Addressing the Needs of LGBT Older Adults in San Francisco: Recommendations for the Future

Addressing the Needs of LGBT Older Adults in San Francisco: Recommendations for the Future Karen I. Fredriksen-Goldsen, PhD, Hyun-Jun Kim, PhD, Charles P. Hoy-Ellis, PhC, Jayn Goldsen, BS, Diana Jensen, MPP, Marcy Adelman, PhD, L. Michael Costa, MPP, Brian De Vries, PhD Report prepared for the LGBT Aging Policy Task Force, San Francisco, CA Report published by Institute for Multigenerational Health, University of Washington, Seattle, WA

Š July 2013

Table of Contents Acknowledgements ......................................................................................2 Executive Summary .....................................................................................3 Introduction ..................................................................................................7 Background Characteristics .........................................................................9 Services and Programs ...............................................................................14 Housing ......................................................................................................21 Resources and Risks ..................................................................................26 Health .........................................................................................................35 Key Findings ..............................................................................................39 Recommendations ......................................................................................41 References ..................................................................................................43 Appendices .................................................................................................48 Methodology .....................................................................................48 Key Terms .........................................................................................50 Tables ................................................................................................55

1

Acknowledgments We would like to extend our gratitude to the individuals, agencies, and organizations whose generous support made this important project a success. Alvin H. Baum, Bob Ross Foundation, Horizons Foundation, Deb L. Kinney, Esq., Meals On Wheels San Francisco, The San Francisco Foundation, The San Francisco LGBT Pride Celebration Committee, San Francisco Human Services Agency, Department of Aging and Adult Services San Francisco, and San Francisco Human Rights Commission all contributed resources or made financial contributions in order to make this report possible. The San Francisco LGBT Senior Policy Research Team played an important role in the planning, development, and implementation of the project including Marcy Adelman, PhD; Jason Alley, PhD; Bill Ambrunn, Esq; L. Michael Costa, MPP; Brian De Vries, PhD; Diana Jensen, MPP; Ashley McCumber; and Tom Nolan – whose knowledge and expertise reflect their leadership and dedication to improving the lives of LGBT older adults. We also want to thank Charles A. Emlet, PhD, and Anna Muraco, PhD, for their assistance with this report. We wholeheartedly want to acknowledge and thank the hundreds of LGBT older adults of San Francisco who so generously participated in this study, sharing their many strengths and unique needs.

2

Executive Summary San Francisco has a large, diverse, and growing population of lesbian, gay, bisexual, and transgender (LGBT) older adults. Population-based research suggests that there are likely between 18,000 to 20,000 LGBT adults aged 60 years or older living in San Francisco (Jensen, 2012). To date, very limited research has systematically examined the unique aging needs, strengths, and challenges facing LGBT older adults. This report is the result of collaborative effort between the San Francisco LGBT Aging Policy Task Force, the Task Force Research Team, and Caring and Aging with Pride of the Institute for Multigenerational Health at the University of Washington. This report highlights key findings and concludes with programmatic, policy, and research recommendations. An important goal of the study was to obtain a diverse representation of LGBT older adults in San Francisco. From April to June 2013, electronic and hardcopy surveys in five languages (English, Spanish, Chinese, Russian, and Tagalog) were distributed through media, service agencies, community events, and community outreach activities. The survey was completed by 616 LGBT City residents, aged 60 to 92 years old. More than two-thirds (71%) of the participants are gay men, 22% lesbians, 4% bisexuals, and 4% transgender. Although the majority of the study sample is non-Hispanic white, 21% identify as Hispanic, Asian American, African American, Native American, “other” race or ethnicity, and multiracial. Because of the targeted nature of the outreach activities, the findings outlined below represent those of the study participants. It is important to recognize that this is one of the most diverse samples to date in LGBT aging research; however, the findings are not generalizable to LGBT older adults living in San Francisco. Several findings suggest risk of isolation and lack of support resources among the participants:  Nearly 60% of the participants live alone.  40% do not have the minimum income necessary to meet their basic needs, based on California Elder Economic Security Index; 30% have incomes below the 200% of the federal poverty level.  Only 15% have children; 60% of whom indicate that their children are not available to help them if needed.  Nearly two-thirds (63%) are neither partnered nor married.  Bisexual, African American, and Hispanic participants are less likely to own a home. Findings related to high need for services and programs include:  The most frequently needed programs and services identified by participants are health services, health promotion, mental health services, housing assistance, having a case manager/social worker, telephone/online referrals, and meal site/free groceries. Services and programs with a high rate of unmet need include health promotion, door-to-door transportation, caregiver support, day programs, housing assistance, in-home care, and telephone/online referrals.  Bisexual and transgender participants report elevated need for most services.  LGBT participants with lower incomes and lower educational attainment, as well as those living alone and those who are not partnered also have higher service needs.

3

Despite needing services, some participants did not access them for a variety of reasons:  Half of those who use alcohol/substance abuse programs and housing assistance indicate that they do not feel comfortable utilizing these services as an LGBT older adult.  About one in six participants do not use meal site/free grocery programs and telephone/online referrals because they feel these services are not LGBT friendly.  The most common reasons given for not accessing services and programs are because they are difficult to access or too expensive. Safe, stable, and affordable housing was identified as an important concern in San Francisco:  Two-thirds of participants are concerned that they may not be able to stay in their home and may need to relocate.  Primary reasons for the potential need to relocate include economics, health, and needs related to aging.  Nearly one-quarter reported needing housing assistance. An elevated need is observed among those with lower socioeconomic status and with HIV/AIDS; yet 42% of housing assistance service users feel unsafe obtaining assistance as an LGBT person.  LGBT respondents who live alone, those with lower incomes, and those with less education are at an elevated risk for housing instability. Many participants have resources and strengths available to assist them in meeting their aging and health needs:  Most participants have moderate levels of social support.  Faith communities are an important source of social support, especially for transgender and Asian American participants.  LGBT participants in legally recognized relationships report better health, less need for community services and programs, higher levels of social support, and higher rates of home ownership and housing stability.  Lesbian and gay male participants have relatively high levels of sexual orientation and gender identity disclosure; bisexual and transgender participants have much lower levels of disclosure. Disclosure has shown to be a positive protective factor for mental health.  There remain some (9%) who say they have no one to turn to for social support. Gay men are at higher risk for lacking social support than lesbians. Many participants also face serious risks, which can increase their vulnerability:  Nearly half of the participants have experienced discrimination in the past 12 months because of their sexual orientation or gender identity. More than a third have experienced age-based discrimination; lesbians are more likely to experience age-based discrimination than gay men.  One in five LGBT participants has been victimized during the past 12 months because of their sexual orientation or gender identity.  5% have been abused by someone in a trusting relationship, including friends, partners and spouses, family members, and paid caregivers.  Only about one-quarter of those abused or victimized reported the crime to the authorities; 9% did not report the crime because they did not trust authorities to treat LGBT people fairly.

4

Poor physical and mental health is of great concern:  Nearly one-third of the participants report poor general health; more than 40% have one or more physical disabilities.  Among the male participants, 33% are living with HIV/AIDS.  10% of participants experience frequent limited activities due to poor mental health.  15% report having seriously considered taking their own lives in the past 12 months.  Transgender participants, those living in poverty, and those not married or partnered are more likely to have poor health. Recommendations Existing population-based research demonstrates that LGBT older adults are an at-risk, health disparate, and vulnerable population. Based on these key findings, programmatic, policy, and research recommendations are as follows: 1. Ensure the development or expansion of services to address the unmet needs of LGBT older adults. The services that emerged with the highest unmet need among the LGBT older adult participants in this study include: health promotion, door-to door transportation, caregiver support, and day programs. 2. Improve the LGBT-friendliness of specific health and social services. Areas in which some LGBT older adults do not feel comfortable include: alcohol/substance abuse programs, housing assistance, and veterans’ services. In these service areas more training is likely needed to create an LGBT welcoming and friendly atmosphere and to ensure competent services. 3. Expand caregiving support programs for LGBT older adults. The participants have a high need for caregiving support, combined with low rates of caregiving arrangements or plans. Most live alone and do not have children available to help them. 4. Improve the availability of LGBT-friendly housing assistance. Housing instability is a major concern; many indicate they may be unable to stay in their home, primarily due to economic and health reasons or changing needs related to aging. Many participants, especially those with lower socioeconomic status and HIV/AIDS, need housing assistance, but many participants feel unsafe obtaining assistance as an LGBT person. 5. Address the distinct needs of lesbians, gay men, bisexuals, and transgender adults as separate sub-groups. Overall, transgender and bisexual participants report elevated need for most services, especially mental health services, meal sites and free groceries, health services, and health promotion. Racial and ethnic minority participants report higher rates of service need: mental health services for Hispanics and African Americans, housing assistance and day programs for Hispanics and Asian Americans. Gay men are less likely than lesbians to turn to friends, family members, or neighbors for social and emotional support. Lesbians are more likely than gay men to experience age-based discrimination. 6. Develop a multifaceted plan to combat discrimination by sexual orientation and gender identity and the abuse of LGBT older adults. Effective campaigns to combat discrimination and abuse have documented the need for a multifaceted approach, including 5

the need for public awareness among the general and affected communities and businesses, as well as the evaluation and development of ways to strengthen the reporting, investigative, legislative, and judicial processes. 7. Establish a suicide prevention program that targets LGBT older adults. An alarming finding is the number of participants that contemplated suicide within the past 12 months. Many health issues emerge in this study including relatively high rates of disability, and poor physical and mental health, which in previous research has been found to be associated with increased risk of depression, which in turn can increase the risk of suicide. 8. Provide training and services to help LGBT older adults as well as providers anticipate future aging and health planning needs. Our findings reveal there are unmet planning needs that warrant attention, including last will and testament, and powers of attorney for health care and finances. An area that warrants attention is that LGBT participants who are not married or partnered report the lowest rates of having a will, powers of attorney for health care and finances, and revocable/irrevocable trust. 9. Promote advocacy to ensure that the needs of LGBT older adults continue to be addressed in local and state planning processes, such as the development of the Area Plan. Given the array of unaddressed needs that have been identified in this report, it is important that LGBT older adults be considered as an at-risk and vulnerable population, with their particular needs addressed in the City’s planning processes designed to address aging related needs of older adults. 10. Collect quality data on the aging and health needs of LGBT older adults. Recent federal mandates have advanced the inclusion of sexual orientation and gender identity questions in public health surveys. It is imperative that San Francisco include sexual orientation, sexual behavior, and gender identity questions in City sponsored aging and health surveys and other data collection tools. Such information is critically needed so that these communities are considered in planning and service development. A more complete understanding of the needs of San Francisco’s aging LGBT population will only be accomplished when sexual orientation and gender identity are in both aging and non-aging related City data-bases. 11. Develop a strategy to successfully reach racial and ethnic minorities, bisexuals, and transgender adults and continue to advance and enhance research with and for diverse LGBT older adults. An important goal of this study was to reach out to diverse LGBT older adults, who have not typically been included in studies of LGBT aging. Even with extensive outreach efforts and making the survey available in five different languages, reaching racial and ethnic minorities, bisexuals, and transgender older adults was difficult. These older adults may experience high levels of isolation as they age, higher than that of the general population. In future research it will be important to test the use of language related to sexual orientation and gender identity, and differing recruitment strategies for diverse populations. It is clear that the LGBT participants have important strengths and resources that can foster their aging, health, and well-being, yet they also face significant risks, which can increase their vulnerability as they age. Through a better understanding of their lived experiences, policymakers and other key stakeholders can initiate program, policy, and research initiatives to better serve the needs of older LGBT adults who live in San Francisco. 6

Introduction In 2012 the San Francisco Board of Supervisors established a time limited LGBT Aging Policy Task Force to explore the health and wellness issues facing San Francisco’s LGBT older adults and to provide the Board with actionable policy and program recommendations by the end of 2013. The Task Force commissioned this community survey after a review of available City data on San Francisco’s lesbian, gay, bisexual, and transgender (LGBT) older adults revealed little information on LGBT older adults in general and for LGBT seniors of color in particular. Outreach efforts for this survey focused on previously underrepresented groups (including racial and ethnic communities, bisexuals, transgender adults, non-English speakers, and those living in SRO’s or experiencing homelessness). This report is a follow-up of LGBT Older Adults in San Francisco: Health, Risks, and Resilience - Findings from Caring and Aging with Pride, which was released in January, 2013, and provided an initial snapshot of 295 LGBT older adults living in San Francisco who participated in the national research project, Caring and Aging with Pride. Understanding the needs of older LGBT adults is critical in light of the aging and increasingly diverse population. Projections based on the 2010 census suggest that the one in five in San Francisco will be 65 or older by 2060, nearly doubling from the current 109,517, to 177,963 (U.S. Census Bureau, 2013a). San Francisco is undergoing a dramatic racial, ethnic, and aging demographic shift; 40% of older adults in San Francisco were non-Hispanic white in 2010 while only 26% will be non-Hispanic white in 2060. LGBT adults are also part of this increasing diversity; their numbers are projected to double to 3 million nationally by 2050 (Auldridge & Espinoza, 2013). Indeed, population-based research suggests that there are likely 18,000 to 20,000 LGBT adults aged 60-years old or older living in San Francisco, based on estimates that among City residents aged 60 and older, 11.1% to 12.4% identify as LGBT (Jensen, 2012). Despite this increasing diversity, a great deal is still unknown about the health and aging needs of LGBT older adults (Institute of Medicine, 2011), which is a serious shortcoming in health disparities research (Centers for Disease Control and Prevention, 2011a). Accumulating research leaves little doubt – LGBT older adults experience significant health disparities (Fredriksen-Goldsen, Cook-Daniels, et al., 2013; Fredriksen-Goldsen, Kim, Barkan, Muraco, & Hoy-Ellis, 2013; Valanis et al., 2000; Wallace, Cochran, Durazo, & Ford, 2011). Health disparities are differences in population health that result from social, economic, and environmental marginalization (U.S. Department of Health and Human Services, 2010). Although population-based research on LGBT older adults is limited, important findings are emerging. A recent study by Fredriksen-Goldsen, Kim, and colleagues (2013) compared lesbian, gay, and bisexual (LGB) adults aged 50 and older to their heterosexual peers. LGB older adults had higher rates of disability and frequent mental distress. They were also less likely to be partnered or married. Legal marriage is no panacea for the challenges that face sexual minorities, but there is evidence that it provides health and mental health benefits (Buffie, 2011; Center for American Progress, 2012; Gay and Lesbian Medical Association, 2008; Herdt & Kertzner, 2006; Herek, 2006, 2007; Rendall, Weden, Favreault, & Waldron, 2011). Although some of this "benefit" is undoubtedly economic in nature, accruing from the more than 1,100 rights, benefits, and privileges extended through federally-recognized marriage (General Accounting Office & Office of the General Counsel, 1997), there is also strong evidence in support of the social and 7

