all 4 (14.8%) persons who acquired infection before migration to Italy were originally from sub-Saharan Africa.

Our study had 2 limitations. First, recruitment was not evenly balanced between centers. However, migrants are unevenly distributed in Italy, with the largest communities in northern Italy. Second, the study acceptance rate was only 73.6%. However, this acceptability rate is similar to rates in other studies in Europe (13,14). Furthermore, sociodemographic characteristics of our sample were homogeneous with those of other studies from the same centers and nationwide (1,15), and we did not anticipate any differences between participants and those who did not participate.

Our data show high HIV prevalence for an illegal migrant population in Italy. For persons originally from sub-Saharan Africa and, as for the native population in Italy, practicing commercial or unsafe sex were independently associated with HIV infection. A large proportion of persons had presumably acquired infections after migration. These data indicate the prominent role of social determinants of HIV infection, including marital status and living conditions. Health education and free access to HIV testing and care for the illegal migrant population in western countries is needed, particularly for persons from sub-Saharan Africa.