Assessment for Social Support

Lesbian, gay, & bisexual older adults: linking internal minority stressors, chronic health

conditions, and depression

Charles P. Hoy-Ellisa* and Karen I. Fredriksen-Goldsenb

aCollege of Social Work, University of Utah, Salt Lake City, UT, USA; bSchool of Social Work, University of Washington, Seattle, WA, USA

(Received 30 January 2016; accepted 15 March 2016)

Objectives: This study aims to: (1) test whether the minority stressors disclosure of sexual orientation; and (2) internalized heterosexism are predictive of chronic physical health conditions; and (3) depression; (4) to test direct and indirect relationships between these variables; and (5) whether chronic physical health conditions are further predictive of depression, net of disclosure of sexual orientation and internalized heterosexism. Methods: Secondary analysis of national, community-based surveys of 2349 lesbian, gay, and bisexual adults aged 50 and older residing in the US utilizing structural equation modeling. Results: Congruent with minority stress theory, disclosure of sexual orientation is indirectly associated with chronic physical health conditions and depression, mediated by internalized heterosexism with a suppressor effect. Internalized heterosexism is directly associated with chronic physical health conditions and depression, and further indirectly associated with depression mediated by chronic physical health conditions. Finally, chronic physical health conditions have an additional direct relationship with depression, net of other predictor variables. Conclusion: Minority stressors and chronic physical health conditions independently and collectively predict depression, possibly a synergistic effect. Implications for depression among older sexual minority adults are discussed.

Keywords: Sexual orientation; depression; older adults; minority stress; structural equation modeling

Introduction

The World Health Organization (WHO) has characterized

depression as a serious public health issue (World Health

Organization, 2012). Current annual health care expendi-

tures for the treatment of depression in the US alone

exceed $22 billion (Soni, 2012). In addition, the annual

per capita health care costs for older Americans with

depression exceed $20,000, which is more than double the

cost of those who do not (Un€utzer et al., 2009). Untreated depression typically becomes chronic in nature (Chap-

man, Perry, & Strine, 2005; Fiske, Wetherell, & Gatz,

2009), negatively impacting quality of life (Chapman

et al., 2005; Fiske et al., 2009), the treatment of co-occur-

ring chronic physical health conditions (Centers for Dis-

ease Control and Prevention and National Association of

Chronic Disease Directors, 2009), and potentially decreas-

ing life expectancy by 5–10 years (Chapman et al., 2005).

Depression is recognized as the most common, treatable

chronic mental health condition among older adults (Cen-

ters for Disease Control and Prevention, 2015). Popula-

tion-based prevalence estimates of depression among

Americans aged 50 and older in the general population

are typically reported to range from 1% to 5% (Centers

for Disease Control and Prevention, 2015). National Sur-

vey on Drug Use and Health (NSDUH) and Behavioral

Risk Factor Surveillance System (BRFSS) data indicate

prevalences among adults aged 50 and older ranging from

about 6% (Substance Abuse and Mental Health Services

Administration, 2013) to about 8%, respectively (Centers

for Disease Control and Prevention and National Associa-

tion of Chronic Disease Directors, 2009). Clinically sig-

nificant depressive symptomatology among older

community-dwelling adults may be as high as 15% (Fiske

et al., 2009).

Census projections suggest that the number of Ameri-

cans aged 50 and older will grow to more than 130 million

by 2030, and will approach 164 million by 2060 (U.S.

Census Bureau, 2015). Current national estimates suggest

that 2.6–4.9 million of these will self-identify as lesbian,

gay, and bisexual (LGB) (Gates & Newport, 2012). Our

knowledge of the health and well-being of LGB older

adults remains a significant shortcoming in health dispar-

ities research (Centers for Disease Control and Preven-

tion, 2011; Fredriksen-Goldsen, Emlet, et al., 2013). Yet,

LGB Americans aged 50 and older have been found to be

a health disparate population, evidencing higher rates of

poor mental health as well as other physical health prob-

lems than heterosexual older adults (Fredriksen-Goldsen,

Kim, Barkan, Muraco, & Hoy-Ellis, 2013; Wallace,

Cochran, Durazo, & Ford, 2011). In large community-

based samples, 29% of LGB older adults (Fredriksen-

Goldsen, Emlet, et al., 2013) and 47% of transgender

older adults (Fredriksen-Goldsen, Cook-Daniels, et al.,

2013) have been found to have clinically significant

depressive symptomatology. While poor mental health

outcomes among lesbian, gay, bisexual, or transgender

*Corresponding author. Email: Charles.Hoy-Ellis@socwk.utah.edu

� 2016 Informa UK Limited, trading as Taylor & Francis Group

Aging & Mental Health, 2016

Vol. 20, No. 11, 1119–1130, http://dx.doi.org/10.1080/13607863.2016.1168362

(LGBT) older adults are being recognized, the underlying

processes tend to be less understood (Institute of Medi-

cine, 2011). A major goal of the Healthy People 2020 ini-

tiative is to improve the health and well-being of LGB

communities, including reducing the incidence of major

depression among LGB adults as a targeted objective

(U.S. Department of Health and Human Services, 2013).

Meeting this objective will require a better understanding

of depression among LGB older adults so that culturally

responsive intervention and prevention efforts can be

developed and implemented.

Depression is not a part of the normative aging pro-

cess. According to the diathesis-stress perspective, depres-

sion due to genetic diathesis is more common among

younger adults; disruptions resulting from significant life

events and cumulative social, psychological, and biologi-

cal stressors are more likely to result in depression among

older adults (Blazer & Hybels, 2005; Fiske et al., 2009;

Zuckerman, 1999). General stressors that increase the risk

for depression in older adulthood are common to both

LGB and heterosexual older adults. These include finan-

cial challenges, decreased social interactions, social isola-

tion, bereavement, and other negative life events (Fiske

et al., 2009). Numerous chronic medical conditions have

been linked to depression among older adults (Blazer,

2003; Chapman et al., 2005; Fiske et al., 2009; Yang,

2007). Adults in the general population living with

chronic health conditions, particularly those aged 40–

59 years old have a significantly increased risk for devel-

oping depression (Pratt & Brody, 2008). Just under 80%

of Americans aged 50 and older have at least one chronic

health condition (AARP Public Policy Institute, 2010;

Centers for Disease Control and Prevention, 2013).

Chronic health conditions most often associated with

depression include asthma, arthritis, cardiovascular dis-

ease (CVD), diabetes, and obesity (Chapman et al., 2005;

Fiske et al., 2009). Emerging evidence indicates that com-

pared to their heterosexual counterparts, LGB adults aged

50 and older are also at heightened risk for a variety of

chronic physical health conditions, including CVD, obe-

sity, and asthma among sexual minority women (Fredrik-

sen-Goldsen, Kim, et al., 2013), and hypertension and

diabetes among sexual minority men (Wallace et al.,

2011). These conditions are among the most prevalent

associated with increased risk of developing or exacerbat-

ing the course of depressive disorders (Chapman et al.,

2005; Fiske et al., 2009).

LGB older adults also experience additional stressors

unique to their sexual orientation, which stem from living

in a heterosexist society and are theorized to contribute to

their ‘excess’ rates of depression (Centers for Disease

Control and Prevention, 2013). Heterosexism can be

described as the collective constellation of societal preju-

dice, attitudes, stereotypes, and beliefs that cast heterosex-

uality as normative and any other form of human sexual

identity, attraction, and/or behavior as abnormal (Herek &

Garnets, 2007). The minority stress model identifies pro-

cesses by which heterosexist-related minority stressors

negatively impact the mental health of LGB people

(Meyer, 2003). Internals of minority stressors, internal-

ized heterosexism and concealment of sexual orientation,

are the most chronic and inescapable (Meyer, 2003) and,

thus, may play a crucial role in heightened risk for depres-

sion among older LGB adults. Internalized heterosexism

refers to early and ongoing socialization processes by

which people internalize society’s prejudicial attitudes,

stereotypes, and beliefs regarding non-heterosexuality.