psychological benefits (Riggle, Rostosky, & Horne, 2010; Wienke & Hill, 2009; Wight, LeBlanc, de Vries, & Detels, 2012). The study by Fredriksen-Goldsen, Kim, and associates (2013) also noted important gender differences in health disparities. Compared to older heterosexual women, lesbians and bisexual older women had higher rates of obesity, and were at greater risk for cardiovascular disease; they were also less likely to have had some health screenings (e.g., mammography). Compared to older heterosexual men, gay and bisexual older men had poorer general physical health, and were more likely to live alone. Findings from the California Health Interview Survey (CHIS) also found health disparities between LGB older adults and their heterosexual peers. In comparisons of LGB adults aged 50 to 70-years old with their heterosexual counterparts, Wallace and associates (2011) found that LGB older adults were more likely to report psychological distress, disability, and poor general health. They also found that although the LGB participants were more likely to routinely see their primary care physician, lesbian and bisexual women were more likely to postpone necessary medical care. Gay and bisexual men were at increased risk for diabetes and hypertension, and for living alone. The available research on transgender older adults is even more limited. Existing research suggests that transgender older adults are at much greater risk for poor health than nontransgender LGB peers. For example, transgender older adults have even higher rates of disability, poor general health, and depression than LGB older adults (Fredriksen-Goldsen, Cook-Daniels, et al., 2013). Policymakers and service providers require accurate and timely information regarding aging and health disparities to prioritize public policies and to develop culturally competent programs and services (National Research Council: Panel on Race, 2004). Although the scope of health disparities is being investigated, LGBT older adults continue to be relatively invisible in services, policies, and research (Fredriksen-Goldsen & Muraco, 2010; Metlife Mature Market Institute & American Society on Aging, 2010). This report provides an overview of findings regarding key experiences, needs, and barriers to services and programs that the LGBT older adult participants living in San Francisco face. It is based on 616 surveys completed by LGBT City residents aged 60 and older. The comparisons included in this report provide a broad context for understanding the data but due to the targeted outreach efforts the results of this survey are best understood as descriptive rather than representational. It is our hope that this report will assist policymakers and other stakeholders in planning and service development to address the growing needs of LGBT older adults. The report is organized as follows: Background Characteristics, Services and Programs, Housing, Resources and Risks, Health, Key Findings, and Recommendations.

8

Background Characteristics In the city of San Francisco, 616 lesbian, gay, bisexual, and transgender (LGBT) adults aged 60 to 92-years old completed the LGBT Aging Policy Task Force Community Questionnaire. The project, Caring and Aging in San Francisco, was undertaken to better understand the health and service needs of older LGBT residents in San Francisco. Outreach for the survey was intended to generate a sample that could highlight the diversity of the sexual orientation, gender and gender identity, race/ethnicity, socioeconomic status, and other sociodemographic characteristics of San Francisco’s LGBT older adults. The project announcement and surveys were distributed through the media, agency contact lists, community events, and outreach efforts. Because of its targeted nature, the background characteristics may not be reflective of LGBT older adults living in San Francisco. Overall participant characteristics can be found in Appendix Table 1.1. A breakdown of characteristics by sexual orientation and gender identity can be found in Appendix Table 1.2. When comparing San Francisco’s LGBT older adult participants to older adults in San Francisco’s general population, some preliminary findings emerge that deserve attention:  58% of LGBT participants live alone, as opposed to 28% of older adults in San Francisco (Ruggles et al., 2010).  LGBT participants are more educated; 70% have a bachelor's degree or higher, compared to 35% of older adults in San Francisco (U.S. Census Bureau, 2011).  San Francisco’s older adults are 43% Asian/Pacific Islander, 41% non-Hispanic white, 9% Latino/Hispanic, and 7% African American (Ruggles et al., 2010). The race and ethnicity of the LGBT participants: 79% non-Hispanic white, 7% Latino/Hispanic, 5% African American, 4% Asian/Pacific Islander, and 2% Native American. Sexual orientation, gender identity, and gender The vast majority of study participants (71%, n = 432) identify as gay men; 22% identify as lesbian (n = 135); 2% as bisexual women (n = 12); 2% as bisexual men (n = 15); 1% as heterosexual (n = 6); and 1% identify as other (n = 6). Four percent identify as transgender (n = 26). Regardless of sexual orientation or gender identity, men make up 73% of the sample, women 25%, and less than 2% identify as other. Age The average age of LGBT participants is 68-years old (M = 67.62, SD = 6.12). For analytic purposes, the participants are grouped into three age cohorts:  60 to 69-years old (71%)  70 to 79-years old (23%)  80-years old and older (6%)

9

Race and Ethnicity In terms of race and ethnicity, 79% of participants (n = 474) identified as nonHispanic white; slightly more than 7% (n = 45) as Hispanic or Latino/a; 5% (n = 30) as African American; 4% (n = 23) as Asian American; and 2% (n = 10) as Native American. One percent (n = 8) identified as two or more races and 2% (n = 13) identified as other. Despite significant outreach to racial and ethnic minority groups, the survey participants are proportionally more non-Hispanic white than the overall older adult San Francisco population.

Education Study participants are well-educated:  Master's degree or higher (46%)  Bachelor's degree (25%)  Some college (22%)  High school education or less (8%) Income Although education and income are highly correlated in the general population, higher education does not necessarily equate to commensurate income among LGBT adults, as mounting evidence indicates (Fredriksen-Goldsen, Kim, et al., 2013; Grant, 2010; Grant et al., 2011). Using household size, reported income, and calculation guidelines provided by the federal government (U. S. Department of Health and Human Services, 2013), we found that 30% of participants have annual household incomes at or below 200% of the federal poverty level (FPL), despite the fact that 70% have a bachelor's, master's, or doctoral degree. In addition, 40% of participants are living at or below the California Elder Economic Security Index (EESI), when living arrangement (living alone or living with a partner), housing type (rent, own with mortgage, or own- mortgage paid off), and annual

10

household income are considered.1 The EESI provides a county-level threshold for making ends meet, which considers actual costs of living, compared to income. Employment status The majority of participants (64%) are not employed; 19% work part time and 17% work full time. Military service Almost one-quarter (22%) of participants report that they have served in the military, despite decades of not being allowed to serve openly. However, less than 2% of lesbians have served in the military. Sexual activity More than half of participants (58%) have been sexually active in past 12 months. Relationship status and living arrangements California is unique in that it temporarily granted marriage licenses to same-sex couples from June 16, 2008 to November 5, 2008 and those marriages continue to be recognized by the state. (Same-sex marriage resumed in California on June 28, 2013).The participants who did take advantage of that brief window of opportunity and were married may be experiencing some of the economic and health benefits that are associated with marriage (Gay and Lesbian Medical Association, 2008), as will be indicated throughout this report. Thirty-seven percent of the participants are married or partnered. Among the 63% who are not partnered or married, 80% are single, 11% are widowed, 7% are divorced, and 2% are separated. More than half of LGBT participants (58%) live alone. Fifteen percent have children; of these, 40% report that children are available to help to the participant and 60% are not. HIV/AIDS More than one third of the gay male participants are living with HIV/AIDS. Among all LGBT participants, 10% identify as HIVpositive, another 15% have AIDS. Three-quarters of the participants do not have HIV/AIDS.

1

We assumed that participants have no grandchildren in their household in the calculation of the California Elder Index. 11

Distinct demographics differences between lesbian, gay, bisexual, and transgender participants While the participants range in age from 60 to 92, gay men are on average older than lesbian participants. Lesbians and gay men aged 60 to 69 comprise the largest age group; those aged 80 and older are the smallest. Gay men are more likely than lesbians to have a high school education or less. They are also less likely to be working. Lesbians report higher household income than gay men. Lesbians are also more likely to be non-Hispanic white than gay men, and less likely to have served in the military. Lesbians are more likely than gay men to be in legally recognized relationships, both registered domestic partnerships and legal marriage. Gay men are more likely to live alone and are less likely to have children. Gay men are also more likely to be sexually active than lesbians. Bisexual women and men do not differ significantly from their lesbian and gay male peers in age, education, or employment. However they are more likely to have annual household incomes at or below 200% of the FPL. Only 8% have annual household incomes above $80,000, compared to 38% of lesbians, and 23% of gay men. Bisexual women and men are more likely to have children than lesbians and gay men. Transgender adults are on average younger than their non-transgender peers; none of the transgender participants are70-years old or older in the sample, while a third of non-transgender participants are in that age range. Transgender adults are less likely to be non-Hispanic white, are more likely to have only a high Demographic comparisons between HIV-positive and HIV-negative participants school education With HIV/HIV Without HIV/AIDS or less, and to Sexual orientation,% have incomes at Gay men 94.63* 63.39 or below 200% Lesbian 0.00 29.24 of the FPL. Distinct demographics differences between HIVpositive and HIVnegative participants Among LGBT adult participants, 25% are living with HIV/AIDS. Those living with HIV/AID are more likely to be gay men. Those with HIV/AIDS, compared to those without HIV/AIDS are

Bisexual 2.01 4.91 Heterosexual 1.34 0.89 Other 2.01 1.56 Gender identity, Transgender, % 4.73 4.00 Age, M(SD) 65.42 (4.65)* 68.33 (6.38) Race/ethnicity, % White (Non-Hispanic) 72.00* 80.71 Hispanic 15.33 4.88 African American 6.67 4.43 Asian American 2.33 4.66 Income, at or below 200% FPL, % 32.41 29.11 Education, some college or less, % 39.19* 26.99 Relationship status, % Partnered, legally recognized 8.72* 20.45 Partnered, not legally recognized 22.82 20.22 Not married or partnered 68.46 59.33 Living alone, % 65.33* 55.11 *Indicates statistically significant relationship between variables at p < .05.

12

also younger, less likely to be non-Hispanic white or to be in a legally recognized relationship, but more likely to be Hispanic, less educated, to live alone, and not be married or partnered. Summary Not unlike their heterosexual counterparts in San Francisco, LGBT participants show wide variability in the demographic contexts of their lives; for example, they are partnered or single; some have children, many do not. While many are well-off financially, many, especially bisexual and transgender women and men live with modest and often insufficient financial resources. LGBT participants also differ from their heterosexual peers in important ways. On average, they are better educated; although this does not appear to translate into commensurate earnings. They are also much more likely to live alone and less likely to have children. Many are in relationships, some legally recognized, some not, but the majority are neither partnered nor married. Significant numbers of men are living with HIV/AIDS. The purpose of this report is to provide community members, policymakers, and other interested parties with information regarding the needs, resources, risks, and experiences of LGBT older adults living in San Francisco. In the following sections (Services and Programs, Housing, Resources and Risks, and Health) we present both descriptive and statistical results of key indicators to foster a better understanding of the needs and resources of this historically marginalized and invisible population. To do this, we offer comparisons by sexual orientation and gender identity. Specifically, we first compare lesbians to gay men. Because of small sample sizes, we are unable to conduct separate analyses for bisexual women and men; instead we compare bisexual men and women together to lesbians and gay men. We also compare transgender adults to the non-transgender participants. While age, income, and education are important demographic factors determining needs, resources, and risks among older adults, sexual orientation and gender identity are associated with age, income, and education in this study as demonstrated above. We adopted statistical analyses that control for the influence of sociodemographic characteristics in the comparisons by sexual orientation and gender identity. When comparing findings by sexual orientation and gender identity, we only report those findings that remain statistically significant after adjusting for these sociodemographics. We conclude each section with a discussion of relationships between the variables in that section and background characteristics: gender, age, race/ethnicity, income, education, relationship status, living arrangement, and HIV/AIDS. Only significant differences are reported.

13

Services and Programs LGBT older adults often have unique needs due to past experiences of discrimination in medical, aging, and social services (see Resources and Risks section). In addition, many LGBT older adults do not have children to help them. While 15% of the participants report having children, 60% of those with children report that their children are not able to help them. Meanwhile, services and programs to assist older adults are frequently geared towards the general population, and cultural sensitivity training may or may not address LGBT-related issues. The sexual orientation and gender identity comparisons summarized below are based on statistical significance tests2 adjusting for age, income, and education. For sexual orientation comparisons: lesbians are compared to gay men; bisexual women and men are compared to lesbian and gay men. For gender identity comparisons, transgender women and men are compared with non-transgender women and men. A breakdown of services and programs findings by sexual orientation, gender identity, and background characteristics can be found in Appendix Table 2, and Figures 1 and 2. When comparing San Francisco’s LGBT older adult participants to older adults in San Francisco’s general population, some preliminary findings emerge that deserve additional attention:  LGBT participants cited the following six services and programs as the most needed: health services, health promotion services, mental health services, housing assistance, case manager/social worker, and telephone/online referrals; only one service, telephone/online referrals was identified as "most needed" by older adults in San Francisco (National Research Center, 2008).  The top five services and programs used by LGBT participants are: health services, mental health service/support groups, health promotion services, housing assistance, and case manager/social worker; again, only one service, the latter, was identified as "most used" by older adults in San Francisco (National Research Center, 2008). Most needed services and programs LGBT participants were provided a list of 14 services that have historically been the more frequently needed services among San Francisco’s older adults. Participants were asked to identify what services and programs they have needed in the past 12 months. An initial analysis identifies the most commonly identified services and programs from that list. Of the 14 services and programs considered, 50% of participants report needing health services. At least one out of five participants report needing the following seven services and programs:  Health services (50%)  Health promotion services (28%)  Mental health services (27%)  Housing assistance (24%)  Case manager/social worker (22%)  Telephone/online referrals (21%)  Meal site/free groceries (21%) 2

See Methodology for overview of tests of significance for these analyses. 14

The other listed services and programs for which participants report need at lower rates are: day programs (19%), in-home care (18%), door-to-door transportation (15%), homedelivered meals (13%), caregiver support (10%), veterans’ services (8%), and alcohol/substance abuse programs (6%). Reported needs for services and programs in the past 12 months were examined by sexual orientation and gender identity and reveal several trends:  The services and programs needed that are in common to all LGBT participants are health services, health promotion, mental health services, and housing assistance.  Other top services and programs vary by sexual orientation/gender identity, including: o In-home care (lesbians, bisexuals, transgender adults) o Door-to-door transportation (lesbians) o Day programs (gay men) o Caregiver support (transgender adults) Further analysis shows important differences in the rates of need for various services and programs. Bisexual and transgender participants report a higher rate of need for nearly all services and programs compared to lesbians and gay men. Gay men report higher rates of need for some services, compared to lesbians. Specifically,  For bisexuals, the need for health services, mental health services, meal site/free groceries, health promotion, and caregiver support are higher than those observed among lesbians and gay men.  Transgender adults report higher rates of need compared to non-transgender adults for in-home care, meal site/free groceries, day programs, mental health services, health promotion services, and caregiver support.  Gay men report significantly higher rates of need for certain services as compared to lesbians: meal site/free groceries, home-delivered meals, and veterans’ services.

