The association between region of birth and sexually transmitted infections among people of black Caribbean ethnicity attending sexual health services in England, 2015

Background/Introduction In England, people of Black Caribbean (BC) ethnicity are disproportionately affected by sexually transmitted infections (STIs), but it is unclear whether this varies by their region of birth. Aim(s)/Objectives To examine differences in STI diagnoses among UK- and Caribbean-born BC people. Methods Data on STI diagnoses in BC people attending specialist sexual health services (SHSs) during 2015 and living in England were obtained from the GUMCAD STI surveillance system, the national surveillance system for STIs in England. Associations between being UK- or Caribbean-born and each of several STI diagnoses were examined, using univariate and multivariable generalised estimated equations logistic regression models adjusted for sexual orientation, place of residence (London vs. non-London), HIV status, area-level deprivation, and STI diagnosis in the last year. All analyses were stratified by age (<25 vs. ≥25 years). Results In 2015, 63,568 BC people made 108,881 attendances at specialist SHSs; 81.9% of these attendances were made by UK-born BCs. The median age (years) was 26 for UK-born and 35 for Caribbean-born people (p≤0.001). Chlamydia, gonorrhoea and non-specific genital infection (NSGI) were the most commonly diagnosed STIs among UK- (5.8%, 2.1% and 2.8%) and Caribbean-born people (4.5%, 1.7% and 3.5%) respectively. Among BCs aged under 25, no significant differences in STIs were found between UK- and Caribbean-born people. Among BCs aged ≥25, compared to Caribbean-born people, those who were UK-born were more likely to be diagnosed with chlamydia (AOR 1.15 [95%C.I. 1.04–1.27]); gonorrhoea (AOR 1.23 [95%C.I. 1.06–1.45]) and genital herpes (AOR 1.23 [95% C.I. 1.10–1.56]) and less likely to be diagnosed with NSGI (AOR 0.89 [95% C.I. 0.80–0.99]) and Trichomoniasis (AOR 0.84 [95% C.I. 0.71–0.99]). Discussion/Conclusion STI diagnoses in BC people aged ≥25 attending specialist SHSs vary by region of birth. Country of birth may have an influence on social and sexual networks and therefore transmission of STIs.


Introduction
Black ethnic minorities bear a disproportionate burden of sexually transmitted infections (STIs) in many high-income countries [1]. In the UK, the highest diagnosis rates of STIs are among people of black Caribbean (BC) ethnicity [2]. These high rates are a consequence of the interaction of cultural, socioeconomic and sexual behavioural characteristics [3][4][5], and these are likely to vary depending on an individual's region of birth.
The BC (i.e. Afro-Caribbean or African Caribbean) ethnic group includes people of African ancestral origins whose family settled in the Caribbean before emigrating to the UK [6]. BC people are a diverse population from at least two perspectives: first, they are from a number of different countries in the Caribbean, each with their own sociocultural influences [7] and, second, they comprise people born in the Caribbean who migrated some decades ago through to sixth-generation people who identify as BC, but who were born in the UK [8,9]. Prevalence of some STIs by region of birth have been described higher in the Americas than in Western Europe [10]. The extent to which region of birth is associated with the risk of STIs is unclear. Thus, we examined differences in the likelihood of being diagnosed with STIs between UKand Caribbean-born BC people attending specialist sexual health services (SHSs) in England.

Methods
Data on STI diagnoses in BC people living in England were obtained for 1 January-31 December 2015 from the GUMCAD STI surveillance system, the mandatory national surveillance system for STIs in England (formerly known as Genitourinary Medicine Clinic Activity Dataset). GUMCAD is a patient-level dataset containing information on STI diagnoses and services provided by all specialist SHSs in England, as well as key sociodemographic data, including country of birth [11]. Specialist SHSs refers to genitourinary medicine (GUM) and integrated GUM/sexual and reproductive health (SRH) services. Attendances made by people who selfidentified as BC but had unknown region of birth (14.2%) or were born outside of the Caribbean or the UK (4.3%) were excluded from the analysis, as were those of unknown gender (0.01%), unknown sexual orientation (2.0%) or living outside England (2.0%).
For this analysis, the following countries and territories were defined as being in the Caribbean: Antigua and Barbuda, Aruba, the Bahamas, Barbados, Cuba, Dominica, Dominican Republic, Grenada, Haiti, Jamaica, Netherland Antilles, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, and Trinidad and Tobago.

Clinical definitions
All STIs reported in GUMCAD were confirmed diagnoses, defined using national STI management guidelines [12], that were coded and reported as such.

Data analysis
Based on the Kolmogorov-Smirnov test [13], the age distribution of BC attendees varied by region of birth, therefore all models were stratified by age, using 25 years as a cut-off due to the higher STI diagnosis rates within people under 25 year olds [2].
Separate models were run for <25 and �25 year old UK-born compared with Caribbeanborn BC attendees to calculate the unadjusted odds ratio (OR) and adjusted odds ratios (AOR) of being diagnosed with each STI. In GUMCAD, a patient is permitted only one record relating to a particular STI in a six week period; repeat codes for the same STI within this period are considered as relating to the same episode of care and so are removed to prevent overcounting of diagnoses. The following STIs were included in the analysis: Chlamydia; gonorrhoea; NSGI; genital herpes (1st episode); HIV; genital warts (1st episode) and trichomoniasis. The following were considered as confounders: gender and male sexual orientation [14] (coded heterosexual male, men who have sex with men, or women), place of residence (London vs. non-London) [15], HIV status [16] (coded negative or unknown vs known positive), a history of STI diagnosis in the last year [17] (coded yes or no), and area-level socioeconomic deprivation defined using quintiles of the Index of Multiple Deprivation (IMD) [18] for each lower super output area (LSOA) of residence.
As this was an attendance-level analysis and many attendees had multiple attendances, all univariate and multivariable models were built using generalised estimated equations logistic regression models to account for these clustered observations.

