HTLV infection among young injection and non-injection heroin users in Spain: prevalence and correlates

Although some studies have described the epidemiology of infection with HIV or hepatitis B and C in young users in Spain - one of the European countries with the highest prevalences - there are no studies of the prevalence of HTLV infection and the most important associated factors.

To evaluate the prevalence and main determinants of HTLV-1 and HTLV-2 infection in young heroin users (including both injection (IDUs) and non-injection drug users (NIDUs)) recruited outside health care services in three of Spain's principal cities.

Cross-sectional cohort study. All participants (981) were street-recruited by chain referral procedures between April 2001 and December 2003. Face-to-face interviews were conducted using a structured questionnaire and dried blood spot samples were collected for serological testing.

No sample was positive for HTLV-1 and 27 samples were positive for HTLV-2; all of these were found only in Spanish IDUs in the cities of Madrid (17, 6.2%) and Barcelona (10, 3.5%). The only two factors significantly associated with HTLV infection in the logistic regression analysis were HIV infection (OR 5.7; 95% CI 2.2-14.8) and having injected in the last 30 days (OR 6.5; 95% CI 1.4-29.8). Having been in prison (OR 2.4; 95% CI 0.9-6.4) and HCV infection (OR 3.8; 95% CI 0.5-30.7), which were strongly and significantly associated in the bivariate analysis, were no longer significant in the logistic analysis. Almost the same variables were selected in the tree analysis, in which subjects could be classified into three groups: high prevalence (28.5%, HIV+ and HBV+ who had injected in the last 30 days), medium prevalence (17.8%) and low (<3%) or zero prevalence (HIV-, HCV- and HBV-).

HTLV-1 was not detected among young Spanish heroin users. HTLV-2 was not found in NIDUs (perhaps due to the low rate of sexual transmission); it was found only in IDUs from Madrid and Barcelona, but not in those from Seville. Its prevalence is very low and the main correlates of infection were HIV infection and injection as the usual route of heroin administration.