In Europe, men who have sex with men (MSM) were affected first and most severely by HIV, when AIDS incidence peaked in the early 1980s.1 2 Substantial changes in the sexual behaviour of MSM that followed resulted in a decline in the incidence of sexually transmitted infections (STI) and reported cases of HIV infection. These trends have been reversed since the late 1990s, with increases in high-risk sexual behaviour,3 incidence of STI and newly diagnosed cases of HIV.4 There is also further potential for HIV transmission with the increased survival of HIV-positive MSM on highly active antiretroviral therapy (HAART).5 Evidence of widespread sexual mixing among MSM in western Europe was provided by the reports of STI outbreaks linked between different countries.6^–9

The aim of this paper is to present HIV surveillance data for MSM in the countries of the European Union (EU) and the European Free Trade Association (EFTA) for the most recent period, from 2000 to 2006, in order to inform the development of a Europe-wide HIV prevention policy for MSM.

METHODS

HIV/AIDS surveillance data for the period 2000–6 have been collected by the European Centre for the Epidemiological Monitoring of HIV/AIDS (EuroHIV) from national correspondents of the 27 member states of the EU and three EFTA countries (Iceland, Norway and Switzerland). Three sources of HIV and AIDS surveillance data on MSM have been analysed.

HIV reporting data

Anonymous individual or aggregate data on newly diagnosed cases of HIV infection have been reported annually by national correspondents. Data by transmission group were available for 27 of the 30 EU/EFTA countries. National HIV reporting data for 2006 were not available from Spain and Italy; Estonia has not consistently reported transmission groups. A further four countries were excluded from trend analysis because HIV reporting systems were established or modified within the study period: France (in 2003), Malta (in 2004), The Netherlands (in 2002) and Portugal (in 2001). Trends analyses for the period 2000–6 were thus performed on data from 23 countries (table 1).

Anonymous individual data for newly diagnosed cases of HIV have been reported by 18 of the 23 countries with consistent HIV reporting systems during the study period and include information on age and clinical stage at the time of HIV diagnosis (countries excluded were Austria, Bulgaria, Germany, Ireland and Romania). Late HIV diagnosis was defined as reported AIDS clinical stage at the time of HIV diagnosis.

AIDS reporting data

Anonymous individual data on AIDS cases have been reported annually by all 30 countries using a standardised case definition.10 Anonymous individual data on year of death among reported AIDS cases were provided by all countries. Adjustments for reporting delays for AIDS and deaths among AIDS cases were made according to country of report.11

HIV prevalence data

Aggregate data on HIV prevalence in various MSM populations were reported from national correspondents and compiled in the European HIV Prevalence Database. The results from 20 HIV seroprevalence studies conducted among MSM during 2000–6 are presented according to the site of recruitment: gay community settings (12 studies)3 12^–22 or healthcare services such as STI clinics (eight studies).

HIV status was based either on testing of serum or saliva samples (seroprevalence) or was self-reported. The testing of serum or saliva samples was performed either by unlinked anonymous testing in which specimens are collected and tested in an unlinked and anonymous manner, or by diagnostic testing in which the results of all testing carried out with the primary objective of providing individuals with the result were systematically reported. Diagnostic HIV testing may have been offered by the clinician as part of routine testing or in the context of clinical care or may have been initiated by the individuals themselves.

Statistical analysis

Statistical analyses were undertaken using SPSS statistical software (release 10.0). Comparisons of proportions were performed using χ² tests. Overall tests for trend for all countries for reported newly diagnosed HIV infection rates, AIDS incidence and mortality during the period 2000–6 were computed using linear regression giving p values based on F distribution.

RESULTS

HIV reporting data

In 2006, the 27 EU and EFTA countries with available national data reported a total of 7693 newly diagnosed cases of HIV infection among MSM (56.7 per million men aged 15–64 years), which constituted 27% of all reported HIV cases (table 1). The highest rate was reported in the UK (2597 cases; 130.1/million) and the lowest in Romania (eight cases; 1.1/million). The rate of newly diagnosed cases of HIV infection among MSM was higher than 100 per million men aged 15–64 years in three countries (The Netherlands, Luxembourg and UK), between 75 and 100 per million in two countries (Portugal and Switzerland), between 50 and 74 per million in eight countries (Belgium, Denmark, France, Germany, Greece, Ireland, Norway, Malta) and lower than 50 per million in all others. MSM represented the majority of newly diagnosed cases in four countries: Czech Republic (56%), Germany (52%), The Netherlands (55%) and Slovenia (74%).