Consciously and unconsciously, LGB people may apply

such internalized representations to themselves and to

other LGB people (Meyer, 2003). Internalized heterosex-

ism has been associated with increased risk for depression

among LGB older adults (Fredriksen-Goldsen, Emlet,

et al., 2013).

Self-concealment of personal information and secrets

of a distressing nature have been consistently linked to

physiological symptoms in the general population (Uysal,

Lin, & Knee, 2010). Concealing one’s non-heterosexual

orientation may provide a degree of short-term protection

by making oneself a less visible target for victimization,

but continued concealment over time is psychologically

stressful (Meyer, 2003), negatively impacting neuroendo-

crine functioning (Meyer, 2003) associated with the

development of chronic health conditions (Cole, Kemeny,

Taylor, & Visscher, 1996). A sample of HIV-negative gay

men in the Natural History of AIDS Psychosocial Study

who concealed their sexual orientation developed cancer

at significantly higher rates relative to gay men who dis-

closed their sexual orientation (Cole et al., 1996). Recent

epigenetic research has identified chronic stress as playing

a role in the expression of the ATF3 gene in breast cancer

metastasis (Wolford et al., 2013). Alternately, disclosure

of one’s LGB sexual orientation is posited to counteract

the negative impacts of chronic minority stress by provid-

ing individual and group-level coping resources (Meyer,

2003). Research findings regarding the role of conceal-

ment and disclosure of sexual orientation and risk of

depression among older LGB adults have been mixed.

Data from the Urban Men’s Health Study (UMHS) indi-

cated that disclosure is associated with greater risk for

depression among gay men aged 50–59, but not for those

aged 60 and older (Rawls, 2004). Another study found

that disclosure of sexual orientation among older LGB

adults is associated with lower levels of depression, but

that relationship is indirectly working through internalized

heterosexism (Hoy-Ellis, 2015). Yet, a different study

found no relationship between concealment or disclosure

of sexual orientation and depression, when controlling for

demographic characteristics and other risk and protective

factors (Fredriksen-Goldsen, Emlet, et al., 2013).

The significance of the current study is that it exam-

ines the relative roles of the most internal of minority

stressors, internalized heterosexism and concealment or

disclosure of sexual orientation, and chronic health condi-

tions in depression among older LGB adults. It also seeks

to explore if disparities in certain chronic physical health

conditions identified in this population may contribute to

disparities in poor mental health. Specifically, this study

aims to test the following hypothesized relationships:

(1) Disclosure of sexual orientation is directly and

inversely related to internalized heterosexism,

chronic health conditions, and depression.

1120 C. P. Hoy-Ellis and K. I. Fredriksen-Goldsen

(2) Disclosure of sexual orientation is inversely and

indirectly associated with chronic health condi-

tions and depression through internalized

heterosexism.

(3) Internalized heterosexism is directly and posi-

tively related to chronic physical health condi-

tions and depression.

(4) Internalized heterosexism is positively and indi-

rectly associated with depression via chronic

physical health conditions.

(5) Chronic physical health conditions have an addi-

tional positive relationship with depression among

LGB older adults, net of disclosure of sexual ori-

entation and internalized heterosexism (see

Figure 1 for model to be tested).

Methods

Sample and procedure

This study is a secondary analysis of data from the

National Health, Aging, & Sexuality Study: Caring &

Aging with Pride Over Time (NHAS), the first of its kind

national study to investigate the health and well-being of

LGB older adults as a population distinct from both their

younger LGB peers and older heterosexual adult counter-

parts. The Institute for Multigenerational Health at the

University of Washington, Seattle, partnered with 11

agencies across the US, which provide programming and

services specific to LGB older adults. A survey was devel-

oped and distributed via agency mailing lists from June

through November of 2010. The survey included ques-

tions to assess standard sociodemographic information, as

well as sexual orientation and gender identity. Also

included in the survey were items particularly relevant to

LGB experience, such as disclosure of sexual orientation

or gender identity, and measures of physical and mental

health. Inclusion criteria for the NHAS required that (1)

potential participants be 50 years old or older at the time

of the survey distribution and (2) self-identify as LGBT.

Along with standard informed consent and anonymity

protocols, participants were offered an opportunity to

enter a raffle to win one of five $500 gift cards for their

time, winners to be chosen randomly. The University of

Washington Institutional Review Board approved all

study materials, procedures, and safeguards for the protec-

tion of human participants; many partnering agencies con-

ducted their own internal reviews. The final dataset was

comprised of surveys completed by 2560 LGBT adults

aged 50–95 years old. For a fuller description of the

NHAS, see Fredriksen-Goldsen, Kim and associates

(2013).

The sample for the current study (n D 2349) consisted of 829 self-identified bisexual and lesbian women (35%)

and 1520 bisexual and gay men. Transgender participants

were excluded and studied elsewhere. Sample participants

ranged in age from 50 to 95 years old (M D 66.9; SD D 9.0), most identified as lesbian or gay (95%), and were

Figure 1. Structural equation model to be tested. Note: Model showing direct and indirect relationships between latent variables concealment and internalized heterosexism; and observed variables chronic health conditions and depression.

Aging & Mental Health 1121

predominantly non-Hispanic white (87.0%). Although the

majority (92%) had at least some college education, about

half (52%) reported annual household incomes of

$49,999. See Table 1 for sample sociodemographic

characteristics.

Measures

Covariates income and education were controlled for, as

the robust associations between these variables and

chronic health conditions and depression have been

widely established (Marmot & Wilkinson, 2006; World

Health Organization, 2003). Age was also treated as a

covariate as it has been related to disclosure of sexual ori-

entation and internalized heterosexism (David & Knight,

2008). Annual household income was coded across six

categories: <$20,000; $20,000–$24,999; $25,000–

$34,999; $35,000–$49,999; $50,000–$74,999; and

$75,000 or more. Educational attainment was categorized

as: kindergarten or none; grade 9–11; grade 12 or GED

(General Educational Development Test, a certification

that is equivalent to a high school diploma); college of 1–

3 years; and college of 4 years or more. Age was calcu-

lated from reported year of birth.

A latent variable to assess the degree of disclosure of

the participants’ sexual orientation was constructed from

a modified version of the 12-item Outness Inventory

(Mohr & Fassinger, 2000), which assesses sexual orienta-

tion disclosure in three primary social domains. Partici-

pants indicated the likelihood that family members (e.g.

parent, sibling), communitymembers (e.g. neighbors, faith

community), and a best friend know or have known their

sexual orientation on a 4-point Likert scale (1 D definitely do not know through 4 D definitely do know). Factor anal- yses indicated that the three indicators (out to friend, fam-

ily, community) loaded well onto a single factor (.63–.91,

p < .001). Internal consistency was acceptable,

Cronbach’s a D .71. Higher scores indicate higher levels of disclosure of sexual orientation.

A separate latent variable with five indicators was

constructed to capture internalized heterosexism, utilizing

the Homosexual Self-Stigma subscale (Liu, Feng, & Rho-

des, 2009). Participants indicated their level of agreement

with five statements such as ‘I wish I weren’t lesbian, gay,

bisexual, or transgender’ coded on a 4-point Likert scale

(1 D strongly agree through 4 D strongly disagree). Fac- tor analyses indicated that all five items loaded well onto

a single latent factor (.48–.79, p < .001), with acceptable

internal consistency (Cronbach’s a D .79). Responses were then reverse-coded so that higher scores indicated

higher levels of internalized heterosexism.

