Sexual minorities are at elevated risk of cardiovascular disease from a younger age than heterosexuals

Abstract

Background:

Cardiovascular disease is the leading cause of death worldwide. In this study, we assessed factors related to cardiovascular disease risk and outcomes among sexual minorities (SM).

Methods:

Data from multiple waves of the PATH study were used in this analysis. Multivariable regression models were used to assess the association between sexual identity and: tobacco or e-cigarette use, adverse cardiovascular events, and age at first diagnosis of adverse cardiovascular disease events.

Results:

In our sample (N=23,205), 1,660 (7.15%) participants identified as SM. SM men, relative to heterosexual men, are more likely to be diagnosed with high blood pressure (aRR=1.27; 95% CI: 1.10, 1.47), high cholesterol (aRR=1.32; 95% CI: 1.12, 1.55), congestive heart failure (aRR=2.29; 95% CI: 1.13, 4.65), stroke (aRR=2.39; 95% CI: 1.14, 5.04), heart attack (aRR=2.40; 95% CI: 1.42, 4.04), and other heart conditions (aRR=1.52; 95% CI: 1.06, 2.18). Although no simple differences were observed among SM women compared to heterosexual women, SM women were more likely to be diagnosed at a younger age for high blood pressure (aRR=−0.69; 95% CI: −1.08, −0.29), high cholesterol (aRR=−0.77; 95% CI: −1.15, −0.38), stroke (aRR=−1.04; 95% CI: −1.94, −0.13), and heart attack (aRR= −1.26; 95% CI: −2.42, −0.10). SM men were only diagnosed at a younger age for stroke (aRR=−1.18; 95% CI: −2.06, −0.30).

Discussion:

Compared to heterosexuals, sexual minorities are at higher risk for cardiovascular disease, more likely to develop cardiovascular disease at an earlier age, and more likely to use tobacco products. Future research should focus on decreasing cardiovascular risk among sexual minorities including reducing tobacco use and stress. Screening recommendations for sexual minority populations should also be reviewed in light of a growing body of literature suggesting elevated risk from a young age.

Introduction

Cardiovascular disease (CVD) is the leading cause of death in the United States (US),¹ responsible for 647,457 deaths in 2017.² Substantial demographic differences exist in risk for CVD. For example, it is well established that men are generally diagnosed with a host of CVD outcomes at a younger age than women although women are at greater overall risk of specific conditions (e.g., stroke).^3,4 Less work, however, has examined these differences among sexual minority (SM) populations. The limited body of work in this area has noted an elevated burden of CVD-related risk factors including inflammation, elevated diastolic blood pressure, and pulse rate.^5–10 More recent research has noted that, among young men who have sex with men and transgender women, systemic inflammation (measured by CRP) is significantly elevated regardless of HIV status.¹¹ Still further evidence suggests higher rates of CVD-related risk factors across a broad variety of categories, including psychosocial factors, drug, alcohol, and tobacco use,^12–14 and sexual minority-specific stressors.^13,15

Minority stress theory suggests that sexual and gender minority individuals experience high levels of internalized and externalized stigma,^16–19 including prejudice¹⁶ and discrimination.^16,18–20 They are also more likely than their cisgender heterosexual counterparts to experience multiple intersecting disadvantages, such as experiences of violence^21,22 and mental illness.^18,20,22 Each of these experiences individually increases stress with evidence suggesting that these experiences can have a synergistic effect on stress, health, and well-being.^21,22 For example, one pathway through which experiences of minority stress may affect physical health is the sympathetic nervous system and the hypothalamic-pituitary-adrenal axis.^23–25 As a result of the activation of these systems, biomarkers of inflammation increase which in turn may potentially lead to detrimental downstream effects on cardiovascular health.^26–28 Beyond biologic responses, experiences of stress and stigma among SMs also contribute to CVD-associated negative health behaviors, such as increased rates of tobacco and substance use.^29–31 Taken together, SM may be at increased risk of chronically elevated inflammation and increased risk for cardiovascular disease from a young age relative to heterosexuals.