15

Service use and comfort using services Among the 14 services and programs considered, the seven most used services largely mirror the needed services (described above) and are:  Health services (45%)  Mental health service/support groups (23%)  Health promotion (19%)  Housing assistance (19%)  Case manager/social worker (20%)  Meal site/free groceries (19%)  Telephone/online referrals (17%) Health promotion services had the largest discrepancy between need (28%) and use (19%). Further analyses revealed important differences in service use by sexual orientation and gender identity. Lesbian participants are less likely to use home delivered meals, meal site/free groceries, and veterans’ services than gay men. Bisexual participants are more likely to use mental health services, caregiver support, and health promotion services than lesbians and gay men. Transgender participants are more likely to use meal site/free groceries, day programs, mental health services, caregiver support services, and health promotion services than nontransgender adults.

A question was included that addressed the level of comfort associated with service or program use. The majority (more than 75%) of the participants who use one or more of the services or programs feel comfortable using each service or program as an LGBT older adult with three exceptions: Alcohol/substance abuse programs, housing assistance, and veterans’ services. 16

Services and programs needed but not used Services for older people need to be available, accessible, and acceptable. If these three criteria are not met, services will go unused (Wallace, 1990). We examined rates of unmet need – the proportion of participants not using a particular service even though they had indicated that they needed that service. Please note that the unmet need rates may be underestimated since those who used a service could possibly still have unmet need (e.g., needing in-home care daily but only receiving it twice weekly). Still, the unmet needs found in this survey have important implications for future services and programs. Overall when we look at all services, about 36% of the participants are experiencing unmet need. The seven highest rates of reported unmet need among the14 services and programs considered are for:  Health promotion (30%)  Door-to-door transportation (28%)  Caregiver support (27%)  Day programs (27%)  Housing assistance (21%)  In-home care (21%)  Telephone/online referrals (19%) Reason for not using services among those who needed services Not LGBT Too Difficult to May not Other Friendly expensive access qualify reasons % % % % % Meal site/free groceries 17.24 0.00 20.69 15.52 46.55 Telephone/online referrals 16.67 1.85 33.33 11.11 37.04 Alcohol/substance abuse program 8.70 13.04 17.39 8.70 52.17 Home-delivered meals 8.33 0.00 20.83 29.17 41.67 Mental health services 7.69 10.26 25.64 17.95 38.46 Day programs 6.78 8.47 28.81 16.95 38.98 Housing assistance 5.81 10.47 34.88 17.44 31.40 Health services 5.77 18.27 14.42 13.46 48.08 Case manager/social worker 5.17 3.45 31.03 13.79 46.55 Health promotion 5.06 13.92 18.99 15.19 46.84 In-home care 3.64 30.91 18.18 23.64 23.64 Caregivers support 3.13 9.38 25.00 21.88 40.63 Door-to-door transportation 1.82 27.27 14.55 27.27 29.09 Note. The results for veterans’ services are not shown due to insufficient data. Bold numbers indicate the most common reason of non-use for each service or program, besides “other.”

Reasons for not using needed services and programs Reasons for non-use of services and programs include difficulty in access, possibly not qualifying, being too expensive, not being LGBT friendly, and “other.” In most cases, the most common reason cited for not using services was “other,” which was unspecified. Of particular relevance to this report, meal site/free groceries (17%) and telephone/online referrals (17%) are cited as the least LGBT friendly compared to other services. 17

The services and programs deemed too expensive among the LGBT participants are inhome care, door-to-door transportation, and health services, despite the fact that many of those same participants have incomes that would make them eligible for free or low-cost services. Many services and programs are perceived as difficult to access. These include housing assistance, case manager/social worker, telephone/online referrals, day programs, and mental health services. About one in three LGBT participants think they may not qualify for homedelivered meals and door-to-door transportation. Just under one-fourth believe they may not qualify for in-home care or caregiver support. Participants reporting that they may not quality for a specific service or program may be a result of the potential user not understanding their eligibility or that eligibility criteria is not made clear by the service provider. Comparisons by sexual orientation and gender identity for reasons of non-use of services and programs were not conducted due to small sample size. Analysis of services and programs by key background characteristics In the following section we report on analyses of services and programs findings by background characteristics (gender, age, race/ethnicity, income, education, relationship status, living arrangement, and HIV/AIDS). Due to insufficient sample size, reasons for not using services were not analyzed; for unmet needs and comfort levels, overall summary scores for the 14 services are applied to the analyses. Only statistically significant findings are summarized below. Gender We find that men report higher levels of need for day programs than women, as well as home-delivered meals and veterans' services. Men are more likely to use home-delivered meals and veterans’ services than women. Age Older participants, especially those aged 80 and older, have a greater need for homedelivered meals, door-to-door transportation, day programs, and veterans’ services than their younger counterparts. In contrast, younger participants have a greater need for mental health services and housing assistance than their older counterparts. We found that use rates for health promotion services and housing assistance decrease as age increases. Older participants are more likely to feel comfortable using services and programs as an LGBT person than younger participants. Race and ethnicity Findings indicate that rates of needed services and programs are related to race/ethnicity, particularly meal site/free groceries, day programs, mental health services, alcohol/substance abuse programs, caregiver support, health promotion, and housing assistance. White, nonHispanic LGBT adult participants tend to have the lowest overall levels of need. They have the least need for meal site/free groceries, Asian Americans have the greatest. Asian American and Hispanic LGBT participants have the greatest need for day programs, African Americans the lowest. However, Hispanic and African American participants have more need of mental health services, and alcohol/substance abuse programs, while non-Hispanic whites have the least. African American participants also report significant need in housing assistance, as do Asian

18

Americans and Hispanics. LGBT Asian American participants have the highest need of caregiver support. The distributions of most service use rates by race/ethnicity reflect the patterns of service need rates described above; racial and ethnic differences in service use rates are found in meal site/free groceries, day programs, alcohol/substance abuse programs, caregiver support, health promotion, housing assistance, and a case manager/social worker. Income It should come as no surprise that income plays a significant role in the need for services and programs, especially among LGBT adults whose annual household incomes are at or below 200% of the federal poverty level (FPL). This low income group has more need for 10 service areas: in-home care, home-delivered meals, meal site/free groceries, door-to-door transportation, day programs, a case manager/social worker, mental health services, telephone/online referrals, health services, and housing assistance. This trend is similarly observed in the relationships between use rates and income at or below 200% of the FPL. Participants living in poverty are also more likely to use health promotion services. We find that the service use rate for day programs among those below poverty, despite their higher need for the service, are not different from the use rate among those above poverty level; this finding indicates that the need for day programs among those under poverty may be less likely to be met. In addition, when all the services are considered, participants at or below 200% of the FPL are more likely to experience unmet need than those above 200% of the FPL. Education In the general population, education is highly correlated with income (Marmot & Wilkinson, 2006), and as such, we would expect to find similar patterns of results examining education and needs. Indeed, participants who have less than a 4-year college degree have many needs that are similar to those whose incomes are at or below 200% of the FPL. Compared to their peers who have a 4-year degree or more, those with less education have more need for inhome services, home-delivered meals, meal site/free groceries, day programs, case manager/social worker, telephone/online referrals, and housing assistance. They also have a greater need for day programs and for veterans' services. Those with less education demonstrate higher use rates for in-home services, homedelivered meals, meal site/free groceries, case manager/social worker, housing assistance, and veterans' services. Relationship status The survey findings in terms of the association of service needs with relationship status provide evidence in support of the benefits of legal relationships among LGBT participants. Those who are in relationships that are not legally recognized tend to have rates of need similar to those who are not in relationships at all. Those in legally recognized relationships tend to have the lowest rates of need for services and programs, including in-home care, home delivered meals, meal site/free groceries, door-to-door transportation, day programs, case manager/social worker, mental health services, and housing assistance. Accordingly, we find that those in legally recognized relationships have the lowest service use rates in in-home care, home delivered meals, meal site/free groceries, day programs, 19

case manager/social worker, mental health services, and housing assistance. In addition, they show the lowest use for telephone/online referrals. Living arrangement LGBT adults who live alone predictably have greater service needs and accordingly greater service use than those who live with others. Those living alone use and have significantly more need for in-home care, home-delivered meals, meal site/free groceries, day programs, a case manager/social worker, and mental health services. HIV/AIDS Due to challenges of living with HIV, we might expect that those who are HIV-positive would have significantly greater need and use for almost all programs and services to support them in the community. We found this to be the case in only a few of the services and programs listed. Compared to their HIV-negative peers, LGBT participants living with HIV/AIDS have a greater need for a case manager/social worker, mental health services, and housing assistance, and greater use for a meal site/free groceries and mental health services. Summary About half of respondents report needing health services; other commonly needed services include health promotion, mental health services, housing assistance, case manager/social worker, telephone/online referrals, and meal site/free groceries. Other common needs observed from each sexual orientation and gender identity group include in-home care (lesbians, bisexuals, and transgender adults), door-to-door transportation (lesbians), day programs (gay men), and caregiver support (transgender adults). Overall, bisexual and transgender participants report elevated need for most services. Common unmet need areas include health promotion, door-to-door transportation, caregiver support, day programs, housing assistance, in-home care, and telephone/online referrals. Many LGBT participants report that inhome care, door-to-door transportation, and health services are too expensive to use and other services are difficult to access. In addition, nearly a fifth did not use meal site/free groceries and telephone/online referrals because they feel those services are not LGBT friendly. In addition, half of those who used alcohol/substance abuse programs and housing assistance report that they did not feel comfortable using the services as an LGBT older adult. Racial/ethnic minority participants report higher rates of service need: mental health services for Hispanics and African Americans, housing assistance and day programs for Hispanics and Asian Americans. In addition, LGBT participants having low incomes, low education levels, and living alone are more likely to need services; those in relationships that are not legally recognized and those who are not married or partnered also report heightened service need.

20

Housing Housing is a basic need, providing not only shelter, but also a refuge, respite, and place of safety. It is a place where family members and friends come together to provide and receive care, to be with one another, to relax and enjoy company. "A household consists of all the people who occupy a housing unit... [and] includes the related family members and all the unrelated people, if any.... such as partners or roomers..." (U.S. Census Bureau, 2012, para.22). Having an additional member in the household has been associated with decreased spending on medical care, as much as $1,040 annually for men aged 65 and older, and $729 for women 65 and older (Halliday & Park, 2009). The sexual orientation and gender identity comparisons summarized below are based on statistical significance tests3 adjusting for age, income, and education. For sexual orientation comparisons, lesbians are compared to gay men; bisexual women and men are compared to lesbian and gay men. For gender identity comparisons transgender women and men are compared with non-transgender women and men. A breakdown of housing findings by sexual orientation, gender identity, and background characteristics can be found in Appendix Table 3. When comparing San Francisco’s LGBT older adult participants to older adults in San Francisco’s general population, some preliminary findings emerge that deserve additional attention:  57% of older San Franciscans own a home (Ruggles et al., 2010), compared to 41% of LGBT participants.  41% of older adults in San Francisco rent (Ruggles et al., 2010), while 54% of LGBT participants do.  57% of older adults in San Francisco are very confident about staying in housing (Kim, Cannon, Cheh, Duda, & Hall, 2010), while only 32% of LGBT participants feel very confident. Household size The average size of LGBT participants' households is 1.48 persons (SD = .65). Lesbians have larger households than gay men. Housing arrangements The majority of LGBT participants in San Francisco (54%) rent their housing. Close to a third (28%) own their home and are making mortgage payments while just over 13% own their homes with their mortgages paid off. About 5% have some other type of financial arrangement that covers their housing, such as living rent free with a friend or relative, in a nursing home, or other health care facility.

3

See Methodology for overview of tests of significance for these analyses. 21

The significant differences in housing arrangement by sexual orientation and gender identity are as follows:  Gay men are more likely than lesbians to rent (56% vs. 36%) and less likely to own their home (40% vs. 55%)  Bisexual women and men are more likely than lesbians and gay men to rent (89% vs. 52%) and less likely to own their homes (7% vs. 43%)

Type of housing By far, most (88%) LGBT participants reside in a house, apartment, or condominium. Slightly less than 7% live in senior housing, assisted living facilities, nursing homes, or an agerestricted community. About 5% are domiciled in single room occupancy (SRO) residential hotels, or are homeless. Analysis of housing type by sexual orientation and gender identity revealed no significant differences. Confident about staying in current housing Even when people currently have housing, the extent to which they can continue in this same setting may be challenged by a variety of reasons, such as changes in health, economics, or safety. We asked participants how confident they are that they would be able to continue living in their current housing for as long as they liked. Only about a third (32%) of LGBT participants are very confident that they will be able to stay in their current housing for as long as they wish. There are no significant differences by sexual orientation or gender identity.

22

Possible need to relocate Reasons for relocating vary, as suggested above, and many participants offered multiple reasons why they might have to move from their current housing type. The most commonly cited reasons for the possible need to relocate are economic reasons (including risk of foreclosure) (54%), health reasons (44%), and changing needs related to aging, such as safety/grab bars, or elevators (40%).