Chronic health conditions were treated as an observed

variable based on participants’ endorsement (‘mark all

that apply’) of whether they had ever been told by a physi-

cian that they had any of the following nine chronic health

conditions identified in the literature as being associated

with depression: angina, arthritis, congestive heart fail-

ure, diabetes, heart attack, high cholesterol, hypertension,

osteoporosis, and stroke. A number of conditions were

summed, producing a range of 0–9, with higher numbers

indicating the presence of more chronic health conditions.

Depression was assessed via the Center for Epidemio-

logical Studies Depression Scale 10-item short form

(CESD-10) (Radloff, 1977), which has well-established

validity and reliability in screening for major depression

across populations (Grzywacz, Hovey, Seligman, Arcury,

& Quandt, 2006; Zhang et al., 2012), including among

community-dwelling older adults (Andresen, Malmgren,

Carter, & Patrick, 1994; Boey, 1999; Irwin, Artin, &

Oxman, 1999). Depression was treated as an observed

variable, making for a more parsimonious the model;

model fit decreases as the number of variables increases

(Kenny, 2014). The CESD-10 calls for participants to

indicate how many days during the past week (0 D <1 day, 1 D 1–2 days; 2 D 3–4 days; 3 D 5–7 days) they had felt or acted in certain ways; for example, ‘I felt

depressed,’ and ‘everything I did was an effort.’ Internal

consistency was good, Cronbach’s a D 0.88. On a range of 0–30, a score �10 is an indicator of depressive symp- toms that meet clinically significant levels (Andresen

et al., 1994; Zhang et al., 2012).

Statistical analyses

Structural equation modeling (SEM) using Stata v. 12 was

employed for all analyses. SEM is a confirmatory statisti-

cal technique useful for testing a priori theorized models

(Bollen, 1989). A sample variance–covariance matrix is

computed and compared to an estimated population vari-

ance–covariance matrix; if the difference between the two

matrices is close to zero, the model is considered to be a

good fit to the data (Bollen, 1989). In SEM, the

Table 1. Sample sociodemographic characteristics.

Variable (%) (n)

AgeM (SD) 66.9 (9.0) 2372

Gender

Women 35.4 840

Men 64.6 1531

Sexual orientation

Lesbian/gay 94.6 2217

Bisexual 5.4 124

Race/ethnicity

Hispanic/non-Hispanic, non-white 13.0 343

Non-Hispanic white 87.0 2198

Education

Grade 1–8 0.2 4

Grade 9–11 0.8 19

Grade 12 or GED 6.7 158

College 1–3 years 18.2 427

College 4 years or more 74.2 1744

Annual household income

<$20,000 18.2 399

$20,000–$24,999 8.3 186

$25,000–$34,999 11.7 269

$35,000–$49,999 14.3 329

$50,000–$74,999 17.0 396

$75,000 or more 30.6 721

1122 C. P. Hoy-Ellis and K. I. Fredriksen-Goldsen

measurement model provides information as to how well

indicators load onto latent variables (i.e. confirmatory fac-

tor analysis); the structural model provides information

on the relationships between variables. SEM has some

advantages over more traditional multiple regression tech-

niques. Standard regression models assume ‘perfect meas-

urement’ which produces biased estimates (Baron &

Kenny, 1986); SEM accounts for measurement error (Bol-

len, 1989), and is more sensitive to detecting suppressor

effects (Cheung & Lau, 2008) and mediation effects

(Iacobucci, Saldhana, & Deng, 2007). Total effects can be

decomposed into their direct and indirect components,

allowing inferences about mediation effects to be made

(Duncan, 1975). Because equations are estimated simulta-

neously, standard errors are smaller and more consistent

(Iacobucci et al., 2007).

In this study, the Maximum Likelihood estimator with

pairwise deletion was used for model-testing. The data

were not normally distributed, therefore, bootstrapping,

resampling with replacement (500 replications), was

employed to derive a sampling distribution for more pre-

cise standard errors and accurate confidence intervals (CI)

(Cheung & Lau, 2008). A Variance Inflation Factor (VIF)

was computed to assess for possible issues of multicolli-

nearity, which preliminary analyses indicated was not an

issue; VIF D 1.07, well below the acceptable upper bound of 10 (StataCorp, 2011). Hooper, Coughlan, and Mullen

(2008) recommend assessing an array of post-estimation

goodness-of-fit (GOF) statistics to examine model fit. The

model x2 is typically reported, yet, with very large sample

sizes (i.e. �200); this statistic will almost always be sig- nificant (Matsueda, 2012), requiring rejection of the null

hypothesis. However, a non-significant difference

between the sample and estimated population variance–

covariance matrices is indicative of a good model fit. Of

other test statistics endorsed by Hooper et al. (2008), the

Comparative Fit Index (CFI) is minimally affected by

sample size, thus, addressing the issue of model x2 signifi-

cance. It contrasts the null model against the sample

covariance matrix and calculates a statistic that ranges

from 0 to 1; a value >.90 suggests a good model fit.

Among the most revealing of fit statistics, the Root Mean

Square Error of Approximation (RMSEA) identifies the

closeness of fit between the population covariance matrix

and sample parameters; a value <.06 indicates a good fit

between the model and the data (Hooper et al., 2008). The

Standardized Root Mean Square Residual (SRMR) is a

measure of the difference between the standardized square

root residuals of the sample and hypothesized population

covariance matrices. While an SRMR < .08 is considered

adequate, a value<.05 suggests a better model fit (Hooper

et al., 2008). In addition, a CI close to zero implies that the

sample and hypothesized population covariance matrices

do not differ significantly.

Results

Overall, 29% of the sample (n D 666) reported clinical symptoms that met the threshold of major depression,

scoring �10 on the CESD-10 (M D 7.2, SD D 6.2). The

average level of disclosure, 3.5 on a scale of 1–4 (SD D .6) was relatively high, and the mean level of internalized

heterosexism, 1.5 on a scale of 1–4 (SD D .6) was rela- tively low. Participants had on average 1.9 chronic health

conditions (SD D 1.4). See Table 2 for sample summary statistics and distributions of chronic health conditions.

To further assess model fit, a Lagrange Multiplier Test

to detect omitted paths and provide estimates of change in

model fit was conducted. Adding omitted paths is method-

ologically sound, provided that such additions are consis-

tent with theory (StataCorp, 2011). Correlated error term

paths were added (not shown), which is theoretically

sound as indicators of observed measures are themselves

typically correlated (see Table 3 for correlation matrix).

The final fitted model is shown in Figure 2. With the

exception of the x2-statistic, post-estimation GOF test sta-

tistics separately and collectively suggest a very close fit

of the model to the data (see Table 4).

Factor loadings and path coefficients in Figure 2 are

standardized to facilitate interpretation of relationships

and effect sizes (Preacher & Kelley, 2011). Initial results

initially indicated that disclosure of sexual orientation did

not appear to have a significant association with either

depression (p D .089) or chronic health conditions (p D .679). However, decomposition of total effects into their

direct and indirect components (see Table 5) suggests that

the indirect effect of disclosure is significantly related to

both depression (p < .001) and chronic health conditions

(pD .030). Indirect effects may be significant even though direct and total effects are not, such as the case when the

indirect effect has an opposite sign, which may indicate

that the mediating variable (i.e. internalized heterosexism)

also acts as a suppressor, strengthening or weakening the

effect of the independent variable on the dependent vari-

able, thereby, obscuring the total effect (Rucker, Preacher,

Tormala, & Petty, 2011). Opposite signs of the indirect

coefficients are seen in Table 5. These relationships are in

line with minority stress theory in that disclosure of sexual

orientation decreases the stressful effects if internalized

heterosexism (Meyer, 2003), which in turn, would attenu-

ate the positive associations between internalized hetero-

sexism with depression and chronic health conditions.