Results
In 2015, among those with a known region of birth, 63,568 BC people made 108,881 attendances at specialist SHSs in England. These attendances were 3.9% of the total attendances that year (2,780,434). Of these 108,881 attendances by BC people, the majority (81.9%) were by UK-born BC people.
The median age of UK-and Caribbean-born BC people was 26 (min:15-max: 83) and 35 (min:15-max: 91; (p<0.001)) years, respectively. Fewer UK-born (52.0%) vs. Caribbean-born (64.3%; p<0.001) BC people resided in London. For both UK-and Caribbean-born people, STIs were most likely to be diagnosed in those living in the most socioeconomically deprived areas of England.
For those <25 years old (Table 1), and diagnosed with chlamydia, gonorrhoea, NSGI, genital herpes, or genital warts, a higher proportion of Caribbean-born people were London residents compared to UK-born BCs. Also, a higher proportion of Caribbean-born BC people with a history of STI diagnosis in the last year were diagnosed with chlamydia compared to UK-born BC people.
For those � 25 years old (Table 2), and diagnosed with gonorrhoea or trichomoniasis, a higher proportion of those Caribbean-born were London residents compared to UK-born BC people.
The unadjusted and factors adjusted associations between diagnosis with selected STIs and attendances by UK-vs. Caribbean-born BC people at specialist sexual health services, by agegroup, England, 2015 are shown in Fig 1. Among BC attendees <25 years, those who were UK-born were more likely to be diagnosed with genital herpes ( In the multivariable analysis, the association with region of birth remained statistically significant for all but two of these STIs (genital warts and HIV) with UK-born BC people being more likely to be diagnosed with chlamydia (1. 15

Discussion
In this analysis, we examined the differences in the likelihood of STI diagnoses among BC people by region of birth. While BC people have previously been shown to have higher STI diagnosis rates [19], we found that the likelihood of being diagnosed with STI varies markedly by region of birth among those aged over 25 years old. UK-born BC people were more likely to be diagnosed with the most commonly diagnosed STIs in England, i.e. chlamydia, gonorrhoea, genital herpes, while their Caribbean-born counterparts were more likely to be diagnosed with trichomoniasis and NSGI. The reasons for this are unknown but may be associated with differences in patterns of health-seeking behaviours, and/ or differences in their sexual networks [20,21].
There is evidence of differences in sexual health outcomes of UK-born compared to migrant populations. For example, a large proportion of all HIV cases among heterosexuals in most countries of the European Union originate from outside of Europe [22]. In England, black ethnic minorities and especially black Caribbean populations are at much greater risk of STIs, especially for gonorrhoea and trichomoniasis [19]. However, there is limited published evidence specifically for STIs in BC people living in England, and this was not differentiated by region of birth [23].
In our analysis, adjusting for confounders had very little impact on the ORs apart from for genital warts in those aged �25. It is interesting to note that the direction of association     between region of birth and chlamydia and gonorrhoea diagnoses differed by age-group. This might suggest that the sexual networks of UK-born vs. Caribbean-born BCs may be similar until the age of 25 and then changes occur in BCs aged �25 potentially due to change in partnership type and sexual mixing patterns, considered key determinants of STI transmission, with respect to demographic characteristics, general health, health behaviours and sexual history [24]. This also could be due to the length of time those born in the Caribbean had spent in the UK (i.e. those <25 years may have, on average, spent less time in the UK and might have had Relationship and Sexual Education in the UK with different sexual health outcomes compared to those who were older). Further investigation of behavioural factors may explain the differences in STI risk by region of birth. The potential role of partnership concurrency in maintaining high rates of bacterial STIs in BC populations has been documented [25][26][27]. It is possible that the differences in the risk of STIs by region of birth are related to different background prevalence of different STIs in different countries[10], and perhaps Caribbean-born who are residents in the UK travel more frequently to the Caribbean than the UK-born BC people, therefore more likely to be exposed there.
While GUMCAD is vast in scale and includes all attendances made at specialist SHSs in England, a limitation is that it does not yet collect detailed data on sexual behaviour (e.g. condom use, number of partners, or partnership concurrency), as well as broader health-related factors, such as alcohol and drug use, which may confound the observed associations. Furthermore, there is no information on factors that might play a role in the risk of STIs such as the length of time Caribbean-born people have lived in the UK and their frequency of travel to the Caribbean.
Historically, many Caribbean-born people were of working age when they migrated with their children to the UK after the British Nationality Act of 1948 [28,29]. This pattern of migration may have changed over time [30], and differences between those migrating to the UK as a child or an adult might help to explain differences in STI risk factors. This information has been collected as part of the work of the National Institute for Health Research Health Protection Research Unit in Blood Borne and Sexually Transmitted Infections [31] and will be explored in future analyses. Another limitation of this study is that, despite the socio-cultural differences and varying levels of development that exist among Caribbean countries, all Caribbean-born people were included in a single category for analysis as there were insufficient observations for each of the 15 countries and territories considered in this group.
This study offers insights into disparities in STIs by region of birth among BC people living in England. These disparities are partially explained by differences in demographics between UK-and Caribbean-born people but also explained by region of birth. However, future research should examine behavioural differences, including the role of sexual networks, concurrency, partnership type and numbers, condom use, and sexual healthcare-seeking behaviour to better understand ethnic disparities in STI diagnoses.