While research regarding differences between SM and heterosexuals may be lacking, differences between men and women have long been examined in CVD research and literature. For example, although research has long considered CVD to be a larger problem for men given their higher rates of poor CVD-related outcomes,^32–34 more recent work has focused on specific areas of high risk among women such as acute coronary syndrome.³⁵ Where research continues to fall behind though is the intersection between sex and SM status, an area of research which may help provide better population-specific targets for CVD interventions. And though few nationally representative studies assess differences between sex assigned at birth and current gender identity, we aim to begin to fill this gap in the literature by assessing whether differences between SM and heterosexuals may differ based between men and women in the PATH study.

Each of these past studies are limited, however, as the focus of the vast majority is minority stress theory itself and not specifically CVD. Those that go further often examine only behavioral risk factors for CVD (e.g., smoking status, diet) or individual biomarkers of inflammation.^7,11,36 Of those that do address individual CVD-related outcomes, several key papers use a single data source, the Behavioral Risk Factor Surveillance System without replicating finding in other sources,^37–40 are studies using older data from nearly a decade ago,^41,42 or focus on special populations (e.g., adolescents).^43,44 Here, we will address these gaps by in the literature by replicating past findings in new dataset among a general, non-specialized cohort of adults. This will enhance currently literature by comparing findings between nationally representative datasets, will provide more up-to-date data and analyses, and will allow for a broader life course perspective of CVD outcome comparisons between sexual minorities and heterosexuals. To achieve these goals we aimed to first examine rates of cigarette and e-cigarette use among SM relative to heterosexuals in an attempt to replicate the findings of past work. Next, we assessed several measures of heart health (high blood pressure, high cholesterol, congestive heart failure, stroke, heart attack, and other heart conditions) for differences based on sexual identity and gender in a dataset previously underutilized to study SMs. Finally, we aimed to fill a key gap in the literature by examining differences in age at first diagnosis by a medical professional for each of the aforementioned conditions.

Methods

Study Population.

Data come from Wave 1 through Wave 3 of the PATH study. The PATH study is a nationally representative, longitudinal cohort study of non-institutionalized adults, young adults, adolescents, and youth in the U.S. who are at least 12 years of age. The study is a collaborative effort between the National Institutes of Health (NIH), National Institute of Drug Abuse (NIDA), and the U.S. Food and Drug Administration’s Center for Tobacco Products and is administered by Westat. Computer-assisted self-interviews, in both English and Spanish, are utilized to collect information on patterns related primarily to tobacco use but also include measures of other licit and illicit substance use, mental health, and general health. Participants are recruited using address-based, area-probability sampling methods with in-person household screeners. Young adults (aged 18–24), African Americans, and adult tobacco users were oversampled. Weighting procedures are utilized to account for oversampling and non-response among participants and to produce representative estimates of the U.S. population. Informed consent is provided by all participants prior to study participation and was approved by an interview administrator, Westat’s Institutional Review Board. Further details regarding the study procedures can be obtained online.⁴⁵ The final analytic sample was limited to adults (aged ≥18 years) as questions regarding participant sexual identity were not assessed among minors. Additionally, gender identity was not included in the public-use files thus these analyses are limited to sexual minorities only.

Sexual Identity.

Our key variable of interest in this analysis was sexual identity, specifically, health disparities that exist between sexual minorities and heterosexuals. Sexual identity was asked as follows, “Do you think of yourself as…?”, and coded as either heterosexual or sexual minority in the public-use dataset provided by the PATH study. Further disaggregation of sexual identity was not possible.

Health Outcomes.

Tobacco use, either cigarette or e-cigarettes, was self-reported at baseline and operationalized as those who currently use these products and those who do not. Adverse CVD events assessed in the PATH Study were diagnosed by a doctor or other health professional and included: high blood pressure, high cholesterol, congestive heart failure, stroke, heart attack, or any other heart condition. For each, data were assessed at Wave 1 as having been diagnosed in one’s lifetime while Waves 2 and 3 assessed past twelve-month diagnoses. For this reason, the analytic sample was further limited to participants who had completed each of the first three study visits. The variable was operationalized as either having been diagnosed or not diagnosed for each condition as of the completion of Wave 3. Meanwhile, among those diagnosed for each condition, age at diagnosis for each was asked in the following manner, “Age range when you were first told you had [condition]” with options being < 18 years of age, 18 to 24 years, 25 to 34 years, 35 to 44 years, 45 to 54 years, or 55 years or older.