Significant direct positive associations were found

between internalized heterosexism and both depression

Table 2. Sample summary statistics and distribution of chronic health conditions.

Variable Range M (SD) Chronic conditions (%) (n)

Disclose to friend 3.9 (0.6) Angina 3.9 92

Disclose to family 1–4 3.4 (0.8) Arthritis 33.8 802

Disclose to community 3.5 (0.7) Congestive heart failure

2.7 63

Disclosure overall 3.5 (0.6) Diabetes 13.7 324

Internalized heterosexism 1–4 1.5 (0.6) Heart attack 5.6 132

Chronic health conditions 0–9 1.9 (1.4) High cholesterol 43.3 1027

Depression (CESD) 0–30 7.2 (6.2) Hypertension 45.5 1079

CESD � 10 29.2% n D 666 Osteoporosis 10.2 243 Stroke 3.9 92

Aging & Mental Health 1123

and chronic health conditions, as well as an additional

indirect association with depression via chronic health

conditions; chronic health conditions have an additional

positive direct association with depression (see Table 5).

The cumulative direct, indirect, and total effects of con-

cealment of sexual orientation, internalized heterosexism,

and chronic health conditions indicate that these variables

account for just under 76% of the variance in depression.

Discussion

Emerging research suggests that LGB older adults have a

significantly greater risk for depression and several

chronic health conditions (Fredriksen-Goldsen, Kim,

et al., 2013; Valanis et al., 2000; Wallace et al., 2011).

Concealment of sexual orientation (Hoy-Ellis, 2015) and

internalized heterosexism may increase the risk for

Figure 2. Fitted structural equation model. Note: Showing direct and indirect relationships between latent variables concealment and internalized heterosexism; and observed varia- bles chronic health conditions and depression. Factor loadings and path coefficients are standardized. �p < .05. ��p < .01. ���p < .001.

Table 3. Correlations of observed measures.

Disclosure (D) Internalized heterosexism (IH)

Family Friend Community A B C D E Chronic CESD Age Income Education

D-family 1.00

D-friend .38 1.00

D-community .49 .45 1.00

IH-A ¡.18 ¡.11 ¡.23 1.00 IH-B ¡.11 ¡.06 ¡.09 .39 1.00 IH-C ¡.17 ¡.13 ¡.20 .71 .09 1.00 IH-D ¡.19 ¡.14 ¡.22 .60 .37 .59 1.00 IH-E ¡.13 ¡.08 ¡.14 .38 .26 .41 .53 1.00 Chronic ¡.08 ¡.04 ¡.05 .07 .04 .06 .08 .04 1.00 CESD ¡.04 ¡.06 ¡.05 .18 .09 .14 .20 .11 .18 1.00 Age ¡.31 ¡.12 ¡.16 .11 .02 .06 .11 .06 .22 ¡.02 1.00 Income .13 .10 .14 ¡.10 .02 ¡.05 ¡.13 ¡.07 ¡.17 ¡.31 ¡.17 1.00 Education .07 .10 .10 ¡.04 .04 ¡.01 ¡.07 ¡.05 ¡.12 ¡.16 ¡.07 .36 1.00

1124 C. P. Hoy-Ellis and K. I. Fredriksen-Goldsen

depression (Fredriksen-Goldsen, Emlet, et al., 2013; Hoy-

Ellis, 2015) among LGB older adults (Fredriksen-Gold-

sen, Emlet, et al., 2013). The results reported here suggest

that disparities in chronic health conditions documented

among LGB older adults may explain some of the dispar-

ity in their rates of depression, aligning with research in

the general older adult population linking chronic health

conditions with increased risk for depression (Blazer &

Hybels, 2005; Chapman et al., 2005; Fiske et al., 2009).

Findings also provide additional evidence that minority

stressors are cumulative in their effects on mental health

outcomes (Meyer, 2003), and that pathways of risk are

complex and may be obscured (Institute of Medicine,

2011). Disclosure of sexual orientation appears to be

related to lower levels of internalized heterosexism,

thereby, reducing the positive associations between both

internalized heterosexism and chronic health conditions

on depression. Internalized heterosexism and chronic

health conditions may have additional impacts on depres-

sion, net of disclosure of sexual orientation, suggesting

that social, psychological, and physical factors be consid-

ered in tandem when examining depression among LGB

older adults.

The finding that higher levels of disclosure of sexual

orientation are inversely related to internalized heterosex-

ism and indirectly with depression mediated by internal-

ized heterosexism is consistent with the minority stress

model. Long-term concealment of a significant aspect of

the self is psychologically costly (Meyer, 2003), which

can be attributed to potential negative consequences of

disclosure, shame, guilt, and distorted thinking that related

to internalized heterosexism (Pachankis, 2007). Through

disclosure of sexual orientation, important individual and

group-level coping processes are activated reducing levels

of internalized heterosexism (Meyer, 2003). When avail-

able, coping resources are deemed to be adequate to meet

perceived threat through secondary appraisals (Lazarus &

Folkman, 1984); the stress response and risk for depres-

sion are significantly diminished (Juster, McEwen, &

Lupien, 2010). Consistent with social comparison theory

(Hogg, Terry, & White, 1995) at the individual level, dis-

closure diminishes feelings of shame and guilt (Pachankis,

2007), and through subsequent positive comparisons of

the self with other LGBs, replacing hitherto negative com-

parisons with heterosexuals, distorted cognitions regard-

ing the self are ameliorated (Meyer, 2003).

The indirect relationship between concealment and

chronic health conditions, mediated via internalized het-

erosexism and the additional direct effect of internalized

heterosexism on both chronic health conditions and

depression, is consistent with social stress theory broadly,

and the minority stress framework in particular. Decades

of social stress research have demonstrated that chronic

psychosocial stressors ‘gets under the skin’ to become

embodied and consequently manifest in chronic disease

(Ferraro & Shippee, 2009; Krieger, 1999), such as CVD,

diabetes (Juster et al., 2010), hypertension, and asthma

(Katon, 2011), particularly among socially marginalized

groups (Aneshensel, 2009). The internalization of stigma

associated with marginalized social status has been char-

acterized as a chronic stressor in and of itself (Hatzen-

buehler, Phelan, & Link, 2013). The hypothalamic-

pituitary-adrenal (HPA) axis is central to neuroendocrine

processes that are activated in response to stressors (Juster

et al., 2010; McEwen, 1998). Cortisol and adrenaline are

primary hormones released in this response process.

When stressors are acute and relatively sporadic, the

release of these hormones may enhance survival. When

stressors are chronic, repeated over-activation of the

Table 4. Model goodness-of-fit statistics.

Statistical test Statistical value

Model x2 (df) 143.64 (42)

Root Mean Square Error of Approximation (RMSEA)

0.035

Confidence interval (CI) (90%) [.029, .042]

Comparative Fit Index (CFI) 0.981

Standardized Root Mean Square Residual (SRMR)

0.023

Coefficient of determination (CD) (model R2) 0.757

Table 5. Decomposition of total, direct, and indirect effects.

Depression

b� se p > z b� se p > z b� se p > z Direct Indirect Total

Disclosure .013 .326 .683 ¡.064 .168 <.001 ¡.051 .309 .089 Internalized heterosexism .186 .418 <.001 .009 .050 .022 .195 .424 <.001

Chronic health conditions .143 .103 <.001 (No path) .143 .103 <.001

Internalized heterosexism

Disclosure ¡.354 .048 <.001 (No path) ¡.354 .048 <.001

Chronic health conditions

Disclosure .032 .064 .249 ¡.021 .023 .030 .011 .060 .679 Internalized heterosexism .060 .079 .022 (No path) .060 .079 .022

Note: b� D Standardized coefficient; se D bootstrapped standard error.

