Anne Johnson and NATSAL: the study of sexual lifestyles and attitudes in Britain
"Human sexual behaviour represents one of the least explored segments of biology, psychology and sociology. Scientifically more has been known about the sexual behaviour of some of the farm and laboratory animals." So wrote Alfred Kinsey in 1948. A 'Nature' article in 1990 - on the first large-scale, cross-sectional study of human sexual behaviour and attitudes in Great Britain - quoted these words, saying they were as true four decades on as they had been when originally written. It is easy to forget how many taboos there were concerning sex in the relatively recent past.
Today, however, Kinsey's proclamation no longer rings true. Our understanding of the subject has increased dramatically, with huge public health benefits in areas such as sex education, teenage pregnancy strategy, HIV prevention and combating other sexually transmitted infections. Various shifts and trends lie behind these changes, and science is always a team effort; however, historians are nonetheless likely to record that Professor Anne Johnson has done more than anyone else to bring about this step-change in human sexual self-knowledge.
Although Johnson studied medicine and trained as a GP, she knew even then that she wanted to do public health research. "From very early on in my medical career, I was very interested in the extent to which public health changes had affected human health. The 19th-century changes such as the provision of water, housing and improved sanitation had huge impacts." A chance meeting at a management training course led to a position as a lecturer and researcher in the epidemiology of sexually transmitted infections at the then Middlesex Hospital in central London in 1985. At this time, the Hospital was treating some of the UK's first cases of AIDS and carrying out early work to determine the prevalence of HIV antibodies among gay men in the capital. As the numbers of cases rose rapidly, fears grew. The Royal College of Nursing predicted the number of infections would reach one million by 1991. There were increasing demands for accurate information and projections; however, with little data available on the numbers of sexually active gay men, how many partners people had and the dynamics of transmission, researchers were operating in the dark.
Johnson had her 'eureka' moment while drinking a cup of tea one afternoon with colleagues in the coffee room of James Pringle House at the Middlesex Hospital, in 1986. "At this stage nobody had any idea how big the HIV epidemic was going to be," she says. "Projections were made, but largely based on extrapolating curves. People were very sceptical because there was an order of magnitude of difference in the predictions." At the time, there were plenty of studies on the sexual behaviours of specific groups such as gay men and sex workers, but studying what the population at large were getting up to in their bedrooms was seen as less acceptable in many quarters. Johnson adds: "I remember saying one day 'we really ought to do a general population-based study on sexual behaviour with a proper random sample'. People thought that was a bit crazy. I remember a senior colleague, a professor who was studying behaviour in gay men, looking at me with an expression that suggested I'd cracked. We forget that discussion of sex was much more hidden back then. People really didn't talk about these things at dinner parties."
Johnson and a group of interested colleagues designed a pilot study, with funding from millionaire property developer Godfrey Bradman. This was carried out by Gallup in 1987 and published in the journal 'AIDS' a year later. Although their response rate was on the low side at 48 per cent, they had at least demonstrated that a randomised study of sexual lifestyles could be carried out in the general population.
Meanwhile, another group, led by Kaye Wellings of the London School of Hygiene and Tropical Medicine, was working on a similar idea, but from more of a qualitative and health promotion perspective. The groups came together and thrashed out a study design and grant proposal for a pilot study involving 1000 participants. Funding was granted by the Economic and Social Research Council (ESRC), and the pilot was carried out in late 1988. With a successful preliminary study in the bag, the support of the ESRC and Sir Donald Acheson, the UK Chief Medical Officer, and the desperate need for the data in the fight against AIDS, the researchers were optimistic that public funding for their main study involving 20 000 people would be forthcoming. However, six months of increasingly ominous silence followed, a mystery that was solved in September 1989 by a front-page story in the 'Sunday Times' under the headline "Thatcher halts survey on sex".
According to the article, Margaret Thatcher, then Prime Minister, saw the study as an invasion of people's privacy and did not want her government to be associated with it. She was also reported to have had doubts over whether the exercise would produce meaningful and useful results. "It was her view that people wouldn't answer the questions properly," says Johnson. "That was particularly infuriating because we'd done such a lot of work to show that people would answer them accurately."
The precise sequence of events remains unclear, but it seems that Sir Donald, who died last year, decided the study was too important to be stymied by political considerations, and so contacted Peter Williams, the then Director of the Wellcome Trust. Things moved very quickly. Johnson sent the original ESRC grant application to Williams on 2 October 1989. Just nine days later, the Trust's scientific committee recommended to the Trustees that they provide the full £900 000 cost of the National Survey of Sexual Attitudes and Lifestyles (NATSAL). Later that same day, the Trustees agreed unanimously to do so. Some of those involved in making the decision were concerned about the implications of appearing to thumb their noses to the Prime Minister. However, on the other side of the scales was a stronger feeling that politicians should not have a veto on a well-designed study that many saw as key to the tackling of a looming medical emergency.