Reasons for moving from current housing for LGBT older adult participants Economic reason(s), including risk of foreclosure

54%

Health reason(s)

44%

Different needs such as safety/grab bars or elevators

40%

Lack of stability in my housing situation

21%

Need to move out of San Francisco

12%

Rising crime rate in my neighborhood Friends moving elsewhere Unsafe environment as a result of being perceived as LGBT

11% 7% 6%

Want to move in with my family or friends 2%

When comparing the top six reasons that LGBT participants might move by sexual orientation and gender identity, the only difference that we found is that bisexual women and men (48%) are more likely than lesbians and gay men (19%) to cite instability in their current housing situation. Analysis of housing by key background characteristics In the following section we report on analyses of housing findings by background characteristics (gender, age, race/ethnicity, income, education, relationship status, living arrangement, and HIV/AIDS). Only statistically significant findings are summarized below. Gender In this study, women have larger average households than men. Women are more likely than men to own their home, to be paying a mortgage, and to have some other type of financial arrangement that covers their housing, such as living rent free with a friend or relative, in a nursing home, or other health care facility. Men are more likely to rent, and to cite rising crime rates as the reason they might need to move out of their current housing than women are. Age LGBT adult participants aged 60 to 69 have larger households than their older counterparts. There does not appear to be any age differences in housing arrangements (e.g., rent or own). LGBT participants aged 80 and older are more likely to live in senior housing, assisted living facilities, nursing homes, or an age-restricted community than younger peers. Those 80 and older are also more likely to feel confident that they will be able to stay in their current 23

housing for as long as they like. LGBT adults aged 60 to 69 are less likely than their older counterparts to cite their health as the reason they might need to move out of their current housing situation. Those aged 60 to 69 are the most likely to cite a need to move out of San Francisco. Race and ethnicity LGBT Asian Americans and non-Hispanic white participants are more likely than African Americans and Hispanics to own their home, and to have their mortgages paid off. NonHispanic whites are the least likely to have some other type of financial arrangement that covers their housing, such as living rent free with a friend or relative, in a nursing home, or other health care facility. Hispanics are the most likely to cite rising crime rates as the reason they might need to move out of their current housing situation; Asians Americans are the least likely. Income Compared to their more affluent peers, LGBT participants whose incomes are at or below 200% of the FPL are less likely to own their home, to have their mortgage paid off, or to be paying a mortgage. They are more likely to be renting. Those with lower incomes are less likely to live in a house or apartment, and are more likely to live in senior housing, assisted living facilities, nursing homes, or an age-restricted community. They are also more likely to be domiciled in a SRO or to be homeless. Those with lower incomes are more likely to consider having to move out of their current housing situation due to changing needs resulting from aging, instability in their current housing situation, and rising crime rates than higher income participants. Education Like those with lower incomes, LGBT participants who have less than a 4-year college degree have smaller households than those with a 4-year degree or more. They are less likely to own their home, to have their mortgage paid off, or to be paying a mortgage. They are more likely to be renting, and to have some other type of financial arrangement that covers their housing. LGBT participants with less education are also less likely to live in a house or apartment, and are more likely to live in a SRO or be homeless. Those with less education are more likely to consider possibly having to move out of their current housing situation due to their health, changing needs resulting from aging, and rising crime rates than those with more education. Relationship status Participants in legally recognized relationships have larger households than their counterparts who are neither partnered nor married. They are more likely to own their home, to have their mortgage paid off or to be paying a mortgage, and less likely to be renting. Compared to those who are not partnered or married, those in legally recognized relationships are also more likely to live in a house or apartment; none report that they live in a SRO or are homeless. They are also more likely to be confident that they will be able to stay in their current housing for as long as they like.

24

Living arrangement LGBT participants who live alone are more likely to rent. They are less likely to own their home, to have their mortgage paid off, or to be paying a mortgage, and to have an alternative housing arrangement, such as living rent free with a friend or relative, in a nursing home, or other health care facility. LGBT adults who live alone are less likely to live in a house or apartment, and are more likely to live in senior housing, assisted living facilities, nursing homes, or an age-restricted community. They are also more likely to be domiciled in a SRO or to be homeless, and less likely to be confident that they will be able to stay in their current housing for as long as they wish. Those who live alone are more likely than those living with others to cite the need to move out of San Francisco as the reason they might need to move out of their current housing situation. HIV/AIDS There do not appear to be any differences in household size or housing arrangement between HIV-positive gay, bisexual, and transgender men and their HIV-negative counterparts. However, those who are HIV-positive are more likely to live in a house or apartment than those who are HIV-negative. Summary Lesbian participants have larger households and are more likely to own their homes than gay men. Very few bisexual and transgender women and men own their homes, and are much more likely to be renting. Most LGBT older adult participants live in an apartment or house, but some are living in SROs or are homeless. About a third of the LGBT participants are very confident that they will be able to stay in their current housing for as long as they wish. The most commonly cited reasons for the possible need to relocate are economic reasons, health reasons, and changing needs related to aging; men are more likely than women to cite rising crime rates as the reason for their current housing instability; transgender participants are more likely to cite economic reasons, and along with bisexuals, they are also more likely to acknowledge instability in their current housing situation.

25

Resources and Risks Resources and risks can influence one’s well-being. Resources can take many forms such as social support, spirituality, and planning for future needs, and we can draw on our resources in times of need. Risks, such as discrimination and abuse, can have a negative impact on physical and mental health. In this section we look at both resources and risks of LGBT older adults. The sexual orientation and gender identity comparisons summarized below are based on statistical significance tests4 adjusting for age, income, and education. For sexual orientation comparisons lesbians are compared to gay men; bisexual women and men are compared to lesbians and gay men. For gender identity comparisons transgender women and men are compared with non-transgender women and men. A breakdown of resource and risk findings by sexual orientation, gender identity, and background characteristics can be found in Appendix Tables 4 and 5. When comparing San Francisco’s LGBT older adult participants to older adults in San Francisco’s general population, some preliminary findings emerge that deserve additional attention:  Of the older adult population in San Francisco, 16% report that no one is available to understand their problems; 29% have no one to help with daily chores when sick (CHIS, 2003); 9% of LGBT participants have no one to turn to for support, encouragement, or short term help.  Estimates of past-year elder abuse/neglect range from 8% to 10% with only about 7% of cases being reported (National Center on Elder Abuse & Administration on Aging, n.d.). About 5% of LGBT participants have experienced past-year abuse/neglect, 21% have been victimized; 28% of those who experienced abuse or victimization reported it to authorities. Resources As human beings, we are by definition social creatures. The very heart of this concept is found in the Nguni word ubuntu. Roughly translated, ubuntu means, ‘we are only fully and truly human through our relationships with others.’ Social support is ubuntu in action. It is also critical to our health and well-being, especially among older adults (Barker, Herdt, & de Vries, 2006; Netuveli, Wiggins, Hildon, Montgomery, & Blane, 2006; Strine, Chapman, Balluz, Moriarty, & Mokdad, 2008). 4

See Methodology for overview of tests of significance for these analyses. 26

Social support LGBT participants were asked how often they received the emotional and social support they needed; responses ranged from never = 1 through always = 5; higher mean scores are indicative of greater levels of social support. LGBT participants appear to enjoy moderate levels of social support (M = 3.67, SD = .96). Lesbians report Levels of social support by sexual orientation and gender identity significantly higher levels of Standard support than gay men. Levels Mean Deviation of social support for bisexuals Lesbians 3.83‡ .72 are similar to those of lesbians Gay Men 3.65 1.01 and gay men, as are the levels Bisexual Women and Men 3.64 .91 between transgender adults Transgender Women and Men 3.52 1.08 and their non-transgender ‡ Indicates that a p value remains <.05 in adjusted logistic peers. regressions after controlling for age, income, and education

Social support networks To better understand how social support operates in their lives, we asked LGBT participants who comprised their social networks -- whom they turned to for support, encouragement, or short-term help, such as running an errand or getting a ride. Almost three-quarters (72%) of the participants indicate a “close friend” as their most common source of social support. Next most common are partner or spouse (36%), therapist (23%), and neighbor (22%). Analysis by sexual orientation and gender identity reveals the following findings:   

Gay men are more likely than lesbians to have no one to whom to turn for social support. Lesbians are more likely than gay men to turn to a partner or spouse, a family member, or a neighbor for social and emotional support. Transgender adults are more likely than non-transgender adults to turn to faith communities for social and emotional support.

Spiritual and religious engagement Engaging in spiritual or religious practices or activities has also been found to be a protective factor in health and quality of life among older adults (Fiske, Wetherell, & Gatz, 2009; McCullough & Laurenceau, 2005; Solomon, Kirwin, Van Ness, O'Leary, & Fried, 2010). It has also been associated with slower progression of impaired immune system functioning and less psychological distress among those living with HIV (Ironson et al., 2002). We asked LGBT participants whether they had attended faith, spiritual or religious services or activities in the past 27

30 days. Overall, 31% of the LGBT participants attended such services or activities including: lesbians (33%), gay men (29%), bisexuals (36%), and transgender participants (58%). The differences by sexual orientation and gender identity are not statistically significant. Identity disclosure Sexual orientation and gender identity are marginalized social identities that are, to varying degrees, concealable. Concealing one's identity can have short-term benefits, making one a less visible target for discrimination, but in the Levels of identity disclosure by sexual orientation and gender long-term, hiding one’s identity identity increases Standard psychological distress Mean Deviation (Meyer, 2003). Disclosing Lesbians 3.51 .69 one's marginalized identity Gay Men 3.57 .64 provides potential access ‡ Bisexual Women and Men 3.07 .83 to important psychological ‡ Transgender Women and Men 3.23 .91 and social resources. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions LGBT-identified after controlling for age, income, and education. individuals have increased opportunities for social interactions with "like" others, which can be socially supportive; comparing one's self to like others also helps LGBT individuals to be more self-accepting (Meyer, 2003). LGBT participants were asked to what extent they openly identify, or are “out”, about their sexual orientation or gender identity to others. Possible responses ranged from not at all = 1, to completely = 4. Higher mean scores are indicative of higher levels of identity disclosure, or being more "out." Overall, LGBT participants appear to have relatively high levels of identity disclosure – to be quite “out” (M = 3.52, SD = .68). Bisexual participants have significantly lower levels of identity disclosure than do lesbians and gay men; transgender adults have lower levels than non-transgender adults. Future planning There are more than 1,100 rights, privileges, and benefits that are automatically conferred on couples in federally recognized marriages (General Accounting Office & Office of the General Counsel, 1997). Many of these benefits make less pressing the need for future planning, such as the right to make medical decisions or funeral arrangements for an ill or deceased spouse, or automatic inheritance and exemption from estate taxes. Even with the recent Supreme Court rulings in favor of marriage equality and federal recognition of same-sex marriage, it is important to note that the federal government does not recognize domestic partnerships 28

or civil unions that are granted at the state level (National Conference of State Legislatures, 2013), and only 13 states (including California) grant marriage licences to same-sex couples. LGBT couples who are not legally married must take time and often considerable expense to put legal instruments into place. Such legal instruments become increasingly important as we age. We asked participants whether they have any future planning items in place. Power of attorney for healthcare About twothirds (61%) of LGBT participants have a durable power of attorney for health care, which designates another adult to make health care decisions should one become incapacitated. Lesbians (72%) are more likely than gay men (60%) to have a power of attorney for healthcare. Wills About half (53%) of LGBT participants have executed a will. Bisexual participants (12%) are significantly less likely to have wills than lesbian and gay adults (55%). Power of attorney for finance A power of attorney for finance allows an individual to designate someone to manage their finances should they become unable. About a third of LGBT participants (30%) have such an instrument in place. Lesbians (38%) are more likely than gay men (29%) to have a power of attorney for finance. Revocable/irrevocable trust Revocable and irrevocable trusts are a mechanism by which assets are protected and distributed after an individual's death, based on the wishes and guidelines put in place by the person establishing the trust. About one-third (29%) of LGBT participants have such a trust in place. Lesbians (38%) are more likely to have a trust than gay men (28%). Other future planning About one in four LGBT participants (23%) has already made funeral plans and about one in six (16%) participants have long-term care insurance. Slightly more than one in ten (11%) have made arrangements for a charitable legacy gift. Less than one in ten (8%) have an informal caregiving arrangement in place. Further analysis showed that there are no significant differences by sexual orientation and gender identity for these other types of future planning. 29

Risks In order to better understand the aging and need of LGBT older adults, it is important to take account of their experiences that may harm their health and well-being, such as discrimination and abuse. This section examines such stressful life events that LGBT participants may face in their everyday life. A breakdown of risk factors by sexual orientation, gender identity, and background characteristics can be found in Appendix Table 5. The sexual orientation and gender identity comparisons summarized below are based on statistical significance tests5 adjusting for age, income, and education. Discrimination Discrimination has been linked to poor physical and mental health outcomes in the general population (Ahmed, Mohammed, & Williams, 2007; Williams, Yu, Jackson, & Anderson, 1997), and among older LGBT adults (Fredriksen- Goldsen, CookDaniels, et al., 2013; FredriksenGoldsen, Kim, et al., 2013). Although San Francisco has a reputation as an LGBT friendly city, reports of discrimination are not infrequent. LGBT participants were asked if in the past 12 months they had been treated unfairly, with less respect than others are treated, or discriminated against due to their gender identity or sexual orientation. Overall, 44% have experienced such discrimination. Transgender participants (81%) experienced significantly more discrimination than their non-transgender counterparts (42%) LGBT participants were also asked if in the past 12 months, they had been treated unfairly, with less respect than others are treated, or discriminated against for other reasons. Overall, 48% of the participants have experienced one of the following types of discrimination:  Age (34%)  Gender (13%)  Disability (16%)  Race (11%)  Socioeconomic status (10%)  Ancestry or national origin (5%)  Language (2%)  Immigration status (1%) Transgender adults experience higher rates of discrimination based on gender (69%), race (38%), and disability (38%) than their non-transgender counterparts.

5

See Methodology for overview of tests of significance for these analyses. 30

Bisexual women and men experience higher rates of gender-based discrimination (30%) than lesbians and gay men (11%). Lesbians are more likely to experience gender-based discrimination (27%) and age discrimination (41%) than gay men (5% and 31%, respectively). Where discrimination is experienced Discrimination takes place in many settings. Although the majority (58%) of LGBT participants experience discrimination due to their sexual orientation or gender identity in public venues, such as in a store, on the sidewalk, or taking public transportation, many have also experienced discrimination in other important settings.