Covariates.

Demographic information was self-reported by participants and included age, sex, race, ethnicity, sexual identity, highest level of education, and yearly income. To prevent potential identification of participants, age is defined as a categorical variable: 18–24, 25–34, 35–44, 45–54, 55–64, 65–74, and 75 years or older. Sex was reported as either male or female. Race was coded based on participant self-identification and coded as those who selected only the white option (“White Alone”), only the black option (“Black Alone”), and those who selected any other or multiple boxes (“Other/Multiracial”). Ethnicity was reported and dichotomized as Hispanic or not Hispanic. Education was self-reported as highest level at the time of the survey: less than high school, high school graduate or GED recipient, some college education, bachelor’s degree, and master’s/doctoral degree. Possession of any health insurance at time of interview (e.g., private insurance, Medicare, Medicaid, Tricare, etc.) was operationalized as a dichotomous variable. All categorical measures were pre-defined as part of the public-use dataset provided by the PATH Study.

Statistical analyses.

Multivariable Poisson regression models with log link functions were utilized to assess the association at baseline between sexual identity and: 1) tobacco or e-cigarette use; and 2) adverse CVD events. Multivariable linear regression models were utilized to assess the association between sexual identity and age at first diagnosis of adverse CVD events. These models were used in place of ordinal logistic regression to facilitate ease of data interpretation, although sensitivity analyses were conducted to ensure consistency of findings. All models were stratified by sex and adjusted for demographic characteristics and were weighted to account for the PATH Study’s stratified cluster sampling design. Statistical significance was established at alpha <0.05. All analyses were performed in Stata 14.0.

Results

At baseline, the analytic sample (N = 23,205) was predominantly between the ages of 18 and 45 (n = 14,516; 62.6%), female (n = 11,866; 51.17%); white alone (n = 16,701; 73.71%), Non-Hispanic (n = 18,975; 82.88%), had some college education or greater (n = 13,329; 57.4%), and had health insurance at the time of survey completion (n = 18,938; 81.61%). See Table 1. Slightly over one-third of the sample reported use of cigarettes (n = 9,569; 41.26%) and one-tenth use of e-cigarettes (n = 2,594; 11.18%). Population statistics are further stratified by sexual identity in Table 1.

Table 1.

Baseline demographic characteristics of participants in the analytic sample, PATH Study

	Total		Heterosexuals		Sexual Minorities
	(N = 23,205)		(n = 21,545)		(n = 1,660)
Variable	n	%	n	%	n	%
Age, years 1
18 – 24	6,409	27.62	5,713	26.52	696	41.93
25 – 34	4,499	19.39	4,110	19.08	389	23.43
35 – 44	3,608	15.55	3,384	15.71	224	13.49
45 – 54	3,613	15.57	3,429	15.92	184	11.08
55 – 64	2,939	12.67	2,818	13.08	121	7.29
65 – 74	1,515	6.53	1,479	6.87	26	2.17
75 or older	618	2.66	608	2.82	10	0.60
Sex
Male	11,323	48.83	10,781	50.01	542	32.71
Female	11,866	51.17	10,751	49.93	1,115	67.29
Race
White Alone	16,701	73.71	15,552	73.85	1,149	71.81
Black Alone	3,697	16.32	3,430	16.29	267	16.69
Other/Multiracial	2,260	9.97	2,076	9.86	184	11.50
Ethnicity
Hispanic	3,919	17.12	3,561	16.75	358	21.92
Not Hispanic	18,975	82.88	17,700	83.25	1,275	78.08
Education
Less than high school	2,973	12.84	2,724	12.68	249	15.01
HS/GED	6,846	29.57	6,343	29.52	503	30.32
Some college	8,228	35.55	7,637	35.54	591	35.62
Bachelor’s Degree	3,316	14.33	3,102	14.44	214	12.90
Master’s Degree or Doctorate	1,785	7.71	1,683	7.83	102	6.15
Health Insurance, any	18,938	81.61	17,665	81.99	1,273	76.69
Cardiovascular Disease
High Blood Pressure	5,262	22.72	4,984	23.18	278	16.83
High Cholesterol	4,177	18.04	3,961	18.42	216	13.08
Congestive Heart Failure	321	1.39	299	1.39	22	1.33
Stroke	415	1.79	391	1.82	24	1.45
Heart Attack	449	1.94	420	1.95	29	1.76
Other Heart Condition	1,209	5.22	1,126	5.24	83	5.02
Tobacco Use
Cigarettes	9,569	41.26	8,716	40.47	853	51.45
E-Cigarettes	2,594	11.18	2,313	10.74	281	16.93

