Sexual orientation health inequality: Evidence from Understanding Society , the UK Longitudinal Household Study

Few studies from the United Kingdom have fully investigated inequalities between members of different sexual minority groups and heterosexuals over range of health outcomes. Using data from over 40,000 individuals, this study explores the health inequalities of sexual minority UK adults. We include respondents who identify as other and those who prefer not to say (PNS). Data come from wave three (2011 – 2012) of the nationally-repre-sentative Understanding Society , the UK Household Longitudinal Study. Sexual orientation was asked in the self-completion portion ofthestudy. Markersofhealth includephysical andmentalfunctioning,minor psychological distress, self-rated health, substance use and disability. Multiplelinear and logistic regression analyses tested for differences in markers of health between sexual orientation groups. Overall, heterosexual respondents had the best health while bisexual respondents had the worst. Gay and lesbian respondents reported poorer health than heterosexuals, speci ﬁ cally with regards to mental functioning, distress and illness status. The other and PNS respondents were most similar to each other and generally experienced fewer health inequalities than gayandlesbianrespondents;theywerelesslikelytousetobaccooralcohol.Insum,sexualminoritiesexperience healthinequality.The inclusion ofother andPNS respondentshas notbeen done inother studies and shows that while they may behealthierthan gay/lesbianand bisexualrespondentsthey still experiencespoorerhealth than heterosexuals. Health promotion interventions are needed for these other and PNS individuals, who might not participate in interventions targeted toward known sexual minority groups.


Introduction
A recent Equality and Human Rights Commission (EHRC) Report on the lives of sexual minorities in the United Kingdom (UK) concluded that more research needs to examine the factors that contribute to the health inequalities experienced by this population in the UK (Mitchell et al., 2009).Specifically, research should be based on nationally representative samples, rather than those based on convenience sampling, to provide a more accurate picture of lesbians, gays and bisexuals (LGB) physical and mental health (Mitchell et al., 2009).Markers of health such as self-rated health (SRH), disability and substance use are predictors of mortality and are associated with increased risk of chronic disease (Mavaddat et al., 2014;DeSalvo et al., 2006).Additionally, substance use is a preventable risk factor for a myriad of chronic diseases such as cardiovascular disease, cancer and liver disease resulting in early mortality (Ezzati et al., 2002;Ronksley et al., 2011).
Few UK studies have examined the relations of self-reported sexual orientation, identity and attitudes with health.The National Survey of Sexual Attitudes and Lifestyles (Natsal) is a UK nationally representative cross-sectional study that asks about sexual behaviors, partnerships and attitudes toward sex and sexuality and self-reported sexual orientation (Mercer et al., 2013).Natsal data have used to report on sexual behaviors and attitudes of the British population, however much less has been published on the health of the UK sexual minority population.One study found no significant differences in self-rated health (SRH) between women who exclusively had sex with men and those who exclusively had sex with women.Women who had sex with both genders had significantly lower SRH than women who exclusively had sex with men or women (Mercer et al., 2007).Both women who exclusively had sex with women and women who had sex with both genders were more likely to experience an illness or visit the hospital compared to women who exclusively had sex with men (Mercer et al., 2007).To our knowledge, no equivalent study has been conducted with men who participated in Natsal.
Two large-scale studies have been conducted in the UK with the sole purpose to understand the health of gays, lesbians and bisexuals in the UK (Guasp, 2013;Hunt & Fish, 2008)

Contents lists available at ScienceDirect
Preventive Medicine  2011(Guasp, 2013;;Hunt & Fish, 2008).Similar percentages of men (76%) and women (80%) rated their health as good or very good, which is slightly higher for men and lower for women compared to the general population (Guasp, 2013;Hunt & Fish, 2008).Higher levels of smoking and drinking compared to the general population were also reported (Guasp, 2013;Hunt & Fish, 2008).Yet, these studies do not adjust for socio-demographic characteristics, which might change the scope of the differences within sexual minority populations or between sexual minorities and heterosexuals.
Much of the UK literature on LGB health focuses on sexually transmitted illnesses (STIs), substance use and suicidal behaviors, which are disproportionately higher in this population.Little has been published on other aspects of health such as anxiety, physical functioning, disability and limiting illness.Additionally, studies tend to focus solely on the LGB population and comparisons with their heterosexual counterparts are rare.Often individuals who respond as other or prefer not to say are dropped from analyses.In the 2014 Integrated Household Survey, 0.3% of respondents responded as other and 3.9% responded as Don't know/refuse (Office for National Statistics, 2015).This is the equivalent of about 2.3 million people living in the UK who identify as other or don't know (Office for National Statistics, 2015).Learning more about their health and what inequalities they might face is important to understanding the health of all sexual minorities.
This study examines variation in markers of physical and mental health among both sexual minority and heterosexual individuals.Additionally we address the potential of multiple minority status, by exploring differences by socio-demographic characteristics.Our research questions are: 1. Are there differences in physical health, e.g.physical functioning, selfrated health, illness status, between heterosexual and sexual minority individuals after adjustment for socio-demographic characteristics? 2. Are there differences in mental health, e.g.lower life satisfaction, greater anxiety, between heterosexual and sexual minority individuals after adjustment for socio-demographic characteristics?3. Are there differences in substance use, e.g.smoking, alcohol consumption, drug use, between young heterosexual and sexual minority individuals after adjustment for socio-demographic characteristics? 4. Are the differences in research questions 1-3 larger or smaller among certain sexual minority groups, e.g.gay men, lesbians, bisexuals, other or prefer not to say?