Rates of discrimination in specific settings due to sexual orientation or gender identity Public place 58% Job or place of employment 19% Medical or health services 17% Faith, spiritual, or religious setting 15% Housing 14% Aging services 10% Other social services (not aging-related) 10% Interaction with police 10%

Abuse and victimization Unfortunately, all older adults can be vulnerable to elder abuse. Those who experience abuse are at increased risk for poor physical and mental health outcomes; their risk of mortality increases by a factor of three (National Center on Elder Abuse & Administration on Aging, n.d.), and the risk for nursing home placement by a factor of four (National Center on Elder Abuse, 2011). Abuse takes many forms, but in order to be classified as abuse, these acts must be committed by someone in a biological, legal, or other trusting relationship with the victim (e.g., family member, partner, friend, caregiver), or in an institutional setting (National Center on Elder Abuse, 2011). When committed by a stranger or someone else, these horrendous acts are classified as criminal victimization. Physical abuse includes being physically hurt, pushed, punched, or assaulted in any way, or being physically threatened by someone. Harassment includes being controlled by another. Verbal abuse also encompasses being threatened by someone. In addition to coerced sexual activity, being touched, grabbed, or groped without consent constitutes sexual abuse. Economic or financial abuse occurs when one feels forced or is tricked into giving someone money or property. Abuse and victimization experienced by LGBT older adult participants Neglect, being left without Abuse by friend, partner/spouse, Victimization by stranger Type of abuse basic needs, family, paid caregiver and other such as food, Verbal abuse 3% 14% water, or Harassed 3% 9% medications by Financially exploited 2% 4% a caregiver also Sexually abused 1% 4% falls under the Neglected 1% 2% rubric of abuse. Physically abused 1% 1% LGBT participants were asked if they had experienced these types of abuse in the past 12 months. Overall, 5% report that they have experienced at least one kind of abuse, and 21% have been victimized. Verbal abuse and victimization are the most common. 31

Among the LGBT participants who were abused in the past 12 months, nearly half (43%) were abuse by a friend, 27% by a partner or spouse, and 20% each by a family member or paid caregiver.

Perpetrators of abuse of LGBT older adult participants Friend 43% Partner or spouse 27% Family member 20% Paid caregiver 20%

Why abuse and victimization are not reported Only about one in four (28%) of LGBT participants who have been abused or victimized actually reported the crime to the authorities. The reasons given by those who did not report such experiences include:  Too ashamed (21%)  Didn't know how (18%)  Didn't trust authorities to treat LGBT people fairly (9%)  Afraid doing so would require disclosure of sexual orientation or gender identity (2%)  Fearful because of immigration status (2%)  Other reason (48%) (not specified)

Analysis of resources and risks by key background characteristics In the following section we report on analyses of resources and risks findings by background characteristics (gender, age, race/ethnicity, income, education, relationship status, living arrangement, and HIV/AIDS). Only statistically significant findings are summarized below. Gender Men in this study are more likely than women to have no one to whom they can turn for support. Women are also more likely than men to turn to a partner or spouse, a family member, a neighbor, and their faith community for social and emotional support. Women are more likely than men to have a power of attorney for health care. Women are more likely than men to experience gender and age discrimination, and to have been victimized in the past 12 months. Age LGBT participants aged 60 to 69 are more “out” about their sexual orientation and gender identity than their older peers. They are less likely to have executed a will, have a power of attorney for health care, a revocable or irrevocable trust, or have funeral plans in place. They also report more discrimination based on sexual orientation and gender identity than their older counterparts, as well as discrimination based on race and economic status. Participants aged 60 to 69 also report higher rates of past-year victimization than their older counterparts. LGBT participants aged 80 and older in this study are less likely than their younger peers to turn to a therapist for social and emotional support. Race and ethnicity LGBT Hispanic participants are the least likely to turn to a partner or spouse for social support, while non-Hispanic white participants are the most likely to do so. Non-Hispanic whites attend religious or spiritual activities or services at the lowest rate, while Asian Americans attend at the highest rate. LGBT Hispanic and African American participants are less “out” than non-Hispanic white participants. Non-Hispanic whites are the most likely to have a will, power of attorney for 32

healthcare, revocable/irrevocable trust, and a power of attorney for finance; African Americans are the least likely to have any of these. African American participants report higher rates of discrimination based on gender identity, sexual orientation, race, and gender, while non-Hispanic white participants report the lowest. Non-Hispanic whites report the lowest levels of abuse among racial and ethnic groups. Income LGBT participants whose incomes are at or below 200% of the FPL have less social support than those with higher incomes. They are also less likely to seek social support from a partner or spouse, close friend, or family member, but are more likely to turn to social services. Compared to those with higher incomes, lower income participants are less likely to have a will, power of attorney for healthcare, a power of attorney for finance, a revocable or irrevocable trust, or long-term care insurance, and less likely to have made a charitable legacy gift. Lower income LGBT participants are also more likely to experience discrimination based on gender identity, sexual orientation, age, disability, race, and socioeconomic status. They are also more likely to experience abuse and victimization. Education LGBT participants who have some college education or less have less social support than those with a 4-year college degree or more. They are also more likely to have no one to whom they can turn for support. When seeking social and emotional support, they are less likely than those with a 4-year degree or more to turn to a partner or spouse, a close friend, a family member, or a therapist, but are more likely to turn to social services. Compared to LGBT adults with a 4-year degree or more, those with less than a 4-year degree are less likely to have a will, power of attorney for healthcare, a power of attorney for finance, a revocable or irrevocable trust, to have made a charitable legacy gift, or long-term care insurance. They also experience more race-based discrimination and abuse. Relationship status LGBT adults in relationships enjoy the highest levels of social support, regardless of whether those relationships are legally recognized, while those who are not in a relationship have less social support. Those not in relationships are more likely to have no one to whom they can turn for support, compared to those who are. Obviously, those in relationships are more likely than those who are not to turn to a partner or spouse for emotional and social support. LGBT participants in legally recognized relationships are the least likely to turn to social services, compared to those in relationships that are not legally recognized, or those not in relationships. LGBT adults in legally recognized relationships are more likely to be out about their sexual orientation or gender identity. They are also more likely to have executed a will, power of attorney for healthcare, a power of attorney for finance, a revocable or irrevocable trust, and long-term care insurance. Those in legally recognized relationships report the least past-year victimization, compared to those in relationships that are not legally recognized, or those not in relationships.

33

Living arrangement LGBT participants who live alone have lower levels of social support than those who live with others. They are also more likely to have no one to turn to for emotional and social support. When they do seek out emotional and social support, they are less likely than those who live with others to turn to a partner or spouse, or a family member, but are more likely to turn to social services. LGBT adults living alone are less open about their gender identity or sexual orientation than those living with others. They are also less likely to have a will in place, have a power of attorney for healthcare, a power of attorney for finance, or a revocable or irrevocable trust. They also report more discrimination based on socioeconomic status. HIV/AIDS Gay, bisexual, and transgender men living with HIV/AIDS are more likely than their HIV-negative peers to seek social and emotional support from a therapist and from social services. They are more likely to have a power of attorney for health care, but are less likely to have long-term care insurance than their HIV-negative peers. They are more likely to have experienced discrimination because of a disability and because of their lower socioeconomic status. Summary In terms of resources, LGBT participants appear to have moderate levels of social support and relatively high levels of identity disclosure. About two-thirds have a durable power of attorney for healthcare, half have executed a will, less than a third have a power of attorney for finance in place, and fewer than one in four have made funeral plans. Lesbians have more social resources than gay men. Faith communities are an important source of social support among transgender participants. Lesbians and gay men tend to have relatively high levels of disclosure, bisexual and transgender much less so, which may place them at increased risk for limited social support. Lesbian participants are more likely than gay men to have a durable power of attorney for healthcare and a power of attorney for finance; bisexual participants are much less likely than lesbians and gay men to have a will in place. In terms of risks, almost half of LGBT participants have been discriminated against in the past 12 months because of their gender identity or sexual orientation, and more than a third because of their age. Lesbians are more likely than gay men to experience age-based discrimination. Compared to their non-transgender peers, transgender participants are also at heightened risk for discrimination based on sexual orientation/gender identity, age, race, and disability. LGBT participants are at a greater risk of being abused by friends than they are by partners/spouses, family members, or paid caregivers. Participants who have lower incomes, less education, who are neither partnered nor married, and who live alone are all at heightened risk of limited social support. African Americans, lower income participants, and those living with HIV/AIDS are at heightened risk for discrimination. LGBT participants in legally recognized relationships have fewer risks than those in non-legally recognized relationships, or those who are neither married nor partnered.

34

Health The Centers for Disease Control and Prevention (CDC) defines health as "a state of complete physical, mental, and social well-being and not just the absence of sickness or frailty" (Centers for Disease Control and Prevention, 2011b, para. 5). The link between mental and physical health is often clouded, especially when physical health promotion is at issue (Sturgeon, 2006). The sexual orientation and gender identity comparisons summarized below are based on statistical significance tests6 adjusting for age, income, and education. For sexual orientation comparisons, lesbians are compared to gay men; bisexual women and men are compared to lesbian and gay men. For gender identity comparisons, transgender women and men are compared with non-transgender women and men. A breakdown of health findings by sexual orientation, gender identity, and background characteristics can be found in Appendix Table 6. When comparing San Francisco’s LGBT older adult participants to older adults in San Francisco’s general population, some preliminary findings emerge that deserve additional attention:  42% of LGBT participants have an ambulatory/physical disability, compared with 22% of older adults in San Francisco (Ruggles et al., 2010).  15% of LGBT participants have contemplated suicide in the past year; about 13% of San Francisco’s older adults have ever contemplated suicide (CHIS, 2009).  Only 83% of LGBT participants report regular doctor visits in the past 12 months, compared to 91% of older adults in San Francisco (CHIS, 2009). General health Most LGBT participants report being in good health, although close to one in three (29%) evaluate their overall general physical health as poor. Nearly one third of gay men (30%), 20% of lesbians, and 24% of bisexual adults have poor general health. Transgender participants are more likely (52%) to report poor health than their non-transgender counterparts (28%). Although 33% of the male participants have HIV, there is no significant difference in overall health status between female and male participants. Physical disability A physical disability involves having a condition that substantially limits one or more basic physical activities, such as walking, climbing stairs, reaching, lifting, or carrying. Just less than half of LGBT participants (42%) report having a physical disability, including 37% of lesbians, 42% of gay men, and 50% of bisexual women and men. Transgender participants are more likely to report physical disabilities than their non-transgender counterparts; in the survey, three-quarters (76%) of transgender participants report physical disabilities, compared to 41% of non-transgender participants. Frequent limited activities due to poor mental health Even if a person does not have a diagnosed mental illness, mental health issues can be debilitating. Participants were asked how many of the past 30 days they are limited in any way, 6

See Methodology for overview of tests of significance for these analyses. 35

in any activities, because of emotional or mental problems. An individual who experiences such limitations for 14 or more days out of 30 can be characterized as having poor mental health (Moriarty, Zack, Holt, Chapman, & Safran, 2009). The majority of LGBT participants report doing well psychologically, but about 8% can be classified as experiencing frequent limited activities due to poor mental health, including 6% of lesbians, 8% of gay men, 17% of bisexuals, and 26% of non-transgender adults. Suicidal ideation Older adults in the general population have the highest risk of suicide of any age group (Cukrowicz et al., 2009). The risk may be even greater for LGBT older adults; discrimination and mental health issues are found to increase the risk of suicide further (de Graaf, Sandfort, & ten Have, 2006). In this study, one in seven (15%) of LGBT participants report having seriously considered taking their own lives in the past 12 months. Among the participants, 13% of lesbians, 14% of gay men, 16% of bisexual women and men, and 32% of transgender adults seriously contemplated suicide in the past 12 months. HIV/AIDS Due to the advent of anti-retroviral therapies that have transformed HIV from a nearly universally lethal disease into one that is chronic and manageable, it is estimated that by 2015 half of the 1.2 million HIV-positive Americans will be aged 50 and older (High, Brennan-Ing, Clifford, Cohen, & Deeks, 2012). About 25% of LGBT participants are living with HIV/AIDS. The difference in rates of HIV/AIDS between lesbians and gay men is marked; while none of the lesbian participants are HIV-positive, a third (33%) of gay male participants are living with HIV. Health services used in preceding 12 months LGBT participants shared the types of health services that they have used in the past 12 months. Usage of health services varied by sexual orientation and gender identity. In addition to community health centers, transgender adults are more likely than their non-transgender counterparts to have used medication management services (46% vs. 12%), and chronic illness management services (42% vs. 16%).

36

Health insurance Health insurance is obtained through the public and private sectors. Public insurance is funded through various levels of the government, and includes Medi-Cal, Medicare, Veterans Health Administration (VHA), and Indian Health Insurance (IHA). Private insurance is generally paid by employers or by individuals and families (private pay). Sixteen percent have Medicare only; another 16% have Medicare and Medi-Cal; 33% have Medicare and other including private insurance; 3% have Medi-Cal; 28% have private insurance; 2% have VHA. Analysis of health indicators by key background characteristics In the following section we report on analyses of health indicators by background characteristics (gender, age, race/ethnicity, income, education, relationship status, living arrangement, and HIV/AIDS). Only statistically significant findings are summarized below. Gender Men participants have higher rates of HIV/AIDS than women. Overall, health status between lesbians and gay men are similar in spite of the HIV/AIDS difference. Women engage in insurance counseling at lower rates than men. Age The only age-related difference noted in health outcomes is that LGBT participants aged 70 and older have lower rates of HIV/AIDS than those aged 60 to 69. LGBT adults aged 60 to 69 are less likely than their older peers to use emergency room services, but more likely to visit a community health center. Those 60 to 69 are more likely to access insurance counseling and use HIV services. Race and ethnicity The only racial or ethnic difference identified in health outcomes is that LGBT Hispanic participants have the highest rate of living with HIV/AIDS, while Asian Americans have the lowest. LGBT Hispanics and African Americans utilize community health centers at higher rates than non-Hispanic whites. Income Compared to those with higher incomes, LGBT participants whose annual household incomes are at or below 200% of the FPL are more likely to be in poor general health and have a physical disability, frequent limited activities due to poor mental health, and suicidal ideation. They are more likely than those with higher income to have used health services in the past 12

37

months. These include community health centers, medication management, chronic illness management, and emergency room visits as well as HIV services. Education Compared to LGBT participants with a 4-year college degree or more, those with less than a 4-year degree are at greater risk of poor general health, physical disability, frequent limited activities due to poor mental health, and HIV/AIDS. They are more likely to use community health centers and medication management. Relationship status LGBT participants who are not partnered or married are more likely to report poor general health than those who are in relationships, regardless of whether those relationships are legally recognized or not. Those not in relationships are at greater risk of physical disability than those in legally recognized relationships. LGBT adults in legally recognized relationships have significantly lower rates of HIV/AIDS and past-year suicidal ideation than those in relationships that are not legally recognized, or those not in a relationship. LGBT adults in legally recognized relationships are less likely to have accessed community health centers than those whose relationships are not legally recognized, or those not in relationships. Living arrangement LGBT participants who live alone are at risk for poorer outcomes on all assessed health indicators, compared to those who live with others. Those who live alone are more likely to have poor general health, a physical disability, poor mental health, HIV/AIDS, and past-year suicidal ideation. HIV/AIDS Compared to their HIV-negative counterparts, LGBT participants living with HIV/AIDS are more likely to have physical disabilities, and to use community health centers, medication management, chronic illness management, and insurance counseling. Summary The majority of LGBT participants enjoy good physical and mental health; even still, there are significant rates of poor general health, physical disability, HIV/AIDS, and a number of LGBT adults have seriously considered taking their own lives in the past 12 months. Those generally at most risk of poor health are transgender participants, those whose incomes are at or below 200% of the FPL (i.e., lower income), those with less than a 4-year college degree (i.e., less education), and those living alone. LGBT participants in legally recognized relationships appear to be at the lowest risk. Nearly all LGBT participants have health insurance.

