The UK: a unique epidemic?

The UK remains a low-prevalence country for HIV, with about one in 500 of the adult population infected compared with (for example) one in about 160 in the USA. However the growth in HIV diagnoses in recent years has been uniquely steep. HIV incidence doubled in the four years between 1998 and 2002, and incidence in heterosexuals trebled during the same period. For comparison, it was thought newsworthy when it was announced recently that HIV incidence in Japan had doubled in the 11 years between 1991 and 2002.

The growth in the UK epidemic has been fuelled largely, though not entirely, by HIV diagnoses among immigrants from sub-Saharan Africa. Two-thirds of HIV-positive test results were among heterosexuals in 2004, and this is likely to underestimate prevalence in heterosexuals as fewer come forward for testing than gay men. But 68% of these (42% of all cases) were acquired in Africa, compared with only 11% (7% of all cases) within the UK. This compares with around 1600 of the 1850 gay male infections (86%) being caught in the UK.

Furthermore, of the 11% of home grown heterosexual infections,  in about 75% of cases the infecting partner acquired their infection outside the EU. The proportion of heterosexual HIV infections in the UK where both partners caught HIV within the UK stood at 70 to 100 cases at the end of 2004 no more than 2.5% of heterosexual infections, and 1.5% of the total.

Nearly half 47% - of all new HIV cases in the UK in 2004 were among people of Black African ethnicity (to use the HPAs term), which would include some African gay men and Africans who acquired HIV within the UK.

The transformation of the UKs epidemic can be seen in the following two graphs. The first includes all risk groups (with heterosexuals divided into men and women) and the total of infections. The second compares the three major risk groups (heterosexuals, gay men and Africans). Africans, as pointed out above, would include Africans from all risk groups, but it can be seen how the growth in African infections echoes the heterosexual curve.

(Source: HPA) To the end of December 2002, 3,917 injecting drug users have been reported to be HIV-positive in the UK; once again over two-thirds are men (2,702 men, 1,215 women). This represents around 7.5% of total reported HIV infections in the UK.

The majority (40%) of all HIV diagnoses among injecting drug users have been in London, and the overwhelming majority of the remainder have been diagnosed in Scotland (32%). Injecting drug users have declined as a proportion of newly-diagnosed HIVpositive individuals in Scotland from 77% prior to 1986 to 3% in 2002.

Injecting drug users make up 37% of all diagnosed HIVpositive individuals in Scotland, compared with 6% across the rest of the UK.

What we see above is not one HIV epidemic, but two: one a long-standing epidemic concentrated almost entirely among gay men and largely acquired in the UK; the other a fringe of the African HIV epidemic, consisting of HIV cases caught in the most high-prevalence countries of the world but diagnosed here.

Is this a unique situation? There are historical reasons why the UK should be particularly affected by HIV in Africa: it is estimated that 54% of the worlds people with HIV live in African countries within the British Commonwealth. But other countries have colonial ties with Africa too.

In Belgium HIV incidence went up by 42% between 1997 and 2002 and it is estimated that 60% of cases in 2002 were among people from sub-Saharan Africa. In Switzerland incidence went up by 25% between 2001 and 2003, and of the 42% of diagnoses that were among heterosexuals, 65% were in people from sub-Saharan Africa. Other western European countries have seen similar increases.

However it is only Ireland and the Scandinavian countries that have seen heterosexual cases from Africa come to dominate in the same way they have done in the UK. Ireland saw HIV incidence jump 33% in once year between 2002 and 2003; 63% of cases in 2002 were among heterosexuals and of those 77% were acquired in Africa, figures very similar to the UK. And in Sweden it is estimated that two-thirds of cases in the last two years have come from people infected in the global epidemic (to quote the UNAIDS fact sheet on the country), with only 21% of cases among gay Swedes. However the Scandinavian countries continue to have very low prevalence about half that of the UK.

The UKs sudden increase in HIV incidence has led to a doubling of prevalence since 1998 (see chart).

(Source: HPA, with extrapolations by author)

This chart is made up by simply subtracting death notices from HIV diagnoses, and this will tend to overstate prevalence because of duplicate records and people leaving the country.

The HPA figures of prevalence, based on the UAPMP surveys, tend to come out with lower figures than this. They estimated that at the end of 2003, there were 53 000 people living with HIV in the UK, of whom, 14 300 (27%) were unaware of their infection. Since then, up to June 2005, there have been another 7526 new diagnoses and 643 deaths of people with HIV recorded. Unless a lot more previously undiagnosed people have been coming forward for testing than hitherto, this would indicate a UK prevalence at June 2005 of around 59,000. This matches the graph above, but would indicate that the prevalence calculated by the graph was about 25% too high, as it only includes diagnosed cases.

However unlinked anonymous surveys, because they only sample certain populations, have an inbuilt uncertainty in the results they come out with. UNAIDS, who have at times issued controversial prevalence figures based on sample surveys, gave 51,000 as the estimated figure of people living with HIV here at the end of 2003, almost agreeing with the HPA. But they gave upper and lower confidence limits of 25,000 (which is 10,000 lower than the number recorded by SOPHID as seeking care) and 81,000.

The HPAs estimate for the number of undiagnosed people is based in part on an estimate of the number of sexually active gay men culled from the 2000 National Survey of Sexual Attitudes and Lifestyles (NATSAL). This found that 2.8% of UK men had had sexual experience with another man in the last five years.

