Sexual health and ethnicity

Racial categorisation, where classification is based on the individual’s physical a ttributes (e.g. eye and skin colour) has been and is increasingly used as a variable for analysis in epidemiological and social research. Proponents argue th a t the po ten tia l benefits are many, and include the ability to detect variations in health across racial groups, investigate disease aetiology and inform targeted in terventions. However, its use has also been criticised (Bhopal et al. 1998, Bhopal and Rankin 1999). Racialisation of health and health outcomes tends to be reductionist, attribu ting observed differences to ‘physical a ttribu tes’ ra ther than social or economic factors, which may be stronger determ inants of the outcome of in terest (Wyatt 1991, Bhopal and Rankin 1999). Racial categorisation may also (intentionally or not) reinforce racial stereotypes, and in fields such as sexual health may lead to wariness among com m unities about the in tentions and values of the researchers (Fenton et al. 1997). O thers argue tha t given the rich cultural diversity in Britain today, knowing tha t the prevalence of a particular disease is higher among ‘Black’ people is of lim ited use in targeting interventions.In an effort to overcome some of these lim itations, the use of the word ethnicity, as opposed to race, has been advocated (Wyatt 1991, Bhopal and Rankin 1999). Ethnicity implies that an individual’s socialisation is part of a collective (and m easurable) identity that is socio-culturally based. Although by no means ideal, it is seen as one m ethod of incorporating and m easuring the influence of cultural factors on health experience and outcomes.Discussions about sexual health and ethnicity, therefore, involve not one bu t two taboos. E thn ic in eq u a litie s in sexual h ea lth ra ise p a r ticu la r sensitivities, with concerns about irresponsible handling of statistical data, racial stereotyping and fears about stigm atising communities (Fenton et al. 1997). This perhaps explains the relative lack of data available to us, which limits our ability to explore variation in sexual behaviour and sexual health status with ethnicity. Surveillance data and ad hoc research studies in the United States have shown variations in sexual health outcomes between broad ethnic groups (‘W hite’/ ‘African American’/ ‘Hispanic’). Investigation of similar inequalities in B ritain is relatively unexplored. This is in part because ethnicity data are generally not collected in routine heath statistics. Ad hoc

surveys on sexual health outcomes have focused on attenders of sexually transm itted disease (STD) clinics or have exam ined differences across racial categories (e.g. ‘Black’, ‘W hite’, ‘O th e r’).Some lim ited data are available from large-scale cross-sectional surveys, for example the National Survey of Sexual A ttitudes and Lifestyles (NSSAL) collected data in 1990 and 1991 from nearly 19,000 respondents aged between 16 and 59 in England, Wales and Scotland. Yet this survey suffered a num ber of methodological problems with regard to ethnicity. These stem partly from classification issues and partly from the size of the sample involved and its implications for analysis. C hief among the problems relating to sample size is the difficulty of adjusting for confounding by socio-economic variables. Self-identification was used to elicit ethnicity, providing a show card from which respondents selected the group they considered they belong to, the categories being those used by the Office for National Statistics (ONS) at the time. The category ‘Black’ was not disaggregated into Black African and Black Afro-C aribbean . This survey is cu rren tly being rep ea ted and the e thn ic ity classification question to be used in the second survey rectifies this oversight.Nevertheless, the quantitative data suggest considerable variation in sexual behaviour with ethnicity, most notably in initial sexual experiences and first sexual intercourse. NSSAL data show the ethnic group called Asian to be powerfully protective in term s of early intercourse. Asian women begin sexual experience 3 years later than their Black and W hite counterparts, with a m edian age of 21 (Johnson et al. 1994). A sm aller proportion of women who reported as Asian (11 per cent) had intercourse before the age of 16; among those who self-identified as W hite the proportion was 19 per cent, and it was higher among those who self-reported as Black, at 26 per cent. These figures are for the total sample for all ages. There has been a progressive decline in age at first in tercou rse and the proportions were h igher for younger respondents in the sample, though the num bers are small. In survival analysis there were significant differences between curves for these groups.Despite later sexual activity and an increased tendency for sex to take place within m arriage where childbearing is generally less problematic, Asian women were more likely to have had an abortion both at some point in their lives, and in the more recent time period of the last 5 years, than W hite women (Johnson et al. 1994). The increased prevalence of term ination of pregnancy seems somewhat incompatible with the evidence of delayed first intercourse occurring with m arriage in this group. The use of contraception, however, showed a steep gradient by ethnic group. Asian women being considerably less likely to have used contraception in the last year than Black and W hite women.More recent population-based analyses of attenders of STD clinics have uncovered evidence of ethnic variations in the prevalence of diagnosed sexually transm itted infections. Lacey and colleagues (1997) recorded details of all residen ts w ith in the boundaries of Leeds H ea lth A uthority who presented with gonorrhoea (culture confirmed) from April 1989 to September

