Secondary Logo

Journal Logo

Recent trends in the HIV epidemic among injecting drug users in Northern Italy, 1993–1999

Sabbatini, Alessia; Carulli, Biancamaria; Villa, Marco; Corrêa Leite, Maria Léa; Nicolosi, Alfredo*for the NISDA Study


Objective To monitor and describe the time trends of the HIV epidemic among intravenous drug users (IDU) attending drug dependence treatment centres (DDTC) in Northern Italy.

Design A cohort of all seronegative IDU attending DDTC in Lombardy between 1993 and 1999; all had been tested for HIV at least twice. Periodic sample interview surveys were done to assess risk behaviours.

Methods The incidence rates of HIV infection were calculated using the person-year (PY) method and expressed as the number of cases per 1000 PY at risk. Background HIV prevalence was calculated by dividing the number of positive cases by the total number of IDU tested at all DDTC in Lombardy.

Results Between 1993 and 1999, 135 seroconversions occurred in 7945 subjects followed for 19 671 PY, yielding an incidence rate of 6.9/1000 PY. Ninety seroconversions occurred among the 6563 males and 45 seroconversions among 1382 females (incidence rates 5.5 and 14.0, respectively). Among the males, the incidence of HIV was 4.5 in those aged less than 25 years and 5.8 in those aged 25 years or more. Among the females, the corresponding figures were 21.1 and 10.3. HIV prevalence decreased over time, and it was higher among females. Sexual behaviours at risk were more common among females.

Conclusions The incidence of HIV infection among IDU in Northern Italy was stable between 1993 and 1999. The higher incidence and prevalence among females and the different prevalence of risk behaviours between genders suggest an increasing role of heterosexual transmission.

From the the Department of Epidemiology and Medical Informatics, Institute of Biomedical Technologies, National Research Council of Italy and the aG.H. Sergievsky Centre, School of Public Health, Columbia University, New York, USA. *See Appendix.

Requests for reprints to Dr A. Nicolosi, Department of Epidemiology and Medical Informatics, Institute of Biomedical Technologies, National Research Council of Italy, Via Fratelli Cervi 93 20090 Segrate (MI), Italy.

Received: 23 March 2001;

revised: 27 June 2001; accepted: 11 July 2001.

Sponsorship: This work was partially funded by a grant from the Istituto Superiore di Sanità, Progetto AIDS.

Back to Top | Article Outline


It is widely held that the best way of studying the epidemic of HIV-1 is to monitor the incidence rates of the infection; the inherent difficulties of studying the incidence of HIV have led most investigators and public health authorities to use reported AIDS cases as a surrogate for HIV incidence. However, the recent abrupt decrease in the number of such cases, attributable to the introduction of new treatment regimens, makes their use obsolete for epidemiological purposes. In this context, the incidence of HIV infection acquires a greater importance to monitor the HIV epidemic.

From 1987, HIV-negative injecting drug users (IDU) attending drug treatment centres in Northern Italy were recruited to the Northern Italian Seronegative Drug Addicts (NISDA) study in order to monitor time trends in the incidence of HIV infection and study the risk factors and their prevalence in this population over time [1–3]. We here provide an update of the incidence rates of HIV infection between 1993 and 1999.

Back to Top | Article Outline


NISDA is based on the observation of a dynamic cohort of HIV-negative IDU [4]. The IDU were recruited from drug treatment centres located in the cities of Lombardy and its bordering regions, in which the incidence of HIV and AIDS is the highest in Italy.

All of the IDU were tested for HIV antibodies at entry, and those who were seronegative were enrolled on the date of their first negative test and then followed until they acquired HIV infection or abandoned the treatment centre. The follow-up was based on the review of the medical records kept by the treatment centres, where all of the visits made by every IDU were recorded, together with demographic, social, drug use, clinical and laboratory data.

In addition, periodical sample surveys of the prevalence of behaviours at risk of HIV infection were carried out using detailed questionnaire-based interviews. The questions concerning borrowing of used syringes and sexual behaviours referred to the 6 months preceding the interview. Individuals were defined as sexually active during that period of time if they reported at least one sexual intercourse, either with a stable (duration of relationship longer than 6 months) or occasional partner.