Aging & Mental Health 1125

HPA-axis results in allostatic load (AL) (Juster et al.,

2010; McEwen, 1998). Among other negative physiologi-

cal effects, AL has been linked to metabolic dysfunctions

such as hyperlipidemia and insulin resistance, which are

associated with diabetes, hypertension, and CVD (Juster

et al., 2010; McEwen, 1998). Regions of the brain

involved in threat appraisal processes are also negatively

impacted by AL, resulting in decreased perceived coping

resources and increased risk for depression (McEwen,

2006).

Chronic health conditions also have an additional

direct association with depression, net of all other rela-

tionships. Having chronic health conditions increases the

risk for developing depression or exacerbating existent

depression (Chapman et al., 2005; Katon, 2011; Wolko-

witz, Reus, & Mellon, 2011). There is also a direct rela-

tionship between increasing numbers of chronic health

conditions and increased risk of developing or worsening

depression (Chapman et al., 2005). It is, thus, plausible

that the heightened risk of chronic health conditions iden-

tified among LGB older adults (Fredriksen-Goldsen, Kim,

et al., 2013; Wallace et al., 2011) plays an important role

in the disparately high rates of depression documented in

this population. The relationship between chronic health

conditions and depression is also consistent with the

broader social stress literature. LGB older adults are mar-

ginalized both by their sexual orientation and their age

(Fredriksen-Goldsen, Hoy-Ellis, Goldsen, Emlet, &

Hooyman, 2014), resulting in social exclusion and lower

social standing. Findings from the Whitehall studies have

advanced our understanding of the relationship between

lower social standing, chronic health conditions, and poor

mental health outcomes by showing that the underlying

mechanism of risk is decreased control over important

aspects of the social environment that accompanies lower

social standing (Marmot et al., 1991; Marmot & Wilkin-

son, 2006). The presence of chronic health can also limit

control over key aspects of one’s life (Blazer, 2003;

Katon, 2011).

Implications

There is a dearth of research that attends to midlife and

older LGB adults as a population distinct from both mid-

life and older heterosexual adults, and from younger adult

and adolescent sexual minorities. The little research that

has made such comparisons indicates that there are impor-

tant differences between these respective groups (Fredrik-

sen-Goldsen, Kim, et al., 2013; Kertzner, Meyer, Frost, &

Stirratt, 2009; Wallace et al., 2011). Today’s LGB older

adults are more likely to conceal their sexual orientation

than their younger LGB counterparts (Floyd & Bakeman,

2006). Within-group differences by age are also beginning

to emerge. For example, LGB adults aged 50–64 years old

report higher rates of discrimination and victimization

than their counterparts aged 65 and older, yet, the latter

age group evidences higher levels of internalized hetero-

sexism and is more likely to conceal their sexual orienta-

tion than the former (Fredriksen-Goldsen, Kim, Shiu,

Goldsen, & Emlet, 2014). Fearing discrimination by staff,

and harassment and isolation from other clients, even

LGB older adults who are open about their sexual orienta-

tion believe that they will need to conceal their identity in

order to access mainstream aging services – at the very

time when advancing age increases the likelihood of need-

ing such services (National Senior Citizens Law Center,

2011). Yet, these findings suggest that to do so, may place

LGB older adults at increased risk for depression.

This study makes a significant contribution to our

knowledge regarding the health and well-being of older

LGB adults by identifying how minority stress risk factors

and chronic health conditions are associated with each

other and with depression. Identifying that chronic health

conditions play a role in the minority stress process may

enhance our understanding of why rates of depression

remain alarmingly high as LGB individuals get older (Fre-

driksen-Goldsen, Kim, et al., 2013; Wallace et al., 2011),

while rates of depression decline noticeably in the general

population as it ages (Blanchflower & Oswald, 2008;

Blazer, 2003; Yang, 2007). Furthermore, results may also

contribute to clarifying the theoretical relationship

between internal minority stressors of concealing LGB

sexual orientation and internalized heterosexism, and

depression. Identifying and understanding the complex

interactions of minority stress processes as they relate to

health will be central to developing culturally sensitive

and effective interventions for LGB older adults living

with depression.

There is evidence that the relationship between

chronic health conditions and depression is recursive

(Chapman et al., 2005; Katon, 2011; Pinquart & Sorenson,

2007). Many chronic health conditions that begin to mani-

fest around the age of 50 may be rooted in chronic stress

that begins in earlier life experience (Kuzawa & Sweet,

2009; Murgatroyd & Spengler, 2011; Seeman, Singer,

Ryff, Dienberg Love, & Levy-Storms, 2002; Wolkowitz

et al., 2011). The corrosive effects of internalized hetero-

sexism that surfaces earlier in life when one begins to

realize a non-heterosexual orientation would fall squarely

in the category of ‘chronic stress that begins in earlier life

experience.’ The same array of complex neurobiological

patterns found between chronic social stress and HPA-

axis dysregulation and AL is found in the relationship

between chronic health conditions and depression (Chap-

man et al., 2005; Katon, 2011; Wolkowitz et al., 2011).

Primary and secondary appraisals of threat and available

coping resources are mediated by the brain (Lazarus &

Folkman, 1984; McEwen, 1998). The ongoing dilemma

of whether, when, where, how, and under what circum-

stances one conceals or discloses sexual orientation, cou-

pled with attempting to gauge potential consequences is a

primary appraisal process. If the individual chooses to

continue concealing her or his sexual orientation, then

concealment itself may be an additional chronic stressor

(Meyer, 2003). On the other hand, disclosure may over

time provide additional coping resources, reduce levels of

internalized heterosexism, and buffer the impact of stress

processes on health. Still, it is possible that those with

depression are more likely to report having been diag-

nosed with chronic health conditions. Longitudinal

1126 C. P. Hoy-Ellis and K. I. Fredriksen-Goldsen

research will be needed to clarify this relationship among

LGB older adults.

This study has also practice implications for address-

ing depression related to sexual orientation among LGB

older adults. Individual appraisals of stressors are central

to social stress processes (Pearlin, Mullan, Semple, &

Skaff, 1990). Subjective appraisals of stressors are more

strongly related to poor health outcomes, including

depression (Mittelman, Roth, Haley, & Zarit, 2004) than

objective stressors (Zarit, Todd, & Zarit, 1986). Accurate

assessment is foundational to effective treatment of

depression among older adults (Zarit & Zarit, 2007).

Therapeutic interventions to address the damaging effects

of internalized heterosexism have typically focused on

supporting the process of disclosure (Herek & Garnets,

2007). While such an approach can positively influence

the stress appraisal process, it also runs the risk of blaming

the individual for their poor mental health (Meyer, 2003).

On the other hand, if the social environment is less threat-

ening, it is likely to be appraised as less threatening,

which would benefit LGB older adults with depression

who do not have access to LGB-affirmative therapy.

Effectively addressing depression among LGB older

adults that is related to factors associated with sexual ori-

entation goes beyond intervening with current depression;

it also requires prevention efforts. More than two decades

ago, Albee and Ryan-Finn (1993) proposed that the occur-

rence of mental distress stemming from societal oppres-

sion can be described as a function of elements in the

social environment that promote marginalization divided

by the capacity of individuals and groups to resist margin-

alization. Taking such a social justice approach to primary

prevention requires empowering LGB older adults to

develop and strengthen their capacity to resist societal het-

erosexism, and that researchers identify and work toward

dismantling heterosexist social structures and institutions

(Kenny & Hage, 2009; Matthews & Adams, 2009). Such

an approach would serve to ameliorate existent depression

among today’s LGB older adults, and contribute to pre-

venting the development of depression among the next

generation of LGB older adults.