In the 23 countries with consistent data for the period 2000–6, there was an 86% increase in the number of reported cases of newly diagnosed HIV infections among MSM between 2000 (3003 cases; 28.8/million) and 2006 (5571 cases; 52.7/million) (p = 0.001; table 1). Increases for the period 2000–6 were reported by all countries except Cyprus, Iceland, Lithuania and Luxembourg. In countries reporting more than 20 new diagnoses of HIV in 2006, more than doubling of cases reported since 2000 was observed in Finland, Germany, Hungary, Norway and Slovenia.

In the 18 countries with consistent HIV surveillance systems that provided individual HIV data, the median age of cases reported in 2006 was 36 years (range 16–85; N  =  5378), similar to data reported in 2000 (median age 36 years, range 16–78; N  =  2795). Six countries reported late HIV diagnosis of more than 20% in 2000 (Denmark, Hungary, Lithuania, Slovenia, Switzerland and UK) and only one did so in 2006 (Denmark). The proportion of cases with late HIV diagnosis decreased from 25% in 2000 (291/1177) to 10% in 2006 (225/2172) (χ²  =  85.7; p<0.001). All but one country (Denmark) with available data reported a decrease in the proportion of late diagnoses since 2000 (data not shown).

AIDS reporting data

In all 30 countries, AIDS incidence among MSM declined by 40% from 2422 (14.5/million men aged 15–64 years in 2000) to 1445 (8.5/million in 2006; p = 0.005) (table 2). The incidence of more than 10 AIDS cases per million men was reported from 12 countries in 2000 and seven countries in 2006. Among countries reporting more than 10 AIDS cases in 2006, the highest AIDS incidence rates were reported in Portugal (18.7/million), Spain (16.6/million) and The Netherlands (16.2/million). Between 2000 and 2006, the AIDS incidence rate decreased in 21 countries and increased in eight.

Mortality among reported AIDS cases decreased by 57%, from 876 (5.3/million in 2000) to 373 (2.2/million in 2006; p<0.001) (table 3). In 2000, nine countries reported mortality rates greater than 5/million and in 2006 only five.

HIV prevalence studies

HIV prevalence among MSM found in 12 studies undertaken in 10 countries in gay community settings between 2000 and 2006 ranged from a high of 18% in Barcelona, Spain, in 2002 to 0% in Lithuania in 2004 (table 4). In western European countries, HIV prevalence was 10% or greater, except in Ireland (5%). In contrast, in two central European countries (Czech Republic and Slovenia), prevalence was less than 5%. In one study, in France in 2004, national data were analysed by region of residence and the self-reported prevalence of HIV was higher among MSM living in Paris (17%) than in the rest of France (12%).

Of the five studies undertaken in 2000–6 among MSM diagnosed with an STI, the prevalence of HIV ranged from 68% in 2001 among MSM diagnosed with syphilis in Paris to 7.6% in 2000 among MSM diagnosed with one of 12 acute STI in clinics in the UK. In Germany, Ireland and Italy levels ranged from 19% to 49%.

Three studies reported HIV prevalence among MSM attending STI clinics (but who may not have been diagnosed with an STI) in the period 2000–6. National (The Netherlands and UK) or regional (Spain) levels ranged from 3 to 5% in the Netherlands, 5 to 6% in Spain and 8 to 19% in UK. HIV prevalence levels were usually higher in large cities compared with the rest of the country (The Netherlands, UK).

DISCUSSION

The number of newly diagnosed cases of HIV reported among MSM has recently increased throughout EU and EFTA countries. These increases have occurred not only in those countries with large and visible gay communities,23^–25 but also in many eastern and central European countries where, although actual numbers remain relatively low, rapid increases have been reported. In those countries where decreases have been reported, monitoring trends is less reliable due to small numbers (eg, Iceland, Luxembourg). A substential proportion of HIV transmission in Europe is thought to occur through sex between men and MSM remain the group at greatest risk.26

An increase of reported HIV diagnoses among MSM may not necessarily reflect a real increase in incidence as newly diagnosed cases of HIV infection represent a mix of incident and prevalent cases. In the UK, the increase in reported HIV diagnoses among MSM has been attributed to an increase in the uptake of HIV testing,27 which is partly corroborated by our observation that, since 2000, more MSM are being diagnosed at an earlier stage of disease, although this should be interpreted with caution as there have been improvements in the completeness of information. Increases in the incidence of HIV among MSM have been reported in different European cities,28^–30 with statistically significant increases in two of these studies.29 30 Incidence in these MSM populations, estimated at 3% or greater annually, is high and underlines the importance of prioritising HIV prevention among MSM. Due to decreased mortality, morbidity and prolonged life expectancy attributed to HAART,31 an increasing pool of HIV-positive individuals may also affect the increase in incidence.