Thatcher was not alone in questioning whether it was possible to get accurate information from surveys that ask people for intimate details of their sex lives. The team designed the survey to include sensitive questions being written on cards so that only numbers or letters would be required in response and a self-completion questionnaire. "We did a great deal of work to help ensure that people were consistent in what they said and designed the questions so that privacy and confidentiality were maximised," says Johnson. "We also took care to define terms such as sexual partner clearly. We've always argued there is measurement error in almost any area of science, be it measuring people's incomes, how much they eat or sleep. Indeed, there is measurement error of the same sort in the laboratory."
This preparatory work included trying to find a form of language that would put respondents at ease and make them most likely to provide accurate information. Kinsey's advice in his landmark publications of 1948 and 1953 was to use everyday terms to describe sexual practices and acts. This was also the practice among researchers working with gay men at the time. However they found that for a study involving a wide range of people of different ages and from different backgrounds, use of clinical language in questions alongside a glossary achieve the best results.
The first survey: NATSAL I
By the end of 1991, nearly 19 000 men and women aged 16-59 had been interviewed for NATSAL I. Key findings that could be used for public health strategy, and for attempts to estimate the magnitude of the HIV epidemic, were published in 'Nature' in December 1992. The study found that 6.1 per cent of men said that they had had some form of homosexual experience, and 3.6 per cent reported having had homosexual partners. These results were consistent with those of studies carried out in other European countries and the USA. There were striking regional variations: in London, the proportions reporting homosexual experiences and partners were close to double the national average.
The proportions reporting injecting non-prescription drugs were 0.8 per cent of men and 0.4 per cent of women. The team found that those with large numbers of heterosexual partners and men with large numbers of homosexual partners had the highest rate of attendance at STD clinics, emphasising these clinics' importance in promoting sexual health. More than four in ten men reporting homosexual partners in the past five years had already had HIV tests. This helped to provide an indication of the level of undiagnosed HIV infection in the population.
Detailed breakdowns of the numbers of heterosexual partners by age and sex were also obtained. They showed, for example, that 26.9 per cent of men aged 16-24 reported having had two or more sexual partners in the previous year. Those in the higher social classes and those who reported having had sex for the first time before the age of 16 were significantly more likely to report multiple heterosexual partners. The overall numbers of sexual partners appeared to be increasing across generations: more of those aged 25-44 reported having had ten or more lifetime sexual partners than did those aged over 45. Age of first sexual encounters was also significantly lower for the younger age groups. The study demonstrated a clear link between age of first sexual experience and sexual health.
The full results of NATSAL I were released in a book published by Blackwell Scientific Publications in 1994. In response to criticisms that the study might attract prurient interest, and was on subjects inappropriate for scientific investigation, the team consciously chose especially dry language when discussing their findings. One headline that sticks particularly in Johnson's memory was "No sex please: we're scientists," over a 'Sunday Telegraph' article on the results of the first survey. Beneath a picture of Johnson, Wellings and fellow team members Julia Field and Jane Wadsworth, all wearing very serious expressions, was the caption "Four women in search of vital statistics".
The data were used to underpin projections of the likely spread of HIV, in the development of AIDS prevention and general sexual health strategies, for sex education, and for contraceptive services planning. "It had very wide uses that had not necessarily been anticipated," says Johnson. For example, the findings were used in the discussions that preceded the lowering the age of consent for gay sex from 18 to 16 in England, Scotland and Wales in the year 2000, bringing it in line with the heterosexual age of consent for the first time.
The second survey: NATSAL II
The main impetus behind NATSAL II came from a working group of the Public Health Laboratory Service, chaired by Professor Nicholas Day of the University of Cambridge, which produced projections for the spread of HIV. Towards the end of the 1990s, it was becoming apparent that the data from the first study were getting outdated. A second survey could not only provide up-to-date information for estimating the extent of the HIV epidemic but also look at changes in behaviour over time.
A significant change in the survey's methods was the use of computer-assisted interviewing and self-interviewing techniques to improve response rates and the quality of the data. Other changes included the taking of urine samples to measure the prevalence of chlamydia, a younger age group of 16-44 to focus resources on those at greatest risk, over-sampling of people in London in order to increase the numbers of those in risk groups in the study, and over-sampling of ethnic minority groups to gain a greater understanding of differences in their sexual lifestyle.
NATSAL II surveyed a little over 11 000 people in 2000 and was funded by the Medical Research Council (MRC). It threw up some unexpected results, most strikingly that there were large increases in behaviours associated with higher risk of transmitting HIV and other sexually transmitted infections. These included increased numbers of heterosexual partners, more concurrent partnerships and more payment for sex.
"At the time that surprised us," says Johnson. "But looking back I think, why did that surprise me? When we measured behaviour in 1990, it was after the national campaigns, and the 'don't die of ignorance' leaflet drops. We'd seen a huge change in sexual behaviour from the mid-1980s and 1990, with a dramatic decline in STIs. From around 1995 these started creeping back up. So the big increase we saw was from the nadir of risky behaviours."