38

Key Findings The participants in this study were largely recruited via the media, agency contact lists, community events, and outreach efforts. As a result of the targeted outreach to obtain a demographically diverse sample, the findings in this report cannot be generalized beyond the participants in the study. In addition, although outreach to diverse and underrepresented communities was an important focus of the study, recruitment was difficult and more work is needed to increase representation of African Americans, Asian/Pacific Islanders, Hispanics, bisexuals, and transgender older adults. The LGBT participants in this study share many of the same health and aging needs as older adults in general; at the same time, they also display important differences. The participants in this study are more likely to live alone and only 15% of the participants have children, and among those with children 60% report that their children are not available to help them. Fully one-third of the gay and bisexual men are living with HIV/AIDS. Bisexual and transgender participants and those living with HIV/AIDS are also more likely to experience poverty. Participant responses indicate a diverse array of needs and use of health and social services. Half of respondents needed health services in the past 12 months; other commonly needed services include health promotion, mental health services, housing assistance, having a case manager/social worker, telephone/online referrals, and meal site/free groceries. Overall, bisexual and transgender participants report elevated need for most services. Racial and ethnic minority participants report higher rates of need for some specific services: mental health services for Hispanics and African Americans, and housing assistance and day programs for Hispanics and Asian Americans. In addition, LGBT participants with lower incomes and lower educational attainment, as well as those living alone, and those not partnered or married are more likely to need an array of services. There are several areas of need in services and programs that may warrant special attention: health promotion services, door-to-door transportation, caregiver support, day programs, housing assistance, in-home care, telephone/online referrals, and health services. About one out of six of the participants who needed but did not use meal site/free groceries and telephone/online referrals report that the reason is that they perceive them as not being LGBT friendly. In addition, half of those who use alcohol/substance abuse programs and housing assistance indicate that they do not feel comfortable utilizing these services as an LGBT older adult. Many LGBT participants believe that in-home care, door-to-door transportation, and health services are too expensive to use, and that several other services and programs are too difficult to access. The need for safe, stable, and affordable housing was identified as an important need in San Francisco. Many are concerned that they may need to relocate due to economics, health, and changing needs associated with aging. Bisexual, African American, and Hispanic participants are the least likely to own a home. LGBT respondents who live alone, those who have lower incomes, and those with less formal education are at particular risk for housing instability, while those in legally recognized relationships appear to have more stability across the differing types of housing. The LGBT older adults also evidenced important strengths and resources. Most of the participants appear to have moderate levels of social support. Lesbian participants enjoy higher levels of social support than gay men; gay men are more likely than lesbians to have no one to turn to for support. Lesbians are also more likely than gay men to seek social support from 39

family members and neighbors. LGBT participants in legally recognized relationships report fewer risks and more resources than those in non-legally recognized relationships or those who are single. While the recent Supreme Court decision making marriage equality again a reality in California is an important step forward, the benefits may be limited for the older adult participants in our study since most are single. Support was also sought, and received, from various other sources. Faith communities are an important source of social support particularly among transgender and Asian participants. Those living with HIV/AIDS are more likely than those who are HIV-negative to seek social support from formal sources, such as social service agencies and therapists while they have similar levels of social support from informal sources. LGBT participants who have lower incomes, those with less education, those who are neither partnered nor married, and those who live alone are all at heightened risk of limited social support. Disclosing one's sexual orientation or gender identity may create opportunities for social support. Lesbian and gay male participants tend to have relatively high levels of disclosure; bisexual and transgender participants much less. Despite living in a city known for its support of the LGBT community, almost half of LGBT participants were discriminated against in the past 12 months because of their gender identity or sexual orientation. More than one-third experienced discrimination as a result of their age; with lesbians more likely than gay men to experience age-based discrimination. Female, transgender, African American, and lower income participants are at heightened risks for various types of discrimination. In addition, those living with HIV/AIDS are more likely than those living without HIV/AIDS to experience discrimination based on disability and poverty. Overall, 21% of the LGBT participants have been victimized and 5% reported that they have experienced at least one type of abuse in the past year. They are at a greater risk of being abused by friends, but abuse also occurred by partners/spouses, family members, and paid caregivers. Unfortunately, only about one-quarter who had been abused or victimized reported the crime to the authorities for reasons such as lack of knowledge and trust. Poor physical and mental health can have serious consequences for older adults as they age. Nearly one third of participants report poor general health; more than 40% have one or more physical disabilities. A third of gay men are living with HIV/AIDS. In terms of mental health, one in ten participants is experiencing frequent limited activities due to poor mental health. Fifteen percent report having seriously considered taking their own lives in the past 12 months. Transgender participants, those living in poverty, and those not married or partnered are more likely to report poor health. Receiving preventive health care is crucial for older adults, but approximately one- fifth of the LGBT participants did not make a regular doctor office visit. Trips to the emergency room are generally acute in nature, and costly; more than one in five participants visited a hospital emergency room during the past 12 months. Few LGBT participants have the full complement of legal arrangements in place to manage serious illness or end-of-life care. Even though six out of ten have a durable power of attorney for healthcare and half have executed a will, less than a third have a power of attorney for finance, and less than one in four have made funeral plans. Lesbian participants are more likely than gay men to have completed planning documents. In addition, bisexual participants are less likely than lesbians and gay men to have a will. Among the racial and ethnic groups, African Americans have the lowest rates of future planning. LGBT participants who are not married or partnered report lower rates of having a will, power of attorney for health care and finance, and revocable/irrevocable trust than those who are partnered or in legally recognized relationships.

40

Recommendations Existing population-based research demonstrates that LGBT older adults are an at-risk, health disparate, and vulnerable population. Based on these key findings, programmatic, policy, and research recommendations are as follows: 1. Ensure the development or expansion of services to address the unmet needs of LGBT older adults. The services that emerged with the highest unmet need among the LGBT older adult participants in this study include: health promotion, door-to door transportation, caregiver support, and day programs. 2. Improve the LGBT-friendliness of specific health and social services. Areas in which some LGBT older adults do not feel comfortable include: alcohol/substance abuse programs, housing assistance, and veterans’ services. In these service areas more training is likely needed to create an LGBT welcoming and friendly atmosphere and to ensure competent services. 3. Expand caregiving support programs for LGBT older adults. The participants have a high need for caregiving support, combined with low rates of caregiving arrangements or plans. Most live alone and do not have children available to help them. 4. Improve the availability of LGBT-friendly housing assistance. Housing instability is a major concern; many indicate they may be unable to stay in their home, primarily due to economic and health reasons or changing needs related to aging. Many participants, especially those with lower socioeconomic status and HIV/AIDS, need housing assistance, but many participants feel unsafe obtaining assistance as an LGBT person. 5. Address the distinct needs of lesbians, gay men, bisexuals, and transgender adults as separate sub-groups. Overall, transgender and bisexual participants report elevated need for most services, especially mental health services, meal sites and free groceries, health services, and health promotion. Racial and ethnic minority participants report higher rates of service need: mental health services for Hispanics and African Americans, housing assistance and day programs for Hispanics and Asian Americans. Gay men are less likely than lesbians to turn to friends, family members, or neighbors for social and emotional support. Lesbians are more likely than gay men to experience age-based discrimination. 6. Develop a multifaceted plan to combat discrimination by sexual orientation and gender identity and the abuse of LGBT older adults. Effective campaigns to combat discrimination and abuse have documented the need for a multifaceted approach, including the need for public awareness among the general and affected communities and businesses, as well as the evaluation and development of ways to strengthen the reporting, investigative, legislative, and judicial processes. 7. Establish a suicide prevention program that targets LGBT older adults. An alarming finding is the number of participants that contemplated suicide within the past 12 months. Many health issues emerge in this study including relatively high rates of disability, and poor physical and mental health, which in previous research has been found to be associated with increased risk of depression, which in turn can increase the risk of suicide. 41

8. Provide training and services to help LGBT older adults as well as providers anticipate future aging and health planning needs. Our findings reveal there are unmet planning needs that warrant attention, including last will and testament, and powers of attorney for health care and finances. An area that warrants attention is that LGBT participants who are not married or partnered report the lowest rates of having a will, powers of attorney for health care and finances, and revocable/irrevocable trust. 9. Promote advocacy to ensure that the needs of LGBT older adults continue to be addressed in local and state planning processes, such as the development of the Area Plan. Given the array of unaddressed needs that have been identified in this report, it is important that LGBT older adults be considered as an at-risk and vulnerable population, with their particular needs addressed in the City’s planning processes designed to address aging related needs of older adults. 10. Collect quality data on the aging and health needs of LGBT older adults. Recent federal mandates have advanced the inclusion of sexual orientation and gender identity questions in public health surveys. It is imperative that San Francisco include sexual orientation, sexual behavior, and gender identity questions in City sponsored aging and health surveys and other data collection tools. Such information is critically needed so that these communities are considered in planning and service development. A more complete understanding of the needs of San Francisco’s aging LGBT population will only be accomplished when sexual orientation and gender identity are in both aging and non-aging related City data-bases. 11. Develop a strategy to successfully reach racial and ethnic minorities, bisexuals, and transgender adults and continue to advance and enhance research with and for diverse LGBT older adults. An important goal of this study was to reach out to diverse LGBT older adults, who have not typically been included in studies of LGBT aging. Even with extensive outreach efforts and making the survey available in five different languages, reaching racial and ethnic minorities, bisexuals, and transgender older adults was difficult. These older adults may experience high levels of isolation as they age, higher than that of the general population. In future research it will be important to test the use of language related to sexual orientation and gender identity, and differing recruitment strategies for diverse populations. It is clear that the LGBT participants have important strengths and resources that can foster their aging, health, and well-being, yet they also face significant risks, which can increase their vulnerability as they age. Through a better understanding of their lived experiences, policymakers and other key stakeholders can initiate program, policy, and research initiatives to better serve the needs of older LGBT adults who live in San Francisco.

42

References Ahmed, A. T., Mohammed, S. A., & Williams, D. R. (2007). Racial discrimination & health: Pathways & evidence. The Indian Journal of Medical Research, 126(4), 318 - 327. Auldridge, A., & Espinoza, R. (2013). Health Equity and LGBT Elders of Color: Recommendations for Policy and Practice. New York: SAGE (Services and Advocacy for LGBT Elders). Barker, J. C., Herdt, G., & de Vries, B. (2006). Social support in the lives of lesbians and gay men at midlife and later. Sexuality Research & Social Policy, 3(2), 1 - 23. Buffie, W. C. (2011). Public health implications of same-sex marriage. American Journal of Public Health, 101(6), 986 - 990. doi: 10.2105/AJPH.2010.300112 Cacioppo, J. T., & Hawkley, L. C. (2003). Social isolation and health, with an emphasis on underlying mechanisms. Perspectives in Biology and Medicine, 46(3 Suppl), S39 - S52. Center for American Progress. (2012). AMA calls for marriage equality across the nation. ThinkProgress: LGBT. Retrieved October 28, 2012, from http://thinkprogress.org/lgbt/2011/06/27/254989/ama-calls-for-marriage-equality-acrossthe-nation/ Center for Disease Control and Prevention. (2012). Behavioral Risk Factor Surveillance System Retrieved April 10, 2012, from http://www.cdc.gov/brfss/ Centers for Disease Control and Prevention. (2011a). CDC Health disparities and inequalities report - United States, 2011. MMWR 2011, 60(Suppl), 1-116. Centers for Disease Control and Prevention. (2011b). Social Determinants of Health. Retrieved October 17, 2012, from http://www.cdc.gov/socialdeterminants/Definitions.html CHIS. (2003). California Health Interview Survey 2003 Adult Public Use File. AskCHIS [data analysis tool]. Retrieved June 26, 2013, from http://healthpolicy.ucla.edu/chis/data/Pages/public-use-data.aspx CHIS. (2009). California Health Interview Survey 2009 Adult Public Use File. AskCHIS [data analysis tool]. Retrieved June 26, 2013, from http://healthpolicy.ucla.edu/chis/data/Pages/public-use-data.aspx Cukrowicz, K. C., Duberstein, P. R., Vannoy, S. D., Lynch, T. R., McQuoid, D. R., & Steffens, D. C. (2009). Course of suicide ideation and predictors of change in depressed older adults. Journal of Affective Disorders, 113(1-2), 30 - 36. doi: 10.1016/j.jad.2008.05.012 de Graaf, R., Sandfort, T. G. M., & ten Have, M. (2006). Suicidality and sexual orientation: Differences between men and women in a general population-based sample from the Netherlands. Archives of Sexual Behavior, 35(3), 253 - 262. doi: 10.1007/s10508-0069020-z 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 Fredriksen-Goldsen, K. I., Cook-Daniels, L., Kim, H. J., Erosheva, E. A., Emlet, C. A., HoyEllis, C. P., . . . Muraco, A. (2013). Physical and mental health of transgender older 43