Limitations

In addition to its cross-sectional design, this study has

other limitations. Surveys were distributed via agency

mailing lists; participants who responded may differ in

important ways from those who did not. For example,

LGB older adults with higher levels of internalized het-

erosexism may be less likely to participate in research.

Similarly, LGB older adults who are not connected with

these service agencies may differ in significant ways from

those who are, for example, differing levels of conceal-

ment and disclosure. The ways in which individuals came

to be on agency mailing lists may also be an issue, as the

majority of respondents in this sample (70.6%) were not

utilizing services at the time that surveys were distributed.

While there is representation across the country, the find-

ings reported here cannot be generalized. Most partici-

pants were concentrated on the West Coast, Eastern

Seaboard, and parts of the Central US in major metropoli-

tan areas. Urban-dwelling LGB older adults likely have

experiences that vary from their rural-dwelling counter-

parts. These limitations may have skewed findings. It is

possible that LGB older adults who are connected with

agencies may differ on both mental and physical health

measures, which if true, likely biases these results.

The psychometric properties of the CESD-10 are well

established; measures to assess internalized heterosexism

and concealment/disclosure are less so. The Outness

Inventory (Mohr & Fassinger, 2000) requires subjective

interpretations of other likely perceptions, rather than

whether participants have actively or passively disclosed

or concealed their sexual orientation. The adapted version

of the Homosexuality Stigma Scale (Liu et al., 2009) may

not differentiate well between current and previous levels

of internalized heterosexism. For example, ‘I have tried

not to be LGB’ can refer to previous decades or current

experience.

Nonetheless, this study has valuable strengths. It is

one of the few to specifically examine LGB older adults

as a distinct population, and to apply the minority stress

framework to this population. In addition to providing

support for the minority stress model in general, it also

suggests that internal minority stressors may play a role

in physical as well as mental health outcomes (e.g.

depression), and that it is important to attend to both.

Through the use of SEM, this study provides further

evidence that may help to clarify the relationships

between disclosure of sexual orientation, internalized

heterosexism, chronic health conditions, and depression,

particularly the role of internalized heterosexism as

mediator suppressor of disclosure in both physical and

mental health.

Conclusion

We must begin to think in terms of health equity and

move toward targeting interventions upstream at commu-

nity and policy levels. Health equity means that every per-

son, regardless of social characteristics (including sexual

orientation), has a right to the best possible health, which

necessitates that any barriers to health that marginalized

groups experience must be addressed (Braveman & Grus-

kin, 2003). Health disparities are the gauge by which

progress toward health equity can be assessed; for LGB

older adults to attain mental health equity in the form of

resolving disparately high rates of depression, we must

attend to the unique barriers that they experience (Fredrik-

sen-Goldsen et al., 2014). Both the perceived and still all

too often real need to conceal an LGB identity – it is still

legal to discriminate based on sexual orientation in the

majority of states (Human Rights Campaign, 2015) – and

internalized heterosexism are barriers to LGB older

adults’ mental health equity. Recognizing that these bar-

riers are ultimately rooted in societal heterosexism

requires that we must also calibrate interventions at com-

munity and policy levels to address macro-level hetero-

sexism that fosters internalized heterosexism and the

perceived need to conceal one’s sexual orientation, which

Aging & Mental Health 1127

eventually manifests downstream in disparately high rates

depression.

Acknowledgments

Some research reported in this publication was supported in part by grants from the National Institute on Aging of the National Institutes of Health under Award Numbers R01AG026526 and 2R01AG026526-03A1 (Fredriksen-Goldsen, PI). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health, National Institute of Aging, the University of Utah, or the Uni- versity of Washington.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

National Institute on Aging of the National Institutes of Health [award number R01AG026526], [award number 2R01AG026526-03A1].

References

AARP Public Policy Institute. (2010). Chronic care: A call to action for health reform. Retrieved from http://assets.aarp. org/rgcenter/health/beyond_50_hcr.pdf

Albee, G.W., & Ryan-Finn, K.D. (1993). An overview of pri- mary prevention. Journal of Counseling and Development, 72, 115–123.

Andresen, E.M., Malmgren, J.A., Carter, W.B., & Patrick, D.L. (1994). Screening for depression in well older adults: Evalu- ation of a short form of the CES-D (Center for Epidemio- logic Studies Depression Scale). American Journal of Preventive Medicine, 10(2), 77–84.

Aneshensel, C.S. (2009). Toward explaining mental health dis- parities. Journal of Health and Social Behavior, 50(4), 377– 394.

Baron, R., & Kenny, D. (1986). The moderator–mediator vari- able distinction in social psychological research: Concep- tual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173–1182. doi:10.1037/0022-3514.51.6.1173

Blanchflower, D.G., & Oswald, A.J. (2008). Is well-being U- shaped over the life cycle? Social Science & Medicine, 66 (8), 1733–1749. doi:10.1016/j.socscimed.2008.01.030

Blazer, D.G. (2003). Depression in late life: Review and com- mentary. Journals of Gerontology Series A: Biological Sci- ences and Medical Sciences, 58(3), 249–265.

Blazer, D.G., & Hybels, C.F. (2005). Origins of depression in later life. Psychological Medicine, 35(9), 1241–1252. doi:10.1017/S0033291705004411

Boey, K.W. (1999). Cross-validation of a short form of the CES- D in Chinese elderly. International Journal of Geriatric Psy- chiatry, 14(8), 608–617.

Bollen, K.A. (1989). Structural equations with latent variables. New York, NY: Wiley.

Braveman, P., & Gruskin, S. (2003). Theory and methods: Defin- ing equity in health. Journal of Epidemiology and Commu- nity Health, 57, 254–258. doi:10.1136/jech.57.4.254

Centers for Disease Control and Prevention. (2011). CDC Health disparities and inequalities report – United States, 2011. MMWR 2011, 60(Suppl), 1–116.

Centers for Disease Control and Prevention. (2013). The state of aging and health in America 2013. Retrieved from http:// www.cdc.gov/features/agingandhealth/state_of_aging_and_ health_in_america_2013.pdf

Centers for Disease Control and Prevention. (2015). Aging and depression. Healthy Aging. Retrieved from http://www.cdc. gov/aging/mentalhealth/depression.htm

Centers for Disease Control and Prevention and National Associ- ation of Chronic Disease Directors. (2009). The state of men- tal health and aging in America – issue brief 2: Addressing depression in older adults: Selected evidence-based pro- grams, 1–12. Retrieved from http://www.cdc.gov/aging/pdf/ mental_health_brief_2.pdf

Chapman, D.P., Perry, G.S., & Strine, T.W. (2005). The vital link between chronic disease and depressive disorders. Pre- venting Chronic Disease, 2(1), 1–10.

Cheung, G.W., & Lau, R.S. (2008). Testing mediation and sup- pression effects of latent variables: Bootstrapping with struc- tural equation models. Organizational Research Methods, 11(2), 296–325. doi:10.1177/1094428107300343

Cole, S.W., Kemeny, M.E., Taylor, S.E., & Visscher, B.R. (1996). Elevated physical health risk among gay men who conceal their homosexual identity. Health Psychology, 15 (4), 243–251.

David, S., & Knight, B.G. (2008). Stress and coping among gay men: Age and ethnic differences. Psychology and Aging, 23 (1), 62–69. doi:10.1037/0882-7974.23.1.62

Duncan, O.D. (1975). Recursive models. Introduction to structural equation models (pp. 25–66). New York: Academic Press.