In the survey, the proportion of men who reported having had homosexual partners increased from 3.6 per cent to 5.4 per cent. In women, the jump was from 1.8 per cent to 4.9 per cent. The researchers, whose work was published in the 'Lancet' in 2001, concluded that a liberalisation of attitudes had led to greater willingness to report behaviours that had previously attracted greater social censure. They also noted that greater tolerance could have led to behavioural changes.
The urine samples showed for the first time that there was as much asymptomatic chlamydia in men as in women, leading to men being included in the national screening programme established in 2003. NATSAL II also helped inform the introduction of HPV vaccination in girls aged 12-13 in 2008.
Up-to-date: NATSAL III
Interviews for NATSAL III began last year and will be completed next summer. It will involve 15 000 people. Again, the design of the study has evolved: "We've moved from a STI/HIV-driven study to one that sees sexual health as something that is construed as a much more positive element of wellbeing," says Johnson. For example, in work being led by Wellings, the team will measure testosterone in saliva samples in both men and women. The aims are to gather information on the normal levels in the population and examine possible links with libido, pleasure and different sexual behaviours. The survey will include a broader age range of 16-74, and include more general health questions. This will allow the team to look at changes in sexual lifestyle over time and to better study the relationship between general physical and sexual health. This time, participants will be tested for five sexually transmitted infections: chlamydia, gonorrhea, HPV, HIV and Mycobacterium genitalium. The Wellcome Trust is contributing £2 million towards the £7.2m cost of the study, with the MRC, the ESRC and the Department of Health paying the remainder. Results are expected in 2013.
A problem that has faced all researchers carrying out survey studies has been falling response rates in recent years. The first and second NATSALs achieved 66.8 per cent and 65.4 per cent respectively. The team was horrified when the response rate for the first pilot for NATSAL III plummeted to 37.9 per cent. "That was a shock," says Johnson. "Then we realised the advance notice letters were far too wordy. We got some plain English advisers to help us to re-word them. One of the problems is that the governance and ethics people want you to include lots of detail, but it can just confuse people. It makes you realise how important science communication is." Measures including the re-wording of the letters helped boost the response rate to around 55 per cent in the early NATSAL III interviews.
"As far as I know, there is no other country in the world that has done as many large, detailed surveys on sexual lifestyles," says Johnson. "It's great to think we've done a lot of the work to lead the field, both in getting the surveys done but also in ways of getting the most accurate reporting."
What differentiates the NATSAL studies from other research on the same topics is their large, randomised and representative samples, which provide unparalleled data quality. While Kinsey was undoubtedly a pioneer, there was an overrepresentation of prisoners, male prostitutes and those most willing to discuss taboo subjects in his samples. There is no surprise that fellow researchers doubted the validity of his findings, such as 37 per cent of men reporting at least one homosexual experience. The multiple and wide-ranging advances that have been achieved in sexual health and education programmes thanks to the NATSAL surveys are the result of the conviction of Johnson and her colleagues, in the face of 1980s academic orthodoxy and political resistance, that the only way to obtain meaningful, useful results was to carry out large, properly randomised household studies.
Johnson worries that the readers of magazines aimed at teenagers may get a false impression of 'normal' sexual behaviour because they do not understand that the surveys in such magazines are unrepresentative. "People prepared to respond to a magazine survey may well be more sexually active than average," she says. "Teen magazines give the impression that everybody is having sex before the age of 16, whereas NATSAL surveys have shown a much lower proportion having sex at this age. It is really important that young people understand what is really going on."
The NATSAL surveys came about as a result the efforts of many individuals, a point Johnson is keen to emphasise. "Although I was principal investigator on the grant applications, this has very much been done as a partnership with others," she says. Wellings has worked alongside Johnson from the very start. Apart from those already mentioned who worked on the first survey, other key individuals who became involved later are Bob Erens, an independent research consultant, Dr Kevin Fenton, now a senior figure at the Centers for Disease Control and Prevention in the USA, and statistician Dr Cath Mercer, at UCL.
Johnson, whose full job title is Professor of Infectious Disease Epidemiology and Co-Director of the Institute for Global Health at UCL, was appointed to the Wellcome Trust Board of Governors in January 2011. When asked whether there will be further NATSALs, she appears uncharacteristically in two minds, at first hoping they will become less necessary before deciding they probably will still be needed. "I would like to think that part of what NATSAL has achieved is an acceptance that measuring people's sexual lifestyles will become part of the regular repertoire of general health questionnaires." She sees evidence of this starting to happen, with sexual behaviour questions being included in the NHS's Health Survey for England for the first time last year, and forming part of the long-term UK Biobank study. "These are huge changes," she says. "However, general health surveys will probably never have the time to collect all the data we'd like, so yes, I expect and hope there'll be a place for NATSAL in the future."
A Witness Seminar on NATSAL was held at the Wellcome Trust in December 2009. It will be published in July 2011 and freely available to download. Overy C et al. (eds). History of the National Survey of Sexual Attitudes and Lifestyles. Wellcome Witnesses to Twentieth Century Medicine, volume 41. London: Queen Mary, University of London; 2011. ISBN 978 090223 874
Nic Fleming is a science writer based in London.
Top image: Professor Anne Johnson. Credit: Wellcome Library