The incidence rate of HIV infection was estimated as the number of seroconversions divided by person-years (PY) of follow-up. Seroconversion was assumed to have taken place on the day midway between the date of the most recent visit and that of the last visit at which the subject had been negative for anti-HIV antibodies. The yearly incidence rates were expressed as moving averages in order to compensate for the fluctuations caused by the small number of cases. The confidence intervals (CI) of the incidence rates were calculated on the basis of the Poisson distribution. The incidence rates and their CI were used to estimate the relative risk of HIV infection by gender and age.

The HIV prevalence data relate to all of the IDU attending drug treatment centres in Lombardy, and was obtained from the Regional Health Authority, which collects such data routinely. Prevalence was calculated by dividing the number of HIV-positive cases by the number of individuals (positive and negative) tested at the treatment centres each year, by gender. Prevalence was separately calculated for ‘old’ (IDU already on treatment before the year in which the HIV test was performed) and ‘new’ clients (IDU who started treatment during the same year in which the HIV test was performed).

Back to Top | Article Outline


Between 1993 and 1999, 8303 IDU were enrolled and 7921 (95.1%: 6543 males and 1378 females) were followed for a mean 2.5 years. There were no statistically significant differences in follow-up duration by gender or age at entry. The median age at enrolment was 26.3 years for males and 25.2 years for females (48.6% of the females and 39.9% of the males were younger than 25 years of age).

During this period, there were 135 seroconversions (90 in males and 45 in females) among IDU followed for 19 611.5 PY (16 395.3 in males and 3216.2 in females), thus yielding an overall incidence of 6.9/1000 PY (95% CI, 5.8–8.2). The incidence rate for the entire period was 5.5 in males (95% CI, 4.4–6.8) and 14.0 in females (95% CI, 10.3–18.7). The rates did not show any statistically significant time trend between 1993 and 1999 in either males or females (Fig. 1).

Fig. 1.

Fig. 1.

When analysed by age group, the incidence rate was 4.5/1000 PY (95% CI, 2.7–7.1) among young male IDU (< 25 years of age); and 5.8 (95% CI, 4.5–7.3) among the older IDU (> 25 years of age); this difference was not statistically significant. The incidence rate among younger females was 21.1/1000 PY (95% CI, 13.4–31.7), double that observed in the older group (10.3/1000 PY; 95% CI, 6.5–15.7). The relative risk of HIV infection among younger female IDU in comparison with their older counterparts was 2.04 (95% CI, 1.14–3.66).

The relative risk of HIV infection between female and male IDU was 4.70 (95% CI, 2.54–8.71) in the younger group and 1.78 (95% CI, 1.10–2.87) in the older group.

The prevalence of HIV among the IDU attending drug treatment centres in Lombardy between 1993 and 2000 decreased, among the ‘old’ clients, from 41 to 29% in males and from 42 to 36% in females; among the ‘new’ clients, it decreased from 18 to 12% in males and from 21 to 17% in females (Fig. 2).

Fig. 2.

Fig. 2.

Between 1993 and 1999, the proportion of individuals borrowing used syringes decreased from 27 to 10% among male IDU, and from 33 to 20% among female IDU (Table 1). The results of the interviews revealed that, among the individuals who reported syringe sharing, 85.7% of the women and 20.7% of the men only did so with their sexual partners. Over the same time period, an increasing proportion of men and women reported condom use during occasional sexual intercourse, and the frequency of condom use during such intercourses also increased. However, the proportion of IDU using condoms during sexual intercourse with their partners was lower than in occasional sexual intercourse, and its frequency somewhat decreased over time. The data in Table 1 show that sexual exposure to HIV was potentially greater among females than males. The proportion of women with a sexual partner who was known to be HIV positive or an IDU was more than three times higher than that of men. Although not very common, another factor of potential sexual exposure (i.e., prostitution) was essentially a female phenomenon. Finally, over the entire study period, 70.4% of younger female IDU versus 54.7% of male IDU (< 25 years of age) were sexually active (P < 0.001); the corresponding figures among the older IDU were 79.3 and 63.0% (P < 0.001).