¹Analytic sample includes only those individuals over the age of 18 as minors did not receive sexual orientation-related questions

²PATH Study public use files include only combined sexual identity variables, sexual minorities include those identifying as lesbian, gay, or bisexual

Table 2 describes weighted Poisson regression models assessing the relationship between sexual identity and use of either cigarettes or e-cigarettes, stratified by sex. Compared to heterosexual women, sexual minority women were significantly more likely to use both cigarettes (Adjusted Risk Ratio [aRR] = 1.56; 95% Confidence Interval [CI]: 1.40, 1.75) and e-cigarettes (aRR = 2.23; 95% CI: 1.87, 2.66), adjusting for demographic characteristics. Sexual minority men, relative to heterosexual men, were more likely to use cigarettes (aRR = 1.18; 95% CI: 1.01, 1.37) but not e-cigarettes.

Table 2.

Multivariable weighted Poisson regression models examining the association between sexual identity, cigarette use, and e-cigarette use at baseline, PATH Study

	Sexual Minority Women1		Sexual Minority Men1
Tobacco Use	aRR2	95% CI	aRR2	95% CI
Cigarettes	1.56***	1.40, 1.75	1.18*	1.01, 1.37
E-Cigarettes	2.23***	1.87, 2.66	1.28	0.98, 1.66

^*p ≤ 0.05

^***p ≤ 0.001

Abbreviations: aRR = adjusted risk ratio; CI = confidence interval

¹Reference group: heterosexual adult women and men, respectively

²Adjusted for: sexual identity, age, race, ethnicity, education level, and insurance status

Next, weighted Poisson regression models were utilized to examine the association between sexual identity and various heart conditions, stratified by sex (Table 3). In addition to demographic variables, all models also adjusted for BMI. Relative to heterosexual men, sexual minority men had significantly higher likelihood of having been diagnosed with high blood pressure (aRR = 1.27; 95% CI: 1.10, 1.47), high cholesterol (aRR = 1.32; 95% CI: 1.12, 1.55), congestive heart failure (aRR = 2.29; 95% CI: 1.13, 4.65), stroke (aRR = 2.39; 95% CI: 1.14, 5.04), heart attack (aRR = 2.40; 95% CI: 1.42, 4.04), and other heart conditions (aOR = 1.52; 95% CI: 1.06, 2.18). No significant differences were observed among sexual minority women relative to heterosexual women.

Table 3.

Multivariable weighted Poisson regression models examining the association between sexual identity and various heart conditions across first three waves, PATH Study

	Sexual Minority Women1		Sexual Minority Men1
Heart Condition	aRR2	95% CI	aRR2	95% CI
High blood pressure	1.05	0.90, 1.24	1.27***	1.10, 1.47
High cholesterol	1.17	0.98, 1.40	1.32***	1.12, 1.55
Congestive Heart Failure	1.33	0.71, 2.49	2.29*	1.13, 4.65
Stroke	1.26	0.72, 2.22	2.39*	1.14, 5.04
Heart Attack	1.30	0.62, 2.72	2.40***	1.42, 4.04
Other Heart Condition	1.23	0.91, 1.68	1.52*	1.06, 2.18

^*p ≤ 0.05;

^***p ≤ 0.001

Abbreviations: aRR = adjusted risk ratio; CI = confidence interval

¹Reference group: heterosexual adult women and men, respectively

²Adjusted for: sexual identity, age, race, ethnicity, education level, insurance status, and BMI