Ferraro, K.F., & Shippee, T.P. (2009). Aging and cumulative inequality: How does inequality get under the skin? The Gerontologist, 49(3), 333–343. doi:10.1093/geront/gnp034

Fiske, A., Wetherell, J.L., & Gatz, M. (2009). Depression in older adults. Annual Review of Clinical Psychology, 5, 363– 389. doi:10.1146/annurev.clinpsy.032408.153621

Floyd, F.J., & Bakeman, R. (2006). Coming out across the life course: Implications of age and historical context. Archives of Sexual Behavior, 35(3), 287–296. doi:10.1007/s10508- 006-9022-x

Fredriksen-Goldsen, K.I., Cook-Daniels, L., Kim, H.-J., Ero- sheva, E.A., Emlet, C.A., Hoy-Ellis, C.P., … Muraco, A. (2013). Physical and mental health of transgender older adults: An at-risk and underserved population. The Geron- tologist, 54(3), 488–500. doi:10.1093/geront/gnt021

Fredriksen-Goldsen, K.I., Emlet, C.A., Kim, H.-J., Muraco, A., Erosheva, E.A., Goldsen, J., & Hoy-Ellis, C.P. (2013). The physical and mental health of lesbian, gay male, and bisex- ual (LGB) older adults: The role of key health indicators and risk and protective factors. The Gerontologist, 53(4), 664– 675. doi:10.1093/geront/gns123

Fredriksen-Goldsen, K.I., Hoy-Ellis, C.P., Goldsen, J., Emlet, C. A., & Hooyman, N.R. (2014). Creating a vision for the future: Key competencies and strategies for culturally com- petent practice with lesbian, gay, bisexual, and transgender (LGBT) older adults in the health and human services. Jour- nal of Gerontological Social Work, 57, 80–107. doi:10.1080/01634372.2014.890690

Fredriksen-Goldsen, K.I., Kim, H.-J., Barkan, S.E., Muraco, A., & Hoy-Ellis, C.P. (2013). Health disparities among lesbian, gay male and bisexual older adults: Results from a popula- tion-based study. American Journal of Public Health, 103 (10), 1802–1809. doi:10.2105/AJPH.2012.301110

Fredriksen-Goldsen, K.I., Kim, H.J., Shiu, C., Goldsen, J., & Emlet, C.A. (2014). Successful aging among LGBT older adults: Physical and mental health-related quality of life by age group. The Gerontologist, 55(1), 154–168. doi:10.1093/ geront/gnu081

Fredriksen-Goldsen, K.I., Simoni, J.M., Kim, H.-J., Lehavot, K., Walters, K. L., Yang, J., … Muraco, A. (2014). The health equity promotion model: Reconceptualization of lesbian, gay, bisexual, and transgender (LGBT) health disparities. American Journal of Orthopsychiatry, 84(6), 653–663 doi:10.1037/ort0000030

Gates, G.J., & Newport, F. (2012). Special report: 3.4% of U.S. adults identify as LGBT. Inaugural gallup findings based on

1128 C. P. Hoy-Ellis and K. I. Fredriksen-Goldsen

more than 120,000 interviews. Retrieved from http://www. gallup.com/poll/158066/special-report-adults-identify-lgbt. aspx

Grzywacz, J.G., Hovey, J.D., Seligman, L.D., Arcury, T.A., & Quandt, S.A. (2006). Evaluating short-form versions of the CES-D for measuring depressive symptoms among immi- grants from Mexico. Hispanic Journal of Behavioral Scien- ces, 28(3), 404–424. doi:10.1177/0739986306290645

Hatzenbuehler, M.L., Phelan, J.C., & Link, B.G. (2013). Stigma as a fundamental cause of population health inequalities. American Journal of Public Health, 103(5), 813–821. doi:10.2105/AJPH.2012.301069

Herek, G.M., & Garnets, L.D. (2007). Sexual orientation and mental health. Annual Review of Clininical Psychology, 3, 353–375. doi:10.1146/annurev.clinpsy.3.022806.091510

Hogg, M.A., Terry, D.J., & White, K.M. (1995). A tale of two theories: A critical comparison of identity theory with social identity theory. Social Psychology Quarterly, 58(4), 255– 269.

Hooper, D., Coughlan, J., & Mullen, M.R. (2008). Structural equation modeling: Guidelines for determining model fit. Electronic Journal of Business Research Methods, 6(1), 53– 60.

Hoy-Ellis, C.P. (2015). Concealing concealment: The mediating role of internalized heterosexism in psychological distress among lesbian, gay, and bisexual older adults. Journal of Homosexuality, 63(4), 487–506. doi:10.1080/ 00918369.2015.1088317

Human Rights Campaign. (2015). Why the equality act? Retrieved from http://www.hrc.org//resources/entry/why- the-equality-act

Iacobucci, D., Saldhana, N., & Deng, X. (2007). A meditation on mediation: Evidence that structural equations models per- form better than regressions. Journal of Consumer Psychol- ogy, 12(2), 139–153. doi:10.1016/S1057-7408(07)70020-7

Institute of Medicine. (2011). The health of lesbian, gay, bisex- ual, and transgender people: Building a foundation for better understanding. Washington, DC: The National Acad- emies Press.

Irwin, M., Artin, K.H., & Oxman, M.N. (1999). Screening for depression in the older adult: Criterion validity of the 10- item Center for Epidemiological Studies Depression Scale (CES-D). Archives of Internal Medicine, 159(15), 1701– 1174.

Juster, R.-P., McEwen, B.S., & Lupien, S.J. (2010). Allostatic load biomarkers of chronic stress and impact on health and cognition. Neuroscience & Biobehavioral Reviews, 35(1), 2– 16. doi:10.1016/j.neubiorev.2009.10.002

Katon, W.J. (2011). Epidemiology and treatment of depression in patients with chronic medical illness. Dialogues in Clini- cal Neuroscience, 13(1), 7–23.

Kenny, D.A. (2014). Measuring model fit. Retrieved from http:// davidakenny.net/cm/fit.htm

Kenny, M.E., & Hage, S.M. (2009). The next frontier: Preven- tion as an instrument of social justice. Journal of Primary Prevention, 30(1), 1–10. doi:10.1007/s10935-008-0163-7

Kertzner, R.M., Meyer, I.H., Frost, D.M., & Stirratt, M.J. (2009). Social and psychological well-being in lesbians, gay men, and bisexuals: The effects of race, gender, age, and sexual identity. American Journal of Orthopsychiatry, 79(4), 500– 510. doi:10.1037/a0016848

Krieger, N. (1999). Embodying inequality: A review of con- cepts, measures, and methods for studying health conse- quences of discrimination. International Journal of Health Services, 29(2), 295–352.

Kuzawa, C.W., & Sweet, E. (2009). Epigenetics and the embodi- ment of race: Developmental origins of US racial disparities in cardiovascular health. American Journal of Human Biol- ogy, 21(1), 2–15. doi:10.1002/ajhb.20822

Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, and cop- ing. New York, NY: Springer.

Liu, H., Feng, T., & Rhodes, A.G. (2009). Assessment of the Chinese version of HIV and homosexuality related stigma scales. Sexually Transmitted Infections, 85(1), 65–69. doi:10.1136/sti.2008.032714

Marmot, M.G., Stansfeld, S., Patel, C., North, F., Head, J., White, I., … Smith, G.D. (1991). Health inequalities among British civil servants: The Whitehall II study. The Lancet, 337(8754), 1387–1393. doi:10.1016/0140-6736(91)93068-K

Marmot, M.G., & Wilkinson, R.G. (2006). Social determinants of health (2nd ed.). New York, NY: Oxford University Press.

Matsueda, R.L. (2012). Key advances in the history of structural equation modeling. In R.H. Hoyle (Ed.), Handbook of struc- tural equation modeling (pp. 17–42). New York, NY: The Guilford Press.