Table 1

Table 1

Back to Top | Article Outline


Our study reveals a general stabilization in the incidence rates of HIV infection among IDU attending treatment centres in Northern Italy between 1993 and 1999, at a level that is considerably lower than the high rates observed in the late 1980s and early 1990s [2]. This trend is consistent with that observed in other Western European and North American IDU populations, in which incidence rates have recently been reported to be declining [5,6]. The incidence rates of HIV infection in our population is comparable with that recently reported in New York City (7/1000 PY between 1992 and 1997) [7], San Francisco (less than 1/100 PY between 1995 and 1998) [8] and Australia (1.7/1000 PY between 1992 and 1995) [9]. An incidence rate of 3.0/100 PY between 1985 and 1996 has been reported in Amsterdam, but with a strongly declining trend [10]. In our study, the observation that the incidence has been stable and low over the last few years suggests that, as confirmed by the results of the behavioural surveys, Northern Italian IDU have began to learn to use safer injection practices than in the past.

There was a general decline in HIV prevalence over time. Prevalence is a function of incidence, disease duration and the speed at which individuals leave the studied population [11]. The higher prevalence among the ‘old’ clients of treatment centres reflects the large number of IDU who became infected during earlier years that were characterized by higher incidence rates [12,13]. The lower prevalence among the ‘new’ clients reflects the lower incidence experienced during more recent years. The higher prevalence among female IDU is consistent with the observed incidence rates, which were also higher in female than in male IDU.

There are various reasons underlying the decrease in the incidence rates among male IDU (especially the young males, who presented the highest incidence rates in the past), including the fact that the habit of sharing syringes with other drug users is more common among men than women, who tend to limit this practice to within the couple [14]. The overall decrease in the frequency of needle sharing means that a larger number of men than women have reduced or eliminated the main mechanism of acquiring HIV infection. There are also explanations for the higher incidence among female IUDs. The women were potentially highly exposed to HIV via the sexual route because they were more sexually active and much more likely to have HIV-positive or IDU partners, practise prostitution or trade sex for drugs. Since women are more susceptible to contracting HIV infection via sexual contacts than men [15], the sexual route of transmission is more likely to increase incidence rates among women [16]. A shift in the predominant mode of transmission among IDU from the parenteral to the heterosexual route has been observed in a number of studies [17–19].

Although positive in regard to the decline and stabilization of the incidence rates of HIV infection, the results of this study should not be considered as favouring any relaxation in prevention efforts. The incidence of HIV may have decreased, but its prevalence will not decrease at the same pace because the longer survival induced by modern treatments will keep a higher proportion of circulating infected individuals than in previous years. Furthermore, prevention efforts regarding safe sexual practices need to be strengthened, particularly to combat the growing increase in the incidence of HIV among female IDU.

