Tobacco use is the top preventable cause of death in Spain, killing more than 53,000 people yearly or one in every seven deaths.1 The top causes of death and disability in Spain are highly associated with tobacco use.2 In addition, the tobacco financial burden in Spain exceeds four billion Euros in yearly healthcare expenditures.3

A recent review of the literature indicated that tobacco use is more prevalent among sexual minorities.4 Hypothesized mechanisms for this association include higher levels of alcohol use, depression, discrimination, homophobia, and targeting by the tobacco industry.4–6 For example, according to the minority stress model, sexual minorities might experience a culturally rooted, chronic stress in the shape of negative experiences and depreciation that increases the risk of poor mental health.4 Poor mental health is associated with health risk behaviours such as smoking that serve as coping mechanisms.7 Sexual minorities are also more likely to be surrounded by people who smoke, which is associated with tobacco use, barriers in cessation and increased morbimortality.8,9 However, many studies assessing tobacco disparities among sexual minorities have used non-probabilistic samples, and there is a lack of representative data. Most of these studies (about 85%) have been conducted in the USA.10

To our knowledge, no Spanish study has assessed the association between sexual minorities and tobacco use at the national level. The 2010 European Men Who Have Sex with Men Internet Survey (EMIS) in Spain show that the prevalence of tobacco smoking in this group is 53.9%.11 This prevalence is markedly higher than the Spanish average in other national surveys (26%).12 However, the EMIS only included a non-probability sample of men who had sex with other men. Only one study has been conducted in Spain to assess the association between belonging to a sexual minority and smoking in a representative sample at the local level. Results from the Barcelona Health Survey show that the prevalence of current smoking among adults attracted by people the same sex was 61.6% whereas the prevalence among adults who had had sex with people the same sex was lower (42.0%).13 The odds of current smoking compared to adults attracted or who had had sex exclusively with people the opposite sex were 2.32 and 1.44, respectively, and statistically significant after controlling for covariates. One of the reasons for the lack of research on LGBT disparities in tobacco use might be the non-inclusion of questions in national surveys on sexual orientation.14 Sexual orientation was not assessed in important national surveys such as the Spanish National Health Survey (ENSE), 15 European Health Interview Survey in Spain (EESE),16 Spanish Survey on Alcohol and Drugs (EDADES);17 while other national surveys assessing sexual orientation focused mainly on sexual health, excluding tobacco use and other important health domains.18,19 The most recent national surveys included information on the gender of those living with the respondent which, like other studies,20 may be used as a proxy for sexual orientation. Therefore, the aim of this study is to analyse the two most recent Spanish national surveys to assess the association between active and passive tobacco use and living with same sex partners (SSP).

In the combined sample, there was a higher percentage of participants living with a SSP compared to an OSP who were younger, not married (36.4% vs 8.6%), non-nationals (21.6% vs 10.9%), had higher levels of education (43.2% vs 25.0% for tertiary education), belonged to a higher social class (53.5% vs 25.5% for class 1), were current active (53.5% vs 25.5%) and passive smokers (37.2% vs 16.0%) and binge drinkers (15.7% vs 7.1%) (Table 1).

Table 2 shows the adjusted associations of living with a SSP with current active and passive smoking controlling for age, gender and social class. In the combined sample, people living with a SSP had 2.71 (95%CI: 1.62-4.52) and 2.88 (95%CI: 1.74-4.79) higher odds of being current active and passive smokers respectively. When stratifying by gender, women living with a SSP had 4.54 (95%CI: 2.20-9.40) and 4.00 (95%CI: 1.88-8.49) higher odds of being current active and passive smokers compared to women living with an OSP. However, the association with active smoking was not statistically significant (OR: 1.86; 95%CI: 0.94-3.70) among SSP men. Men living with SSP had 2.20 (95%CI: 1.10-4.43) higher odds of being current passive smokers compared to men living with OSP. When analysing the separate surveys, women living with SSP were consistently more likely to be current active and passive smokers, whereas their male counterparts were only more likely to be passive smokers in the 2014 survey.

To our knowledge, this is the first attempt to examine the association between living with a SSP and smoking at the national level in Spain. A strength of this study is that we used a sample from surveys representative of the Spanish population. We found that compared to people living with OSP, those living with SSP were more likely to be current active and passive smokers. We also found that the association between current smoking and living with a SSP is especially strong among women living with SSP, as they are about four times more likely to be current active and passive smokers than women living with OSP.