However NATSALs figures are usually regarded as conservative because people dont tend to tell the whole truth to sex researchers and it may also underestimate the HIV incidence and prevalence among gay men who frequent the commercial gay scene but do not attend GUM clinics (see below).

Whatever the exact figures, it is clear that the upward curve in prevalence was not foreseen. In 2000, the CDSC (forerunner to the HPA) issued its own figures for prevalence up to 2000 and the estimated increase up to 2005, and this forecast makes an interesting contrast with what actually happened:

(Source: CDSC 2000, from table in previous edition of AIDS Reference Manual)

The way that heterosexual infections were about to outstrip gay ones was completely unforeseen, and shows how, even in the developed world, political and economic turmoil can suddenly transform the shape of epidemics. In 2003, for instance, the biggest single nationality among newly-diagnosed people (larger than UK nationals) was Zimbabweans, many seeking asylum from the Mugabe regime.

This near-tripling of prevalence since highly active antiretroviral treatment (HAART) became available has in turn had a knock-on effect on HIV services and created considerable political turmoil and policy changes, such as the withdrawal, from April 2004, of entitlement to NHS HIV treatment to anyone who could show they had been in the UK more than 12 months. These changes are documented elsewhere in this Manual, under African communities in the UK.

This also led to a situation where, in 2003, gay men became for the first time less than half the total of those living with HIV in the UK they were 51% of the total in 2002 and 45% in 2003.

The overlap gay men and ethnicity

Of course, we are not seeing two completely separate infections. Gay men exist in Africa and the rest of the developing world too. They may both be a source of HIV infections and be vulnerable to infection by UK gay men.

The HPA has hitherto only collected the likely country of infection data for heterosexuals, though it plans from next year to collect these figures for gay men too. In gay men, country of possible infection, country of birth, and STD clinic attendance have only been collected through clinician reporting, and so are not available for diagnoses.

What we do know about the country of infection and ethnicity of gay men is presented here:

• The UK was the country of probable infection for 83% of newly-diagnosed gay men, with 7% from the rest of Europe, and between 3% and 1% from other continents.
• Seventy-two per cent of newly diagnosed gay and bisexual men were born in the UK, 14% in other parts of Europe, and between 5 and 2% born in other continents. We can see from this that when it comes to gay men the UK is more likely to be the place where gay men catch HIV than where they are born: in contrast to the heterosexual pattern, it is a ‘net exporter’ of HIV.
• The majority of men (87%) were white, 6% ‘other’ or mixed, 3% black Caribbean, 2% black African, 1% black ‘other’, and 1% Asian.

Prevalent diagnosed HIV infections

Information on prevalent diagnosed HIV infections is provided by the survey of prevalent diagnosed HIV infections (SOPHID) in England, Wales, and Northern Ireland, and CD4 monitoring in Scotland.

In 2003, the latest year with full figures available, 35,428 individuals were receiving care of which 23,298 (64%) were receiving antiretroviral therapy.

The proportion of newly diagnosed infections attributed to heterosexual contact has increased to 65% of the total reported in 2004, and heterosexuals accounted for the majority of the prevalent diagnosed caseload for the first time in 2003 (see chart).

(Source: HPA)

The decline of AIDS

There has been a marked fall in the number of new AIDS diagnoses amongst people who have been aware of their HIV status for some time, suggesting that anti–HIV therapy is having an impact on disease progression. An increasing proportion of new AIDS diagnoses is occurring in people who had never been tested or had only been diagnosed with HIV less than three months before the onset of an AIDS-related illness. For them, their HIV diagnosis probably comes too late for anti-HIV therapy to delay the onset of AIDS. (Source: HPA)

In the UK, AIDS diagnoses have fallen from a high of 1851 in 1994 to 777 cases diagnosed so far in 2004. In the period 1994-2001 new AIDS diagnoses fell in all regions of the UK. Deaths due to AIDS have declined from 1531 in 1994 to 214 so far in 2004, while death from other causes has increased slightly, from 169 in 1994 to 242 in 2004, This was the first year that more deaths in people diagnosed with AIDS were due to non-AIDS-defining illnesses than AIDS itself.

Although there is no evidence that Africans in the UK have a shorter life expectancy once they have been diagnosed with AIDS (compared with other groups of people with AIDS), the rate of new AIDS diagnoses amongst African people has hardly fallen at all. Every year Africans diagnosed with AIDS form a larger proportion of UK AIDS cases, and whilst new AIDS diagnoses fell 35% amongst gay men in 1997, they fell by only 6% amongst Africans.

London loses some of its dominance

Since the beginning of the HIV epidemic London has been the epicentre of HIV in the UK, with 56% of all diagnoses made in the Greater London area and another 9% in the south-east. However, though London and the south-east remain dominant, they are much but less so than previously. In 2004 41% of HIV diagnoses were made in London and an additional 12% in the south-east. See chart below.

The proportion of all new HIV diagnoses that were made in London has declined steadily: from 64% in 1990 and 60% in 2000 to 53% in 2002 and 41% in 2004. In the same time period the proportion of all new HIV diagnoses made in the south-east has risen from 7.5% to 12% and in the Eastern region the proportion has increased from 2% to 8%.

The North West region has the third largest number of HIV diagnoses in the UK after London and the south-east, accounting for 8.5% of all new diagnoses in 2004.