1993 at Leeds General Infirmary, the only STD clinic serving the city. High rates of gonorrhoea were observed among Black m en, and younger men were at highest risk with an incidence of 2-3 per cent per year. The neighbourhoods with the highest rates of infection were inner-city areas with high proportions of ethnic m inority groups. However, after controlling for age, sex and socio­economic group, Black men and women in Leeds were more than ten times more likely than W hite men and women, and fifty times more likely than Asian men and women, to have had one or more episodes of gonorrhoea during the study. W hite men and women were nearly five times more likely than Asians to have had one or more episodes of gonorrhoea during the study period.Low and others (1997) contacted sixteen departm ents of genito-urinary medicine (GUM) in Lam beth, Southwark and Lewisham and collected data from eleven departm ents during 1994-95. Again the highest rates of STD were found among young Black men with incidence rates of 1-2 per cent. Women from Black minority ethnic groups had around ten times the rate of gonococcal infection than tha t seen in W hite women. Rates were higher among Black men than among Black women for all age groups except the youngest - aged 15-19 years. Again inequalities in gonorrhoea rates persisted after adjusting for socio-economic confounding; men from Black m inority ethnic groups being eleven times more likely than those from W hite groups to acquire gonorrhoea. Roughly one in ten gonorrhoea episodes reported from departm ents of GUM in England occurred among residents of Lewisham, Southwark and Lam beth.Both these studies show m em bership of a Black ethnic group to be associated with a higher risk of acquiring gonorrhoea, as judged by incidence rates, even after controlling for socio-economic status. By any standards, the differences in disease incidence between the groups give cause for concern. Why are rates so high among ethnic populations in the U nited Kingdom? T here are no known biological reasons to explain why racial or ethnic differences alone should alter the risk for STD or unwanted conceptions. The explanatory factors may be epidemiological or socio-cultural.From an epidemiological perspective, the association between ethnicity and sexual health may be due to chance, bias or true association. Chance is unlikely to account for the association, given the consistency of findings across research studies. Biases in the design of studies may account for some of the differences. A ttenders of STD clinics are more likely to have high-risk sexual lifestyles, increased num bers of p a rtn e rs and to have been previously diagnosed with an STD (Johnson et al. 1996). H ealth-seeking behaviours will also influence who goes to STD clinics and when, which leads to recruitm ent bias among STD attenders. These factors make STD clinic data unrepresen t­ative of the wider population and may lead to an overestim ation of high-risk sexual behaviours if findings are extrapolated to the wider community.If there is a true association between ethnicity and sexual health outcomes what factors may explain this? The risk of spread of a sexually transm itted

infection (STI) within a given population is denoted by the basic reproductive rate (R0), which is dependent upon the transm ission probability (P), the rate of partner acquisition (c), and the average duration of infectiousness (D). P and D are the two biological factors within the model. However, all may be influenced by cultural factors.For example, condoms reduce the probability of disease transm ission by providing an effective barrier. NSSAL failed to find significant differences between the ethnic groups in term s of condom use. Nevertheless, the study revealed a tendency, docum ented elsewhere, for a preference amongst Black men and women for systemic (non-barrier) methods - tha t is, those not directly re lated to the act of in tercourse (Elam et al. 1999). Some have suggested that this may be explained by a cultural tendency to equate virility with fertility and a consequent preference for m ethods, which are not specifically intercourse related; there was also less tolerance for surgical methods of contraception among either Black or Asian men and women com pared with those self-defining as W hite. The im portance of specific cultural practices, for example female and male circumcision and the use of vaginal drying agents, in de term in in g sexual and reproductive hea lth outcomes have been highlighted in ethnographic studies (B row ns aL 1993, Tyndall et al. 1996). Here, culturally prescribed practices can increase the risks associated with disease transm ission. Although such practices are theoretically am enable to intervention, effecting change may be difficult where practices are seen as being a part of a cultural identity.V aria tio n s in sex u a l h e a l th are also ex p la in e d in te rm s o f th e in terrelationship between an individual’s culture and ethnic background on the provision and use of sexual health services. This in tu rn may influence the duration of infectiousness (D). The stigm a of sexually transm itted diseases and of attendance at GUM clinics varies with cultural group (Elam et al. 1999), and it may m ean that individuals delay accessing these services or taking advantage of available interventions. Negative attitudes towards GUM clinics, clinic staff and doctors have been docum ented across a variety of ethnic groups (F en to n^ al. 1999a). Indeed, there is some evidence that, for Asian and African communities, the general practitioner (GP) rem ains the choice of first call for advice and information on sexual m atters. Among Black gay and bisexual men the negative attitudes of clinic staff, insensitivity of clinic doctors and confidentiality concerns have been cited as the main reasons for not attending STD clinics and seeking care outside this sector (Fenton et al. 1999b). The attitudes are not without foundation and many inner-city minority communities may be further disadvantaged by the delivery of sub­standard sexual health services and interventions (Monk 1992). Negative attitudes towards minority groups in a society may be reflected in the views of health service providers, and the resultant prejudice and discrimination may serve not only to hinder health service use but also to dim inish the benefits received from such use. Negative experiences associated with accessing GUM clinics, such as racism, insensitive or intolerant staff and