adults: An at-risk and underserved population. The Gerontologist. Advance online publication. doi: 10.1093/geront/gnt021 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 population-based study. American Journal of Public Health. Advance online publication. doi: 10.2105/AJPH.2012.301110 Fredriksen-Goldsen, K. I., & Muraco, A. (2010). Aging and sexual orientation: A 25-year review of the literature. Research on Aging, 32(3), 372 - 413. doi: 10.1177/0164027509360355 Gates, G. J. (2011). How many people are lesbian, gay, bisexual, and transgender? Los Angeles: The Williams Institute: UCLA School of Law. Gates, G. J., & Newport, F. (2012). Special Report: 3.4% of U.S. Adults Identify as LGBT. Inaugural Gallup findings based on more than 120,000 interviews. Retrieved October 23, 2012, from http://www.gallup.com/poll/158066/special-report-adults-identify-lgbt.aspx Gay and Lesbian Medical Association. (2008). Same-sex marriage and health. Washington, DC: Gay and Lesbian Medical Association. General Accounting Office, U. S., & Office of the General Counsel. (1997). GAO/OGC-97-16 Defense of Marriage Act (pp. 75). Washington, DC: United States General Accounting Office, Office of the General Counsel. Grant, J. M. (2010). Outing Age 2010: Public policy issues affecting lesbian, gay, bisexual, and transgender elders. Washington, DC: The National Gay and Lesbian Task Force Policy Institute. Grant, J. M., Mottet, L. A., Tanis, J., Harrison, J., Herman, J. L., & Keisling, M. (2011). Injustice at every turn: A report of the National Transgender Discrimination Survey. Washington, DC: National Center for Transgender Equality and National Gay and Lesbian Task Force. Halliday, T. J., & Park, M. (2009). Household size, home health care, and medical expenditures. Institute for the Study of Labor (IZA). University of Hawai'i at Manoa. Retrieved from http://www.economics.hawaii.edu/research/workingpapers/WP_09-16.pdf Herdt, G., & Kertzner, R. (2006). I do but I can't: The impact of marriage denial on the mental health and sexual citizenship of lesbians and gay men in the United States. Sexuality Research and Social Policy: Journal of NSRC, 3(1), 33-49. Herek, G. M. (2006). Legal recognition of same-sex relationships in the United States: A social science perspective. American Psychologist, 61(6), 607 - 621. doi: 10.1037/0003066X.61.6.607 Herek, G. M. (2007). Science, public policy, and legal recognition of same-sex relationships. American Psychologist, 62(7), 713 - 715. doi: 10.1037/0003-066X.62.7.713b High, K. P., Brennan-Ing, M., Clifford, D. B., Cohen, M. H. C., J., & Deeks, S. G. (2012). HIV and aging: State of knowledge and areas of clinical need for research. A report to the NIH office of AIDS research by the HIV and aging working group. Journal of Acquired Immune Deficiency Syndromes, 60(Suppl 1), S1 - S18.

44

Institute of Medicine. (2011). The health of lesbian, gay, bisexual, and transgender people: Building a foundation for better understanding. Washington, DC: The National Academies Press. Ironson, G., Solomon, G. F., Balbin, E. G., O'Cleirigh, C., George, A., Kumar, M., . . . Woods, T. E. (2002). The Ironson-woods Spirituality/Religiousness Index is associated with long survival, health behaviors, less distress, and low cortisol in people with HIV/AIDS. Annals of Behavioral Medicine, 24(1), 34 - 48. Jensen, D. (2012). Lesbian, gay, bisexual, and transgender (LGBT) seniors in San Francisco: Current estimates of population size, service needs, and service utilization. San Francisco: The San Francisco Human Services Agency Planning Unit. Kim, J., Cannon, J., Cheh, V., Duda, N., & Hall, J. (2010). Community Partnerships for Older Adults (CPFOA) Program Survey of Older Adults, 2008 [United States]. Ann Arbor, MI: Inter-university Consortium for Political and Social Research. Liptak, A. (2013, June 26). Supreme Court bolsters gay marriage with two major rulings. The New York Times. Retrieved July 8, 2013, from http://www.nytimes.com/2013/06/27/us/politics/supreme-court-gaymarriage.html?pagewanted=all&_r=0 Marmot, M. G., & Wilkinson, R. G. (2006). Social determinants of health. (2nd ed.). New York: Oxford University Press. McCullough, M. E., & Laurenceau, J. P. (2005). Religiousness and the trajectory of self-rated health across adulthood. Personality and Social Psychology Bulletin, 31(4), 560 - 573. doi: 10.1177/0146167204271657 Metlife Mature Market Institute, & American Society on Aging. (2010). Still Out, Still Aging: The Metlife study of Lesbian, Gay, Bisexual, and Transgender Baby Boomers. New York: Metlife Mature Market Institute and American Society on Aging. 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 Moriarty, D. G., Zack, M. M., Holt, J. B., Chapman, D. P., & Safran, M. A. (2009). Geographic patterns of frequent mental distress: U.S. adults, 1993-2001 and 2003-2006. American Journal of Preventive Medicine, 36(6), 497 - 505. doi: 10.1016/j.amepre.2009.01.038 National Center on Elder Abuse. (2011, 07/19/11). Elder Abuse FAQs. Retrieved April 15, 2013 from http://www.ncea.aoa.gov/faq/index.aspx National Center on Elder Abuse, & Administration on Aging. (n.d.). Statistics/Data. Retrieved May 31, 2013, from http://www.ncea.aoa.gov/Library/Data/index.aspx National Conference of State Legislatures. (2013). Civil Union and Domestic Partnership Statutes. Issues & Research. Retrieved July 8, 2013, from http://www.ncsl.org/issuesresearch/human-services/civil-unions-and-domestic-partnership-statutes.aspx National Research Center, Inc. (2008). City of San Francisco aging and adult services telephone survey. Boulder, CO: National Research Center, Inc.

45

National Research Council: Panel on Race. (2004). Understanding racial and ethnic differences in health in late life: A research agenda. Washington, D. C.: The National Academies Press. Netuveli, G., Wiggins, R. D., Hildon, Z., Montgomery, S. M., & Blane, D. (2006). Quality of life at older ages: Evidence from the English Longitudinal Study of Aging (Wave 1). Journal of Epidemiology & Community Health, 60(4), 357 - 363. doi: 10.1136/jech.2005.040071 Padilla-Frausto, D. I., & Wallace, S. P. (2013). Elder Economic Security Standard Index for California, 2011, Methodology Report. Los Angeles: UCLA Center for Health Policy Research. Pew Research Center. (2010). Baby Boomers approach 65- Glumly. Pew Social & Demographic Trends. Washington, DC: Pew Research Center. Rendall, M. S., Weden, M. M., Favreault, M. M., & Waldron, H. (2011). The protective effect of marriage for survival: A review and update. Demography, 48(2), 481 - 506. doi: 10.1007/s13524-011-0032-5 Riggle, E. D. B., Rostosky, S. S., & Horne, S. G. (2010). Psychological distress, well-being, and legal recognition in same-sex couple relationships. Journal of Family Psychology, 24(1), 82 - 86. doi: DOI 10.1037/a0017942 Ruggles, S., Alexander, J. T., Genadek, K., Goeken, R., Schroeder, M. B., & Sobek, M. (2010). Integrated Public Use Microdata Series: Version 5.0 [Machine-readable database]. Minnesota Population Center [producer and distributor], Minneapolis, MN. Solomon, R., Kirwin, P., Van Ness, P. H., O'Leary, J., & Fried, T. R. (2010). Trajectories of quality of life in older persons with advanced illness. Journal of the American Geriatric Society, 58(5), 837 - 843. doi: 10.1111/j.1532-5415.2010.02817.x Strine, T. W., Chapman, D. P., Balluz, L. S., Moriarty, D. G., & Mokdad, A. H. (2008). The associations between life satisfaction and health-related quality of life, chronic illness, and health behaviors among U.S. community-dwelling adults. J Community Health, 33(1), 40-50. doi: 10.1007/s10900-007-9066-4 Sturgeon, S. (2006). Promoting mental health as an essential aspect of health promotion. Health Promotion International, 21(Suppl 1), 36 - 41. doi: 10.1093/heapro/dal049 UCLA Center for Health Policy Research. (2012). California Health Interview Survey: CHIS Questionnaires. Retrieved January 25, 2013, from http://healthpolicy.ucla.edu/chis/design/Pages/questionnaires.aspx U.S. Census Bureau. (2011). Population 60 Years and Over in the United States: 2011 American Community Survey 1-Year Estimates, San Francisco County, California. Retrieved July 2, 2013, from http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_ 11_1YR_S0102&prodType=table U.S. Census Bureau. (2012). Current Population Survey (CPS) – Definitions. Retrieved June 5, 2013, from http://www.census.gov/cps/about/cpsdef.html# U.S. Census Bureau. (2013a). Annual Estimates of the Resident Population for Selected Age Groups by Sex for the United States, States, Counties, and Puerto Rico Commonwealth 46

and Municipos: April 1, 2010 to July 1, 2012. Retrieved June 21, 2013, from http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk U.S. Census Bureau. (2013b). State & County QuickFacts: San Francisco (city), California. Retrieved June 21, 2013, from http://quickfacts.census.gov/qfd/states/06/0667000.html U. S. Census Bureau. (2013c). American Community Survey: Questionnaire archive. Retrieved July 2, 2013, from http://www.census.gov/acs/www/methodology/questionnaire_archive/ U. S. Department of Health and Human Services. (2010). Foundation health measures: Disparities, Healthy People 2020. Retrieved October 26, 2011, from http://www.healthypeople.gov/2020/about/disparitiesAbout.aspx U.S. Department of Health and Human Services. (2011). Newsroom: Improving Data Collection for the LGBT Community. HealthCare.gov. Retrieved April 15, 2013 from http://www.healthcare.gov/news/factsheets/2011/06/lgbt06292011a.html U. S. Department of Health and Human Services. (2013). Computations for the 2013 annual update of the HHS poverty guidelines for the 48 contiguous states and the District of Columbia. Retrieved May 1, 2013, from http://aspe.hhs.gov/poverty/13poverty.cfm 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. (1990). The no-care zone: Availability, accessibility, and acceptability in community-based long-term care. Gerontologist, 30(2), 254 - 261. 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. Wienke, C., & Hill, G. J. (2009). Does the "marriage benefit" extend to partners in gay and lesbian relationships?: Evidence from a random sample of sexually active adults. Journal of Family Issues, 30(2), 259-289. Wight, R. G., LeBlanc, A. J., de Vries, B., & Detels, R. (2012). Stress and mental health among midlife and older gay-identified men. American Journal of Public Health, 102(3), 503 510. doi: 10.2105/AJPH.2011.300384 Williams, D. R., Yu, Y., Jackson, J. S., & Anderson, N. B. (1997). Racial differences in physical and mental health: Socioeconomic status, stress and discrimination. Journal of Health Psychology, 2(3), 335 - 351.

47

Methodology This is a follow-up to the original report, LGBT Older Adults in San Francisco: Health, Risks, and Resilience - Findings from Caring and Aging with Pride, which suggested that older LGBT residents of the City are at potentially increased risk for poor health. In collaboration with the San Francisco LGBT Aging Policy Task Force, we developed a targeted outreach strategy to better understand the health and aging needs of older diverse LGBT adults who reside in the City. The project announcements and surveys were distributed through the media, agency contact lists, community events, and outreach efforts. This targeted recruitment was needed to improve sample sizes for statistical comparisons and was not intended to produce a representative sample. Because of its targeted nature, it is likely not reflective of LGBT older adults living in San Francisco. Both electronic and hardcopy versions of the self-administered survey materials were distributed by the LGBT Aging Policy Task Force. The survey materials consisted of the project announcement and the survey packet, which included a questionnaire cover letter, questionnaire, and raffle/participation form. The announcement described the purpose of and criteria for inclusion in the study. The cover letter provided additional detail regarding the study, as well as informed consent. The questionnaire included questions pertaining to demographics, service and program use and needs, housing, health, resources, and risks. As an honorarium, participants also had the opportunity at the end of the survey to enter a raffle for a chance to win one of five $100 Macy's gift certificates. In order to reach a more racially and ethnically diverse sample, both electronic and hardcopy versions of survey materials were provided in English, Spanish, Russian, Chinese, and Tagalog. Surveys were distributed and collected over a 9-week period, from April to June of 2013. The LGBT Aging Policy Task Force disseminated the electronic version of the project announcement to potential participants via e-mail lists. A hyperlink was embedded in the electronically-distributed version, taking participants directly to the online survey. The announcement and survey was available in all five languages. The LGBT Aging Policy Task Force also distributed hardcopy versions of the project announcement and survey packets in all five languages at community and public settings such as bookstores and housing agencies. In addition, some outreach occurred through community centers, mental health and wellness centers, health and aging resource fairs, and agencies serving those with HIV. For example, some outreach to older transgender adults was through community centers and mental health and wellness centers. Outreach to some older LGBT Hispanics included contacts through community agencies serving those living with HIV. To be eligible, participants were required to be 60 years of age or older and living in San Francisco. In addition, participants either identified as LGBT, or were attracted to or had an intimate or sexual relationship with someone of the same sex or gender. Although both electronic and hardcopy versions of the survey were available in English, Spanish, Russian, Chinese, and Tagalog, only 12 of the 616 completed surveys that met study inclusion criteria were in Spanish, the remaining 604 were in English. For data analysis, descriptive statistics were initially conducted. Next, differences in service need and use, housing, resources and risks, and health-related indicators by sexual orientation (lesbians compared to gay men;bisexual women and men compared to lesbians and gay men) and gender identity (transgender adults compared to non-transgender adults) were examined, utilizing Student’s t-tests, Pearson’s χ2 tests, or Fisher’s exact tests, as appropriate. We 48

also conducted adjusted logistic regressions, controlling for age, income, and education in the comparisons by sexual orientation and gender identity. The summary of findings were based on the results of the adjusted analyses. In addition, we examined how service need and use, housing, resources and risks, and health-related indicators were associated with background characteristics (gender, age, race/ethnicity, income, education, relationship status, living arrangement, and HIV/AIDS status) utilizing Student’s t-tests, one-way ANOVAs, Pearson’s χ2 tests, or Fisher’s exact tests, as appropriate. In this study, lesbians, gay men, and bisexuals are treated as distinct groups, but bisexual women and men were combined due to the small sample sizes. We stratified participants into three age groups: those aged 60-69, those 70-79, and those 80 and older. For statistical comparisons for race and ethnicity, we included Hispanic, non-Hispanic white, African American, and Asian American; other racial and ethnic groups were excluded from the analyses due to insufficient sample size. Information regarding measures is detailed in Key Terms. Unavailable in most other studies, the sample of LGBT older adults in this study are age 60 and older and diverse in many respects. However, there are limitations that are important to consider. First, the design and sampling procedures used in this study do not allow for the generalizability of the findings. Thus, the findings can’t be generalized beyond those that participated in the study. Recruitment of the underrepresented groups were a primary focus of the study, and while we achieved greater diversity than most previous studies, more work is needed to find ways to effectively reach diverse communities. In addition, only self-report data were collected and likely based on participants' perceptions and interpretations rather than behaviors; such measures do not replace objective indicators of the variables under study.