Matthews, C.R., & Adams, E.M. (2009). Using a social justice approach to prevent the mental health consequences of het- erosexism. Journal of Primary Prevention, 30(1), 11–26. doi:10.1007/s10935-008-0166-4

McEwen, B.S. (1998). Stress, adaptation, and disease. allostasis and allostatic load. Annals of the New York Academy of Sci- ences, 840, 33–44.

McEwen, B.S. (2006). Protective and damaging effects of stress mediators: Central role of the brain. Dialogues in Clinical Neuroscience, 8(4), 367–381.

Meyer, I.H. (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychological Bulletin, 129(5), 674– 697. doi:10.1037/0033-2909.129.5.674

Mittelman, M.S., Roth, D.L., Haley, W.E., & Zarit, S.H. (2004). Effects of a caregiver intervention on negative caregiver appraisals of behavior problems in patients with Alzheimer’s disease: Results of a randomized trial. Journals of Gerontol- ogy Series B: Psychological Sciences and Social Sciences, 59B(1), P27–P34.

Mohr, J., & Fassinger, R. (2000). Measuring dimensions of les- bian and gay male experience. Measurement and Evaluation in Counseling and Development, 33(2), 66–90.

Murgatroyd, C., & Spengler, D. (2011). Epigenetic program- ming of the HPA axis: Early life decides. Stress, 14(6), 581– 589. doi:10.3109/10253890.2011.602146

National Senior Citizens Law Center. (2011). LGBT older adults in long-term care facilities: Stories from the field. Retrieved from http://www.lgbtlongtermcare.org/authors/

Pachankis, J.E. (2007). The psychological implications of con- cealing a stigma: A cognitive-affective-behavioral model. Psychological Bulletin, 133(2), 328–345. doi:10.1037/0033- 2909.133.2.328

Pearlin, L.I., Mullan, J.T., Semple, S.J., & Skaff, M.M. (1990). Caregiving and the stress process: An overview of concepts and their measures. The Gerontologist, 30(5), 583–594.

Pinquart, M., & Sorenson, S. (2007). Correlates of physical health of informal caregivers: A meta-analysis. Journal of Gerontology, 62B(2), P126–P137.

Pratt, L.A., & Brody, D.J. (2008). Depression in the United States household population, 2005–2006. NCHS Data Brief, (7), 1–8. Retrieved from http://www.ncbi.nlm.nih.gov/ pubmed/19389321

Preacher, K.J., & Kelley, K. (2011). Effect size measures for mediation models: Quantitative strategies for communicat- ing indirect effects. Psychological Methods, 16(2), 93–115. doi:10.1037/a0022658

Radloff, L.S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psycho- logical Measurement, 1, 385–401.

Rawls, T.W. (2004). Disclosure and depression among older gay and homosexual men: Findings from the Urban Men’s Health Study. In G. Herdt & B. de Vries (Eds.), Gay and les- bian aging: Research and future directions (pp. 117–41). New York, NY: Springer.

Rucker, D.D., Preacher, K.J., Tormala, Z.L., & Petty, R.E. (2011). Mediation analysis in social psychology: Current

Aging & Mental Health 1129

practices and new recommendations. Social and Personality Psychology Compass, 5(6), 359–371. doi:10.1111/j.1751- 9004.2011.00355.x

Seeman, T.E., Singer, B.H., Ryff, C.D., Dienberg Love, G., & Levy-Storms, L. (2002). Social relationships, gender, and allostatic load across two age cohorts. Psychosomatic Medi- cine, 64(3), 395–406.

Soni, A. (2012). Trends in use and expenditures for depression among U.S. adults age 18 and older, civilian noninstitution- alized population, 1999 and 2009. Retrieved from http:// meps.ahrq.gov/data_files/publications/st377/stat377.pdf

StataCorp. (2011). Stata: Release 12 (Vol. Stata). College Sta- tion, TX: StataCorp LP.

Substance Abuse and Mental Health Services Administration. (2013). Results from the 2012 national survey on drug use and health: Mental health findings. Retrieved from http:// www.samhsa.gov/data/sites/default/files/2k12MH_Findings/ 2k12MH_Findings/NSDUHmhfr2012.htm#sec2-3

U.S. Census Bureau. (2015). 2014 National population projec- tions: Summary tables. Table 9. Projections of the popula- tion by age and sex for the United States: 2015 to 2060. Retrieved from https://www.census.gov/population/projec tions/data/national/2014/summarytables.html

U.S. Department of Health and Human Services. (2013). Les- bian, gay, bisexual, and transgender health. 2020 Topics & Objectives. Retrieved from https://www.healthypeople.gov/ 2020/topics-objectives/topic/lesbian-gay-bisexual-and-trans gender-health/objectives

Un€utzer, J., Schoenbaum, M., Katon, W.J., Fan, M. Y., Pincus, H.A., Hogan, D., & Taylor, J. (2009). Healthcare costs asso- ciated with depression in medically Ill fee-for-service medi- care participants. Journal of the American Geriatric Society, 57(3), 506–510. doi:10.1111/j.1532-5415.2008.02134.x

Uysal, A., Lin, H.L., & Knee, C.R. (2010). The role of need sat- isfaction in self-concealment and well-being. Personality and Social Psychology Bulletin, 36(2), 187–199. doi:10.1177/0146167209354518

Valanis, B.G., Bowen, D.J., Bassford, T., Whitlock, E., Charney, P., & Carter, R.A. (2000). Sexual orientation and health:

Comparisons in the Women’s Health Initiative sample. Archives of Family Medicine, 9(9), 843–853.

Wallace, S.P., Cochran, S.D., Durazo, E.M., & Ford, C.L. (2011). The health of aging lesbian, gay and bisexual adults in California. Los Angeles, CA: UCLA Center for Health Policy Research.

Wolford, C.C., McConoughey, S.J., Jalgaonkar, S.P., Leon, M., Merchant, A.S., Dominick, J.L., … Hai, T. (2013). Tran- scription factor ATF3 links host adaptive response to breast cancer metastasis. Journal of Clinical Investigation, 123(7), 2893–2906. doi:10.1172/JCI64410

Wolkowitz, O.M., Reus, V.I., & Mellon, S.H. (2011). Of sound mind and body: Depression, disease, and accelerated aging. Dialogues in Clinical Neuroscience, 13(1), 25–39.

World Health Organization. (2003). Social determinants of health: The solid facts. Retrieved from http://www.euro. who.int/en/what-we-publish/abstracts/social-determinants- of-health.-the-solid-facts

World Health Organization. (2012). Depression. Mental Health. Retrieved from http://www.who.int/mediacentre/factsheets/ fs369/en/

Yang, Y. (2007). Is old age depressing? Growth trajectories and cohort variations in late-life depression. Journal of Health & Social Behavior, 48(1), 16–32.

Zarit, S.H., Todd, P.A., & Zarit, J.M. (1986). Subjective burden of husbands and wives as caregivers: A longitudinal study. The Gerontologist, 26(3), 260–266.

Zarit, S.H., & Zarit, J.M. (2007). Mental disorders in older adults: Fundamentals of assessment and treatment (2nd ed.). New York, NY: The Guilford Press.

Zhang, W., O’Brien, N., Forrest, J.I., Salters, K.A., Patterson, T. L., Montaner, J.S.,… Lima, V.D. (2012). Validating a short- ened depression scale (10 Item CES-D) among HIV-positive people in British Columbia, Canada. PLoS One, 7(7), e40793. doi:10.1371/journal.pone.0040793

Zuckerman, M. (1999). Diathesis-stress models. Vulnerability to psychopathology: A biosocial model (pp. 3–23). Washing- ton, DC: American Psychological Association.

1130 C. P. Hoy-Ellis and K. I. Fredriksen-Goldsen

Copyright of Aging & Mental Health is the property of Routledge and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.