poor facilities, serve to m aintain the stereotypes of STDs afflicting those at the margins of society.The m ain behavioural factor influencing disease transm ission is the rate of partner change, c, or the rate of acquisition of new partners. Few studies have looked at variations across ethnic groups. Although the total num ber of reported sexual partners may give some indication of the m agnitude of partner change, it is difficult to distinguish between concurrent and sequential partners. Univariate analysis in NSSAL suggested that Black men reported having twice the num ber of sexual partners in a lifetime than W hite m en did (with medians of eight compared to four respectively), but the num ber of observations was small and m ultivariate analysis was not carried out to adjust for confounding factors. However, a twofold increase in the number of partners among Black m en seems unlikely to account fully for the tenfold difference in the incidence of disease. Qualitative studies highlight variations across ethnic groups in a ttitudes towards m ultiple and concurrent relationships (Elam et al. 1999).Partner selection and the question of ‘who mixes with whom’ is also an im portant aspect in the transm ission of STDs. People tend to have partners w ith in th e ir own e thn ic group, thereby increasing the risk of onward transm issio n of in fection w ith in p a r ticu la r com m unities. Barlow and colleagues (1997) exam ined patterns of assortative sexual mixing among GUM attenders, enquiring about the ethnicity of the last sexual partner, and found a close and closed pa tte rn of sexual mixing within ethnic groups.Taking into account differences in sexual behaviour, the transm ission dynamics of sexually transm itted infection, and the widespread deficiencies in appropriate sexual health services, we begin to get a sense of the differential weighting which should be attached to the predisposing factors. Yet anxieties around the tension between these positions may inhibit public health efforts. T here are moves on the part of in terested groups to improve services. Blackliners, for exam ple, are working with staff to improve services and researchers are evaluating ways of effectively involving communities in sexual health-related research and development initiatives (Fenton et al. 1999c). For such interventions to be effective, however, the potential for racism in sexual health services needs to be recognised and addressed across a wide range of agencies and sectors.From a sociological perspective, ethnic and cultural background have the potential to influence sexual health in a num ber of ways: how we learn about sexual m atters, our a ttitudes and practices related to sex, our choice of partners, how we relate to and accept sexual health prom otion messages and utilise sexual health services. In concert, these influences may either act to place some groups at increased risk of adverse sexual health outcomes, such as infection or unplanned conception, or lead to more protective m easures in other groups.In all communities, cultural norms operate to proscribe certain attitudes and behaviours and to prescribe others. In general, social mores correspond

closely with cultural norms. Societies which condone, or have relaxed social attitudes towards, for example, early sexual experience, concurrent sexual relationships or contact with sex workers, are more likely to have higher prevalence of these behaviours (Cleland and Ferry 1996). W here certain behaviours are effectively censured, the impact on sexual health status may well be protective, examples being delayed coitarche and sex within m arriage. As STD prevalence in Britain is highest within m id-late adolescence this ‘natural intervention’ has a potentially protective effect on STD acquisition by reducing the num bers of sexual partners, rate of partner change and the probability of contact with the prevalent pool of infection.Cultural influences on sexual lifestyles and attitudes are not static however. Com plex p a tte rn s of sexual lifestyle occur, where com m unities are in transition and where there is age-related diversity within the group. Young people may, for example, increasingly share the social norms of the community into which they are integrating, while their parents and other older members of the group may retain more traditional norms. Living within a more sexually open society can have im m ediate effects on an individual’s a ttitudes and lifestyle. In Britain, exposure to a more open society has resulted in increased knowledge about sexual and reproductive health in general, alternative sexual lifestyles and a range of sexual practices, even am ong first generation im m igrants but most markedly among successive generations (Elam et al. 1999). For those born in Britain, exposure to individuals from other ethnic backgrounds (including the ethnic majority), to sex education in schools, to ease of access to sexual m aterial and to the onslaught of m edia in terest in sex has opened the discourse on sexual behaviour. Even among cultures where it is currently the norm not to discuss sexual m atters, external informal and formal influences will change the m anner in which successive generations of young people are socialised regarding sexual m atters. Less positively, rapid in tegration into another culture may lead to discordance between sexual behaviours as currently practised and the cultural norms governing such behaviours. In such cases, behaviours may be more covert and less am enable to preventive intervention and this may have negative implications for sexual health status.C ultural variations in teract with social structural factors. As with other health -related conditions, the socio-economic experience of many ethnic minority communities in Britain places them at a disadvantage as far as sexual health status is concerned. Sexual health promotion in socially deprived or economically disadvantaged areas m ust compete against other priorities. Im m igration concerns, economic survival, child rearing and social integration are often more pressing and im m ediate concerns for many ethnic minorities. The effect of religion may reinforce cultural influences on sexual behaviour. Among some com m unities, the influence of religion may be difficult to untangle from tha t of culture, for example in many Islamic communities. Also, religious dictates may be used to strengthen cultural messages relating to behaviours and practices (Elam et al. 1999).