No increase in sexual risk taking by gay men in England, 2001 to 2008

This article is more than 9 years old. Click here for more recent articles on this topic

Analysis of Gay Men’s Sex Survey data shows that, overall, the proportion of men reporting unprotected anal intercourse (UAI) did not change between the 2001 and 2008 surveys. There was an increase in the number of men with diagnosed HIV reporting unprotected sex, but numerous other indicators showed no increase in risk taking.

Moreover, there wasn’t evidence of more men practicing ‘serosorting’ or ‘strategic positioning’. The analysis is reported in the March issue of Sexuality Research and Social Policy (full text freely available).

The results contrast with those of a mathematical modelling study, widely reported by the media in February, which suggested that new HIV infections have increased in gay men since the mid-1990s, despite the greater use of HIV treatment – because of a modest rise in the number of men having unprotected sex. However, it is important to note that this rise in UAI was an estimation derived from the modelling, rather than empirical data.

Gay Men's Sex Survey

Sigma Research conducted this large, cross-sectional survey annually between 1997 and 2008.


unprotected anal intercourse (UAI)

In relation to sex, a term previously used to describe sex without condoms. However, we now know that protection from HIV can be achieved by taking PrEP or the HIV-positive partner having an undetectable viral load, without condoms being required. The term has fallen out of favour due to its ambiguity.

mathematical models

A range of complex mathematical techniques which aim to simulate a sequence of likely future events, in order to estimate the impact of a health intervention or the spread of an infection.

odds ratio (OR)

Comparing one group with another, expresses differences in the odds of something happening. An odds ratio above 1 means something is more likely to happen in the group of interest; an odds ratio below 1 means it is less likely to happen. Similar to ‘relative risk’. 


Studies aim to give information that will be applicable to a large group of people (e.g. adults with diagnosed HIV in the UK). Because it is impractical to conduct a study with such a large group, only a sub-group (a sample) takes part in a study. This isn’t a problem as long as the characteristics of the sample are similar to those of the wider group (e.g. in terms of age, gender, CD4 count and years since diagnosis).

statistical significance

Statistical tests are used to judge whether the results of a study could be due to chance and would not be confirmed if the study was repeated. If result is probably not due to chance, the results are ‘statistically significant’. 

Their new analysis compares responses from 2001 and 2008, two years in which the same set of detailed questions about sexual behaviour were asked. Most questions were about whether an individual had engaged in a sexual practice at least once during the previous year – the data cannot, therefore, give an insight into the frequency of risk taking.

Moreover, as the same men did not complete the survey in each year, changes over time must be interpreted with some caution. To minimise bias, only those men recruited online were included in this analysis (face-to-face recruitment methods differed from year to year).

Nonetheless, there were some differences between the samples of 3517 and 1382 men recruited in 2001 and 2008, respectively.

In 2008, the sample was considerably older and better educated, somewhat more ethnically diverse, and had a slightly different geographical spread from the first survey. While the researchers did adjust for these factors in their statistical analyses, it is possible that there were other confounding factors that they did not identify.

As has been previously reported from this survey, the number of men taking HIV tests has increased over time. The number who had ever taken a test went up from 45.8 to 69.3%. As a consequence, the number of men with diagnosed HIV went up from 3.4 to 9.1%.

After statistical adjustment, men in 2008 were around half as likely to report that they had never tested for HIV (odds ratio 0.45, 95% confidence interval 0.39-0.51).

Sexual behaviour

Looking at the whole samples, there were no statistically significant changes in the number of sexual partners reported by participants. However, among men who had either tested negative or never tested, there were statistically significant falls in the number of men who had had five or more sexual partners in the past year. For HIV-positive men, there wasn’t a change.

In terms of unprotected anal intercourse (UAI), this was reported by 59% of men in 2001 and 57.5% of men in 2008 – i.e. no change.

The one exception was for men with diagnosed HIV. In 2001, 73.7% reported unprotected sex in the past twelve months, rising to 82.1% in 2008. After statistical adjustment, in the later survey men were twice as likely to report unprotected sex (odds ratio 2.13, 95% confidence interval 1.09-4.17).

The survey gathered detailed information on sexual role, partner’s HIV status and use of condoms, in order to identify a further eight anal intercourse behaviours. For all of these, there were no statistically significant changes from 2001 to 2008. (The only exception was a small increase in the number of men reporting any anal intercourse.)

When looking only at men who had tested HIV negative or tested HIV positive, there weren’t any significant changes in these specific behaviours either.

The authors note that if more men were ‘serosorting’, they would expect to see more men reporting unprotected sex with men of the same HIV status as them. Similarly, if more men were adopting ‘strategic positioning’, they would expect to see more HIV-positive men taking the receptive role with partners who were not HIV positive. However there was no evidence for either phenomenon in the data.

The authors say that their data are suggestive of “a growing concentration of risk among men with diagnosed HIV”, who appear to be more likely to report having unprotected anal intercourse and multiple partners.

They also note the possibility that uptake of HIV testing has increased disproportionately in men at greatest risk of HIV and men most likely to report unprotected sex. In other words, with such substantial changes in HIV-testing behaviours in recent years, the nature and characteristics of the population of gay men with HIV may have also changed.

Whatever the explanation, this confirms “the ongoing need for prevention programmes to prioritise men with diagnosed HIV”.

Comparison with other studies

The authors point to two other analyses of UK gay men over a similar time frame.

Surveys of gay men recruited in Glasgow and Edinburgh bars observed an increase towards the end of the 1990s in the number of men with multiple unprotected anal intercourse partners, but found no evidence for further increases between 2002 and 2008.

Surveys of men recruited in London gyms similarly reported an increase in the number of men reporting UAI with men of an unknown or different HIV status between 1998 and 2001, followed by a decrease and a levelling off between 2005 and 2008. In terms of men reporting any unprotected anal intercourse in the previous year, the data did suggest an increase between 1998 and 2008, but this did not quite reach statistical significance.

Finally, it is worth noting the recent mathematical modelling study. This suggested that, although antiretroviral therapy has significantly reduced HIV transmission, the annual number of new HIV infections has nonetheless risen since the late 1990s. The modelling suggested that this could most logically be explained by a modest rise in unprotected sex between 1998 and 2010.

The key source of sexual behaviour data for the model were the National Survey of Sexual Attitudes and Lifestyles (NATSAL) findings from the year 2000. In that survey, 35% of men who have sex with men reported UAI with a man of an unknown or different HIV status.

The model suggested that this proportion could have risen from 35 to 44% by 2010 (a 26% rise). But the latter figure is an output of the model, rather than empirical data that was fed into it.

NATSAL has recently been repeated, with data collected in 2010 to 2012. When its results are published – probably in the next year or two – they are likely to be of considerable interest, due to NATSAL recruiting a large, representative sample of the adult UK population.


Hickson F et al. HIV Testing and HIV Serostatus-Specific Sexual Risk Behaviour Among Men Who Have Sex with Men Living in England and Recruited Through the Internet in 2001 and 2008. Sexuality Research and Social Policy 10: 15-23. (Full text available here)