Sample
Data come from waves two and three of the Understanding Society: the UK Household Longitudinal Study (UKHLS).This nationally, representative study began collecting data in 2009.Respondents are interviewed annually and all adults in the household 16 and older are asked to participate in the main survey.The survey contains two parts: a computer-assisted personal interview (CAPI) and a self-completion survey conducted on a computer.In 2009, just over 50,000 individuals in over 35,000 households were interviewed.Sampling scheme details, data collection methods and annual response rates are available (Burton et al., 2011;Lynn, 2009).

Measures
UKHLS covers a wide range of topics including but not limited to socio-demographic characteristics, employment and educational attainment, marital status and family structure and health.
Sexual orientation was asked in wave 3 using the question "Which of the following options best describes how you think of yourself?"Responses were "heterosexual or straight", "gay or lesbian" (GL), "bisexual", "other" and "prefer not to say" (PNS).Sexual orientation was asked of all adults who consented to complete the self-completion portion of UKHLS.Heterosexual or straight is the reference category in all analyses.

Markers of physical health
Self-rated health responses ranged from "excellent" to "poor".Due to small numbers in the highest and lowest categories, categories of SRH were combined into: Good health (excellent and very good), moderate health (good) and poor health (fair and poor).Illness status was determined by two questions the first asks about disability and the second about specific types of disability.Three categories were calculated to indicate illness status: No illness, non-limiting long-standing illness (NLLSI) and limiting long-standing illness (LLSI).The SF-12 is a wellestablished and validated measure of health functioning (Ware et al., 1998;Busija et al., 2011).The SF-12 provides two summary scores, the physical component score (PCS) and the mental component score (MCS).Both the PCS and the MCS were scored to have a range of 0-100 with a mean of 50 and a standard deviation of 10 (Ware et al., 2001).

Markers of mental health
The General Health Questionnaire (GHQ-12) is a measure of generalized psychiatric morbidity and the Likert-scoring method to produce a total score with a range of 0-36 (Goldberg & Williams, 1988;Goldberg et al., 1997).Two questions scored on 7-point Likert scale were used to assess health and life satisfaction.For all markers of mental health, higher scores indicated better health.

Substance use
Smoking status and history questions and alcohol consumption behaviors were asked at wave 2. Smoking status was created from a combination of two questions and was categorized as: Never smoker, current smoker and former smoker.Alcohol consumption was assessed from consumption in the past 12 months.Due to small numbers in the highest and lowest categories, the seven responses were combined into five categories ranging from "Once a week or more" to "never had a drink".Young people aged 16-21 were given an additional self-completion module at wave 3 in which they were asked about ever smoking, past month alcohol consumption, binge drinking in past four weeks, and drug (i.e.cannabis, solvent and other) use and frequency of drug use since last interview.

Covariates
Socio-demographic characteristics were included to describe the different sexual minority groups as well as covariates in the regression analyses.Age was included as a continuous variable in regression analyses, but for descriptive purposes age was grouped.Males are the reference group for gender.Ethnicity was grouped into five categories: White British (reference category), Asian (Indian, Bangladeshi and Pakistani), Black African/Caribbean, Other and Mixed.Marital status had three categories: single (reference category), partnered and previously partnered.Economic activity was also a three category variable: employed (reference category), economically inactive and unemployed.Highest educational qualification was a six category variable with a range of no qualification (reference category) to degree (e.g.University).Religion was categorized as Christian (reference category), Muslim, Hindu, Buddhist and Other.UK generation status ranged from 1st to 4+ (reference category) and includes an "other" category.

Statistical analysis
Chi-square tests were used to test for equal distributions of categorical variables across sexual orientation groups while general linear models (GLM) were used to test differences in the means of continuous variables across groups.GLM models controlled for age and gender.Associations between sexual orientation and health were tested via linear and logistic regressions.All regression models controlled for covariates.
j o u r n a l h o m e p a g e : w w w .e l s e v i e r .c o m / l o c a t e / y p m e d . The Lesbian and Bisexual women's health check was conducted in 2007 and a complementary study the Gay and Bisexual Men's Health Survey was conducted in http://dx.doi.org/10.1016/j.ypmed.2017.06.010 0091-7435/© 2017 The Authors.Published by Elsevier Inc.This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).