With some exceptions, the association between sexual minorities and cigarette smoking is widely established in the USA, but evidence is lacking elsewhere.5,10 Associations found in this manuscript resemble those found among adults attracted for people the same sex in Barcelona.13 The prevalence of smoking found in this study is also similar to the prevalence found in the 2010 EMIS in Spain.11 Results from this study are in line with findings from a recent review showing that the association between sexual minorities and smoking is stronger among women.10 Our findings are particularly similar to findings from the 2003–2010 National Health and Nutrition Examination Surveys in which associations between sexual orientation and different outcomes of active smoking were only found among lesbians and associations with passive smoking were stronger in this group too.8 Stronger associations between passive smoking and women living with a SSP suggest that the social network of women living with a SSP might be more likely to be formed by smokers. Evidence shows that smoking by social network members is linked to higher rates of smoking and lower rates of cessation.27,28 The findings showing the higher prevalence of passive smoking among adults living with SSP are consistent with data from the USA National Longitudinal Study of Adolescent Health in which same-sex-attracted youth had a higher frequency of tobacco use in their peer networks than did opposite-sex-attracted youth.29

Demographic characteristics including the younger age and higher percentage of non-nationals among adults living with SSP are consistent with results from the 2011 Spanish census.30 Younger cohorts might be more likely to live together due to an increasingly higher acceptance of same sex relationships, especially on behalf of people their generation. For example, a survey conducted in 39 countries, found that the view that homosexuality should be accepted by society increased in Spain from 82% in 2007 to 88% in 2013.31 Results from this study also found that younger groups had more tolerant views. The favourable conditions of Spain both in legal and tolerance terms might increase immigration of SSP from other countries.31,32

This study has some limitations. First, we only included participants living with their partner. Previous studies have used this approach as a proxy for sexual orientation in population surveys.20 This approach fails to include respondents not living with a partner and is insensitive to bisexuality or other reasons for being in a relationship. People living with SSP might be less likely to disclose such information leading to lower associations. Second, transgender people were not included in the study. Third, the institutionalized population was excluded which might have a different smoking prevalence. Fourth, all measures were self-reported. Measurement of active and passive smoking was one of the many aims of the two surveys, and therefore, using biomarkers such as carbon monoxide would have been impractical. Fifth, causality may not be inferred given the cross-sectional design. Sixth, possible non-response bias should be taken into account given that response rates were 61% and 71% in the 2011 and 2014 surveys. Seventh, the sample size of people living with SSP was small despite combining samples from the two surveys affecting statistical power. The consequence of this small sample size of people living with SSP is that estimates for this group have relatively large margins of error.33 Therefore, the sample is useful and a good first approach, but far from ideal, for the purpose of studying sexual minorities. Moreover, the small sample size does not allow meaningful analysis of subgroups based on sociodemographic or other characteristics.

Research is needed to ascertain the mechanisms explaining the association between sexual orientation and smoking including longitudinal studies, mediation analysis and research on exposure of LGBT people to tobacco industry campaigns. Spain and other countries need to include items on different domains of sexual orientation (behaviour, attraction and identity) in national adult and youth surveys to improve health disparities monitoring and to implement policies and interventions accordingly. Several methods have been suggested to decrease reluctance of some LGBT participants to identify themselves to researchers and to obtain quality samples with relatively small populations.33 A particularly interesting approach is oversampling sexual minorities in population surveys to allow richer analyses. Meanwhile, countries conducting surveys that gather data on the household members’ gender and their relation to the respondent may conduct analyses like the present study. Tobacco control strategies including raising tobacco taxes should be combined with efforts to tackle disparities among sexual minority populations including prevention of discrimination and homophobia or prevention of targeting by the tobacco industry and anti-tobacco campaigns in the general and LGBT media.5,34 Anti-tobacco messages should include the harms caused by second hand smoke to themselves, to their family, friends and other people around them and take into account the potential role of gender.9 The high association between people living with a SSP and passive smoking suggests that prevention and smoking cessation efforts addressed at same-sex couples might be especially effective. This approach is similar to approaches suggested to prevent sexually transmitted infections among stable same-sex couples.35

We found that current active and passive smoking were more prevalent among Spanish adults living with SSP compared to those living with OSP. Associations were especially strong among women living with SSP. The Spanish government should make further efforts to monitor and implement initiatives to improve the health of sexual minorities as part of their plans to tackle health disparities.

Carlos Álvarez Dardet.

The corresponding author on behalf of the other authors guarantee the accuracy, transparency and honesty of the data and information contained in the study, that no relevant information has been omitted and that all discrepancies between authors have been adequately resolved and described.

J. Perales was involved in the conception and design of the work and carried out the analysis. J. Perales, I. Checa and B. Espejo were involved in the interpretation of data. The first version of the manuscript was written by J. Perales and was subsequently improved by I. Checa and B. Espejo, with important intellectual contributions. All authors have approved the final version and are jointly responsible for adequate revision and discussion of all aspects included in the manuscript.

None.

None.