Back to Top | Article Outline


1. Nicolosi A, Lazzarin A for the NISDA Study. HIV seroconversion rates in intravenous drug abusers from Northern Italy. Lancet 1989, ii: 269.269.
2. Nicolosi A, Corrêa Leite ML, Musicco M, Molinari S, Saracco A, Lazzarin A. Incidence and prevalence trends of HIV infection in intravenous drug users attending treatment centres in Milan and Northern Italy, 1986-1990. J Acquir Immune Defic Syndr 1992, 5: 365–373.
3. Nicolosi A, Corrêa Leite ML, Musicco M, Molinari S, Lazzarin A. Parenteral and sexual transmission of human immunodeficiency virus in intravenous drug users: a study of seroconversion. Am J Epidemiol 1992, 135: 225–233.
4. Nicolosi A, Corrêa Leite ML, Osella A, Lazzarin A. The NISDA study: methods and results of a longitudinal study of HIV seroconversion. In:HIV Epidemiology: Models and Methods. Edited by Nicolosi A. New York: Raven Press; 1994: 13–28.
5. Hamers FF, Batter V, Downs AM, Alix J, Brunet JB. The HIV epidemic associated with drug use in Europe: geographic and trends. AIDS 1997, 11: 1365–1374.
6. Holmberg S. The estimated prevalence and incidence of HIV in 96 large US metropolitan areas. Am J Public Health 1996, 86: 642–654.
7. Des Jarlais DC, Marmor M, Friedmann P. et al. HIV incidence among injection drug users in New York City, 1992–1997: evidence for a declining epidemic. Am J Public Health 2000, 90: 352–359.
8. McFarland W, Kellogg TA, Louie B, Murrill C, Katz MH. Low estimates of HIV seroconversions among clients of a drug treatment clinic in San Francisco, 1995 to 1998. J Acquir Immune Defic Syndr 2000, 23: 426–429.
9. van Beek I, Dwyer R, Dore GJ, Luo K, Kaldor JM. Infection with HIV and hepatitis C virus among injecting drug users in a prevention setting: retrospective cohort study. Br Med J 1998, 317: 433–437.
10. Langendam MW, van Brussel GH, Coutinho RA, van Ameijden EJ. Methadone maintenance treatment modalities in relation to incidence of HIV: results of the Amsterdam cohort study. AIDS 1999, 13: 1711-1716.
11. Freeman J, Hutchison GB. Incidence, prevalence, and duration. Am J Epidemiol 1980, 112: 707–723.
12. Arcà M, Perucci CA, Spadea T. The epidemic dynamics of HIV-1 in Italy. Stat Med 1992, 11: 1657–1684.
13. Downs AM, Heisterkamp SH, Brunet JB, Hamers F. Reconstruction and prediction of the HIV/AIDS epidemic among adults in the European Union and in the low prevalence countries of central and eastern Europe. AIDS 1997, 11: 649–662.
14. MacDonald M, Wodak AD, Ali R. et al. HIV prevalence and risk behaviour in needle exchange attenders: a national study.The Collaboration of Australian Needle Exchanges. Med J Aust 1997, 166: 237–240.
15. Nicolosi A, Corrêa Leite ML, Musicco M, Arici M, Gavazzeni G, Lazzarin A. The efficiency of male-to-female and female-to-male sexual transmission of the human immunodeficiency virus: a study of 730 stable couples. Epidemiology 1994, 5: 570–575.
16. Solomon L, Astemborski J, Warren D. et al. Differences in risk factors for human immunodeficiency virus type 1 seroconversion among male and female intravenous drug users. Am J Epidemiol 1993, 137: 892–898.
17. Leads From the Morbidity Weekly Report. AIDS associated with injecting-drug use – United States, 1995. JAMA 1996, 275: 1628–1630.
18. Ronald Peter JM, Robertson JR, Wyld R, Weightman R. Heterosexual transmission of HIV in injecting drug users. Br Med J 1993, 307: 1184–1185.
19. Estebanez PE, Russell NK, Aguilar MD, Beland F, Zunzunegui MV. Women, drugs and HIV/AIDS: results of a multicentre European study. Int J Epidemiol 2000, 29: 734–743.
Back to Top | Article Outline


Participants in the Northern Italian Seronegative Drug Addicts (NISDA) study: Emanuela Arrigoni, Federica Albini Riccioli, Emanuela Bernasconi, Elena Bertolini, Eugenio Brianza, Livia Brignolo, Carmela Bruni, Fiorella Cadoria, Maria Caminada, Anna Maria Camisani, Guglielmo Campione, Antonina Cardia, Bianca Carmela Bruni, Fiorella Cadoria, Maria Caminada, Anna Maria Camisani, Guglielmo Campione, Antonina Cardia, Bianca Maria Carulli, Alberto Chiesa, Antonio Colaianni, Paola Colleoni, Maria Léa Corrêa Leite, Edoardo Cozzolino, Giovanni De Micco, Piera Dettori, Enrico Donadeo, Enrico Elba, Maria Grazia Fasoli, Maurizio Fea, Claudio Filippi, Sergio Fonzi, Roberto Gaggini, Roberto Giuseppe, Abbasali Heydari, Bruno Iacopino, Alfio Lucchini, Roberta Mangili, Maria Merlo, Ronaldo Moioli, Roberto Mollica, Alfredo Nicolosi, Giuseppe Pedrazzani, Giuseppe Pennisi, Piero Perotti, Cinzia Priora, Serenella Quaresima, Fabio Reina, Franco Riboldi, Maria Richina, Roberto Giuseppe, Alessia Sabbatini, Elisabetta Secchi, Vincenzo Stefano, Laura Tidone, Biagio Tinghino, Claudio Tosetto, Dario Valsecchi, Concettina Varango, Marco Villa, Rossano Vitali, Lucia Zazio.


HIV; injecting drug users; heterotransmission

© 2001 Lippincott Williams & Wilkins, Inc.