Weighted multivariable linear regression models were subsequently used to assess differences in age at first diagnoses for each condition and whether any differences existed based on sexual identity and sex (Table 4). Sexual minority women, compared to heterosexual women, were significantly more likely to be diagnosed at younger ages with high blood pressure (Adj. Coef. = −0.69; 95% CI: −1.08, −0.29), interpreted as a higher likelihood of diagnosis for SM women in the 35–44 age category while heterosexual women were diagnosed in the 45–54 age category. Similar risk of diagnosis at a younger age was observed among sexual minority women with regards to high cholesterol (Adj. Coef. = −0.77; 95% CI: −0.15, −0.38), stroke (Adj. Coef. = −1.04; 95% CI: −1.94, −0.13), and heart attack (Adj. Coef. = −1.26; 95% CI: −2.42, −0.10). Respectively, this indicates that sexual minority women were more likely to be diagnosed for high cholesterol closer to age 35 while heterosexual women were diagnosed closer to age 45, stroke in the 35–44 age category compared to heterosexual women in the 45–54 age category, and heart attack in the 45–54 age category relative to heterosexual women in the 55–64 age category.

Table 4.

Multivariable weighted linear regression models examining the association between sexual identity and age at first diagnosis of heart conditions, PATH Study

	Sexual Minority Women1		Sexual Minority Men1
Heart Condition, Age at Diagnosis	Adj. Coef.2	95% CI	Adj. Coef.2	95% CI
High blood pressure	−0.69***	−1.08, −0.29	−0.36	−0.74, 0.03
High cholesterol	−0.77***	−1.15, −0.38	−0.26	−0.63, 0.11
Congestive Heart Failure	0.29	−0.71, 1.29	−1.47	−3.27, 0.34
Stroke	−1.04*	−1.94, −0.13	−1.18**	−2.06, −0.30
Heart Attack	−1.26*	−2.42, −0.10	−0.64	−1.55, 0.27

^*p ≤ 0.05;

^**p ≤ 0.01;

^***p ≤ 0.001

Abbreviations: Adj. Coef. = adjusted coefficient; CI = confidence interval

¹Reference group: heterosexual adult women and men, respectively.

²Adjusted for: sexual identity, race, ethnicity, education level, insurance status, and BMI

Among sexual minority men, the only age differences observed occurred for stroke where sexual minority men were diagnosed at significantly younger ages relative to heterosexual men (Adj. Coef. = −1.18; 95% CI: −2.06, −0.30), indicating that they are more likely to be diagnosed at age 45–54 compared to 55–64 for heterosexual men.

Discussion

Our findings from this nationally representative, longitudinal cohort study contribute to the current literature on CVD risk among sexual minorities by providing a more nuanced set of results. We observed elevated risk for SM men relative to heterosexual men for high blood pressure, high cholesterol, congestive heart failure, heart attack, and other heart conditions. No differences were noted between SM and heterosexual women. We did observe, however, that SM women were much more likely to be diagnosed at younger ages for all conditions except for congestive heart failure while SM men where only diagnosed at younger ages for stroke. The results presented here suggest that not only do broad gender-based differences persist among SM populations relative to their heterosexual counterparts but also that the risk begins at a much younger age SM men.

There is a growing body of literature that links minority stress, defined as chronic stress caused by multilevel experiences of stigma and discrimination related to sexual identity,⁴⁶ to physical health disparities among SM.^47–49 This represents a potentially key mechanism through which elevated risk for CVD is observed among this population. For example, minority stress throughout the life course may lead to repeated and prolonged activation of physiological stress responses, leading to increased inflammation,^23–25 which increases risk for CVD.^26–28 In fact, past work has demonstrated elevated levels of inflammation among young men who have sex with men and transgender women (aged 16–29), suggesting they may be at elevated risk for poor CVD-related health outcomes from an early age.^7,11 Notably, this past work has also observed elevated levels of inflammation to be independent of HIV infection¹¹ and to be partially mitigated by high levels of marijuana use.⁹ This may help explain why our own findings noted age-based differences only for SM men, they may experience elevated inflammation from a much younger age than previously thought, although future research is needed to assess this hypothesis. Together with the work presented in this study, these results suggest that earlier screening of SM populations for cardiovascular disease may be warranted and updated recommendations should be explored in future research.