Throughout Europe there are wide differences in the prevalence of HIV among MSM in community settings, with a higher prevalence reported in those countries with larger and more visible communities than in countries with smaller or less visible communities. In addition to prevalence data, some studies assessed risk behaviour, which remains high.12^–1820^–22 Therefore, even if a low HIV prevalence has been reported in some communities of MSM in Europe, the high levels of risk behaviour in these communities combined with increasing sexual mobility leaves open the potential for the rapid spread of HIV among MSM in countries with a lower prevalence. This highlights the need for a Europe-wide strategy for HIV prevention among MSM.

In western Europe, studies performed in community settings collected self-reported HIV status information, the validity being lower than in studies in which the HIV status was ascertained by testing of biological samples, especially as it has been demonstrated that a substantial proportion of MSM are unaware of their true HIV status.3 Nonetheless, self-reported HIV status information is more easily obtained in the field as the difficulties of collecting biological samples in a community setting are circumvented. The added value of combining community-based behavioural surveys with the anonymous collection of saliva samples has been demonstrated.3 When feasible, such studies can contribute to a better understanding of HIV epidemiology, to the evaluation of harm-reduction strategies (especially those of negotiated safety with a steady partner and serosorting) and to the development of better targeted and more appropriate interventions among MSM.

We have shown that the increase in new HIV diagnoses in most countries was paralleled by the decrease in late diagnosis, suggesting an increased proportion of MSM being tested at an earlier stage of HIV infection. Individuals unaware of their HIV-positive status contribute disproportionately to the transmission of HIV,32 and testing has been increasingly promoted to reduce the proportion of individuals unaware of their status who present late at healthcare facilities3 to undergo testing and receive counselling to prevent secondary transmission.33 The promotion of HIV testing is the cornerstone of the proposed European Centre for Disease Prevention and Control (ECDC) policy for HIV prevention in Europe and these efforts should be supported by appropriate surveillance data to monitor the uptake of HIV testing among MSM.34

There have been decreases in AIDS incidence and mortality among MSM with AIDS in many European countries, which can be attributed to relatively good access to HAART introduced widely after 1996.5 Trends in mortality should be interpreted with caution as we have reported only mortality among individuals previously diagnosed with AIDS. The increasing numbers of HIV-positive MSM reported in Europe as well as the large proportion of MSM diagnosed with an STI who were also HIV positive highlight the need for effective sexual health promotion in this group to prevent secondary transmission. The population of MSM with an acute STI represents those at greatest risk of HIV infection because these men have recently engaged in high-risk sexual behaviour and the presence of an acute STI can facilitate the transmission of HIV.35 In the USA, increased levels of serosorting have been reported among MSM and this may have contributed to the stabilisation of HIV incidence despite increases in diagnoses of acute STI and high-risk sexual behaviour.36 In Europe, an increase in serosorting has only been reported in London among HIV-positive, but not HIV-negative, MSM.37 HIV prevention campaigns have often targeted younger MSM and this should be continued, but increasing HIV incidence among older MSM reported in some countries38 indicates the need for HIV prevention messages to be appropriately targeted.

We have reported a recent increase in the number of HIV diagnoses among MSM in nearly all EU and EFTA countries, and in some countries this probably represents a true increase in incidence. This, combined with the high prevalence of HIV reported in many gay community settings, the high prevalence of HIV among MSM diagnosed with STI and the high sexual mobility of this population, highlight the need for a Europe-wide HIV prevention strategy. Such a strategy should facilitate the exchange of information on best and innovative practices to promote safer sex and HIV testing among MSM between various national public health agencies, health promotion specialists and community organisations. Surveillance data providing robust information on newly diagnosed HIV cases, HIV incidence and prevalence, high-risk behaviours and the uptake of HIV testing are vital in the development of evidence-based national and international HIV prevention strategies for MSM as well as monitoring the interventions.