Several of our key findings differ between SM men and women. While past research has found that women generally have higher levels of CRP than men, more recent, albeit limited, work has suggested this trend may be reversed among SM men and women.⁴³ One reason for this reversal may be due to differences in gender non-conforming personality traits related to coping mechanisms,⁵⁰ a finding which may lead to gender differences in response to perceived stress and stigma.^51–53 These differential responses have also been reported to influence conditions associated with CVD and systemic inflammation (e.g., depression and anxiety).^54–56 Further, young SM men demonstrate gender non-conforming responses to perceived stress and stigma which may explain their worse immune functioning⁵⁷ and, thus, broader increased risk for diagnosis of CVD relative to heterosexual men observed here. New, broad health initiatives are sorely needed among SM populations to investigate hypotheses such as these and to address the persistent gaps in national studies. Only through comprehensive research among this population will we be able to address and prevent health disparities among SM relative to heterosexuals.

Another possible pathway for elevated CVD risk may be through differences in health seeking behaviors. SM experience discrimination within the healthcare system⁵⁸ that impacts their care seeking behaviors,⁵⁹ and in some cases influences the quality of care they receive.⁶⁰ This is particularly salient among transgender populations with one study noting that two-thirds of transgender individuals report experiencing mistreatment in healthcare settings and forty percent avoided healthcare services in the past twelve months.⁶¹ These avoidant behaviors could lead to missed opportunities to address modifiable risk factors for CVD and to adequately manage medical conditions, such as hyperlipidemia and high blood pressure.⁶²

Tobacco use is an independent risk factor for CVD and approximately 29% of tobacco related deaths are caused by coronary artery disease.⁶³ Although initially thought to be a safer alternative to smoking, e-cigarette users inhale harmful chemicals and nicotine, which are believed to increase CVD risk.⁶⁴ In our analysis, we found that SM are more likely than their heterosexual counterparts to use both tobacco and e-cigarettes. This result supports prior work that shows higher rates of tobacco use among SM^7,31,65,66 and could contribute to the increased risk for CVD among SM.¹⁵ Compounding these rates are experiences of minority stress; SM who experience discrimination are more likely to use tobacco⁶⁵ and smoking is used by some as a coping mechanism⁶⁷ in response to stress. This past work, combined with our own findings, suggests that minority stress contributes to CVD indirectly by influencing tobacco use.

This study has several limitations. First, these data are cross-sectional meaning they cannot assess causal relationships and that confounders in the models may distort the associations between outcome and predictors. Second, the public-use PATH dataset does not differentiate among sexual minority subgroups, participants are categorized as a sexual minority or heterosexual. This has two significant implications. First, this excludes individuals that have same-sex partners but do not identify as a sexual minority. This means it is very possible that SM were underrepresented in the sample, and there could be important differences in stress exposure, smoking, and CVD risk among this population that could influence our results. Second, we were unable to assess differences in CVD risk factors and outcomes among sexual minority subgroups and among those with intersecting identities (e.g., race, ethnicity, age) as the aggregated public-use data does not allow for meaningful, nuanced analyses. Future studies should examine how tobacco use, and CVD disparities evolve over time in SM, especially given that our analysis demonstrate that SM are diagnosed with CVD at younger ages than their heterosexual counterparts.

SM are at increased risk for high blood pressure and other heart conditions and are more likely to be diagnosed with these conditions at an earlier age than their heterosexual counterparts. Future research should assess the role of minority stress early in the life course in the development of premature CVD. Additionally, future research should develop and test multilevel interventions that aim to reduce CVD risk among sexual minorities, including improving experiences within the healthcare system and reducing tobacco use. Finally, CVD screening recommendations for SM populations should be reviewed in light of a growing body of literature suggesting elevated risk from a young age.

Availability of Data and Material

All data are publicly available at https://pathstudyinfo.nih.gov/.