49

Key Terms Sexual Orientation: Participants were asked to select from the following categories: gay/lesbian/homosexual, bisexual, heterosexual or straight, or other. Gender: Participants were asked to select their current gender from the following categories: women, men, or other. Gender Identity: Assessed by the following question: Are you or have you ever been transgender? (yes or no). Age: Calculated from participant’s year of birth. Participants were grouped into age 60-69, 70 79, or 80 and older. Race and Ethnicity: participants were asked to identify their race and ethnicity by selecting one or more of the following categories: White, Hispanic or Latino(a), Black or African American, Asian, Native Hawaiian or other Pacific Islander, American Indian, Alaskan Native, or Other. Non-Hispanic participants who marked more than one race were categorized as multiracial. Education: Determined by the highest level of education completed. Categories included: less than high school, high school or GED, less than 4 years of college, 4 years of college (bachelor’s degree), or more than 4 years of college (master’s degree or higher). Education was dichotomized into either less than 4 years of college or 4 years of college or more. Employment: Participants were asked how many hours per week they work in paid employment. Categories included: none, 1-14 hours, 15-34 hours, 35 hours or more. Participants were grouped into unemployed, employed part-time (1 to 34 hours), or employed full-time (35 hours or more). Income: Participants selected their annual household before taxes in 2012 from the following categories: $20,000 or less; $20,001 – $30,000; $30,001-$40,000; $40,001 - $50,000; $50,001$60,000; $60,001 - $70,000; $70,001 - $80,000, Greater than $80,000. Income was dichotomized by factoring annual household income with household size to determine whether participants were above 200% of the federal poverty level (FPL) (U.S. Department of Health and Human Services, 2003) or at or below 200% of the FPL. We also determined whether participants have income to meet their basic needs based on their household income, living arrangement (living alone or living with partner) and housing type (rent, own with mortgage, or own but paying mortgage) by applying the Elder Economic Security Index (Padilla-Frausto & Wallace, 2013), which is county-specific measure of the minimum income necessary to cover basic expenses. Military service: Participants were asked if they have served in the military (yes or no). Relationship Status: Participants were asked to select their current relationship status: registered domestic partnership, not married; married, legally recognized; married, not legally recognized; partnered, not married; single; divorced; widowed; separated; other. Relationship status was categorized into partnered or married, legally recognized; partnered or married, not legally recognized; and not partnered or married. 50

Sexual activity: Participants were asked with whom they had sexual relations with in the past 12 months. Categories included: male, female, both male and female, and no sexual relations. Participants were grouped into being sexually active vs. being sexually non-active. Living arrangement: Participants selected from the following categories: living alone, living with a partner/spouse, with other legal or biological family member, friend, or with others. Living arrangement was dichotomized into living alone or living with others. Children: Participants were asked if they have children or step-children that are available to help you. Categories are: no, I do not have children; yes, I have children and they are available to help me; yes, I have children but they are not available to help me. Service need: Participants were asked whether they needed the following services in the past 12 months: Alcohol/substance abuse programs, caregiver support, case manager/social worker (advocate, form-filling assistance, translation), day programs (senior/activity center, social club, adult day services), door-to-door transportation, health promotion (health education, wellness, exercise classes), health services, home-delivered meals, housing assistance (help finding or maintaining suitable housing), in-home care (home health aide, visiting nurse, personal care attendant, homemaker), meal site/free groceries (group meal site/senior lunch/free groceries), mental health services (mental health services/support groups), telephone/online referrals (telephone helpline or online information or referral for seniors), veterans’ services (VA benefits, VA hospital/clinic, CVSO for eligibility assistance), other. Service use and level of comfort using services as LGBT person: Participants were asked, in the past 12 months, to what extent they felt comfortable using the services listed above as an LGBT older adult, or as someone who is attracted to or has had an intimate or sexual relationship with someone of their same sex or gender. Categories were: very comfortable; somewhat comfortable; somewhat uncomfortable; very comfortable; and did not use service. Those who selected one of the first four categories were coded as having used service. The level of comfort using each service as LGBT persons was dichotomized into comfortable (very comfortable and somewhat comfortable) vs. uncomfortable (very uncomfortable and somewhat comfortable). For the purpose of subgroup comparisons, we also calculated overall level of comfort by averaging the comfort level of the 14 services; the range is 1(very uncomfortable) to 4 (very comfortable). Unmet service needs: In order to measure unmet service needs, we calculated what proportion of participants who indicated service needs did not use the services. Reasons for not using needed services and programs: Participants who needed but did not use services in the past 12 months were asked what was the primary reason for not using the services. Participants selected for the following list: not LGBT friendly; too expensive; difficult to access; may not quality; or other reason. Household size: Determined by asking participants how many people currently live in their household.

51

Housing arrangement: Participants were asked, regarding their current housing, whether they own, rent, or something else. Categories included: own or rent; own with mortgage paid off; own and paying mortgage payments; rent; live rent-free with relative or friend; and other (including nursing home or other health care facility). Housing type: Participants were asked to select from the following list: senior housing or agerestricted community, assisted living facility, nursing home or other health care facility, residential hotel/SRO (single room occupancy), house or apartment/condominium, currently homeless. Housing stability: Participants were asked how confident they were that they will be able to continue living in their current housing for as long as they like (Kim et al., 2010). Housing stability was dichotomized into very confident vs. somewhat confident, a little confident, and not confident at all. Primary reasons participant might move from current housing: Those who indicated not very confident regarding their housing stability were asked which of the following best describe the primary reasons they might move: health reason, economic reason, including risk of foreclosure, lack of stability in my housing situation, different needs as I age such as safety/grab bars or elevators, unsafe environment as a result of being perceived as LGBT, want to move in with my family or friends, rising crime rate in my neighborhood, friends moving elsewhere, need to move out of San Francisco, and other reason (adopted and modified from Kim et al., 2010). General health: Participants were asked, in general, how they would rate their health during the past 4 weeks (Center for Disease Control and Prevention, 2012). Response categories were dichotomized as poor (very poor, poor, fair) and good (good, very good, excellent). Physical disability: Participants were asked whether they had a condition that substantially limits one or more basic physical activities such as walking, climbing stairs, reaching, lifting, or carrying (yes or no) (US Census Bureau, 2013). Limited activities due to poor mental health: Participants were asked, “in the past 30 days how many days were you limited in any way in any activities because of emotional or mental problems?” Fourteen or more days were coded as frequent limited activities due to poor mental health (moified from Center for Disease Control and Prevention, 2012). HIV/AIDS: Participants were asked if they had ever been told they had HIV or AIDS. Suicidal ideation: Assessed by the following question, “in the past 12 months have you seriously thought about committing suicide?” (yes or no) Health services used: Participants were asked whether in the past 12 months they had used the following health services: community health centers or clinics, medication management, chronic illness management, insurance counseling, HIV services, regular doctor office visits, emergency room visits.

52

Health Insurance: Types of health care coverage that participants had were asked: Medicare; Medi-Cal; private insurance: health/medical; private insurance: long-term care; Veteran’s Administration; Indian Health Service; uninsured; other. Social support: The degree of social support received was measured by asking, “How often do you get the social and emotional support you need?” using a 5-point Likert scale (Center for Disease Control and Prevention, 2012). The range of the score is 1 (never) to 5 (always). Social support network: Participants were asked to whom they turn for support, encouragement, or short term help (such as run an errand or get a ride) and to select one or more from the following list: partner/spouse, close friend, other legal or biological family member, neighbor, faith, spiritual, or religious community, privately paid caregiver, social service provider, service agency, or organization, therapist or support group, other, or no one (adopted and modified from Metlife Mature Market Institute & American Society on Aging, 2010). Religious and spiritual activities: Participants were asked how many days, during the past 30 days, they attended faith, spiritual, or religious services or activities. Disclosure of sexual orientation or gender identity: Participants were asked to what extent, in general, they openly identified, or they were “out,” about their sexual orientation or gender identity to others. The score is 1 (not at all) to 4 (completely) (adopted and modified from Metlife Mature Market Institute & American Society on Aging, 2010). Future planning: We asked whether participants had completed any of the following for themselves: will, power of attorney for health care, revocable/irrevocable trust, power of attorney for finance/management, funeral plans, purchased long-term care insurance, provided a charitable legacy gift to an LGBT organization, and informal caregiving arrangements. Discrimination due to sexual orientation or gender identity: Participants were asked how often in the past 12 months they had been treated unfairly, with less respect than others were treated, or discriminated against due to their sexual orientation or gender identity (adopted and modified from UCLA Center for Health Policy Research, 2012). The responses were dichotomized into having experienced discrimination (rarely, sometimes, often) and not having experienced discrimination (never). Setting of discrimination due to sexual orientation or gender identity: Participants who experienced discrimination due to their sexual orientation or gender identity were asked in what setting(s) they had experienced such discrimination. Participants selected one or more from the following list: medical or health services, aging services, other social services (not aging-related), job or place of employment, faith, spiritual, or religious setting, housing, public place (such as a store, sidewalk, public transportation), interaction with police, other. Discrimination due to other reasons: Participants were also asked whether in the past 12 months they were treated unfairly, with less respect than others were treated, or 53

discriminated against for any of the following reasons: gender, race or skin color, ancestry or national origin, age, disability, speaking a language other than English, immigration status, poverty, some other reason. Abuse: Types of abuse participants had experienced in the past 12 months were measured. They include: physically hurt, pushed, punched, or assaulted in any way or physically threatened by someone; felt that someone was controlling or harassing you; verbally abused or threatened by someone; touched, grabbed, or groped without your consent or forced to do sexual acts; left without basic needs (such as food, water, medications) by someone who was supposed to take care of you; felt forced or tricked to give someone money or property. Perpetrators of abuse: For those who experienced any abuse listed above, we asked by whom the experience happened and to select one or more of the following list: an intimate partner or spouse, a friend, other legal or biological family member, a paid caregiver, a stranger, someone else. Reasons not reporting abuse: For those who experienced and did not report abuse, we asked what the primary reason was they did not report it. Participants were asked to select one of the following list: didn’t know how to report it; ashamed of the experience; didn’t trust the authorities to be fair to LGBT people; reporting it would require me to disclose my sexual orientation or gender identity; my immigration status; other reason.

54

Table 1.1. LGBT Older Adult Participants Living in San Francisco: Socio-Demographic and Background Characteristics (n = 616) % (n) Sexual orientation Lesbians Gay men Bisexual women Bisexual men Heterosexual Other Transgender Gender Men Women Other Age, mean (SD) 60-69 70-79 80 and older Race and ethnicity White (Non-Hispanic) Hispanic African American Asian American Native American Other Multiracial

22.28 (135) 71.29 (432) 1.98 (12) 2.48 (15) 0.99 (6) 0.99 (6) 4.26 (26) 73.45 (451) 25.24 (155) 1.30 (8) 67.62 (6.12) 70.94 (437) 23.38 (144) 5.68 (35) 78.61 (474) 7.46 (45) 4.98 (30) 3.81 (23) 1.66 (10) 2.16 (13) 1.33 (8)

% (n) Education High school or less Some college 4 years of college More than 4 years of college Employment Not employed Part-time Full-time Household income $20,000 or less $20,001 – $40,000 $40,001 - $60,000 $60,001 - $80,000 $80,001 or more At or below 200% poverty level Military service Relationship status Registered domestic partnership Married, legally recognized Married, not legally recognized Partnered, not married Not married or partnered Living alone Children Sexually active HIV/AIDS

8.30 (40) 21.59 (130) 24.58 (148) 45.51 (274) 63.94 (383) 18.86 (113) 17.20 (103) 26.53 (156) 21.94 (129) 15.99 (94) 10.54 (62) 25.00 (147) 29.84 (171) 21.70 (130) 8.29 (51) 8.78 (54) 1.95 (12) 18.37 (113) 62.61 (385) 57.91 (355) 14.92 (91) 58.08 (356) 25.04 (154)

Table 1.2. Socio-Demographic and Background Characteristic Comparisons by Sexual Orientation and Gender identity Lesbians

Sexual orientation,% Gay men

Bisexualsa

Transgender

Gender identity, % Non-transgender

Age, mean (SD) 66.66 (4.49)* 68.17 (6.67) 66.26 (4.32) 63.88 (2.27)* 67.82 (6.20) 60-69 75.56* 67.82 77.78 100.00 69.40 70-79 22.96 24.54 22.22 0.00 24.62 80 and older 1.48 7.64 0.00 0.00 5.98 Education High school or less 3.03* 9.69 12.00 16.00* 7.87 Some college 21.21 21.04 24.00 40.00 20.80 4 years of college 19.70 24.59 36.00 32.00 24.30 More than 4 years of college 56.06 44.68 28.00 12.00 47.03 Employment Not employed 50.38* 67.86 64.00 70.83 63.68 Part-time 22.56 17.62 24.00 12.50 19.12 Full-time 27.07 14.52 12.00 16.67 17.19 Household income $20,000 or less 20.93* 25.91 60.00* 44.00* 25.63 $20,001 – $40,000 13.18 24.94 12.00 24.00 21.86 $40,001 - $60,000 13.18 16.22 20.00 12.00 16.13 $60,001 - $80,000 14.73 9.93 0.00 16.00 10.22 $80,001 or more 37.98 23.00 8.00 4.00 26.16 At or below 200% poverty level 24.60 28.86 60.00* 54.17* 28.75 Race and ethnicity White (Non-Hispanic) 84.09 79.48 72.00 40.00* 80.28 Hispanic 3.79 8.49 12.00 8.00 7.33 African American 4.55 4.72 0.00 20.00 4.36 Asian American 3.03 3.77 8.00 12.00 3.49 Native American 0.76 1.42 0.00 12.00 1.22 Other 2.27 1.65 4.00 4.00 2.09 Multiracial 1.52 0.47 4.00 4.00 1.22 Military service 2.26* 27.62 24.00 25.00 21.58 Relationship status Registered domestic partnership 15.56* 6.50 3.70 0.00 8.73 Married, legally recognized 12.59 7.42 7.41 11.54 8.73 Married, not legally recognized 3.70 1.39 0.00 3.85 1.88 Partnered, not married 17.78 19.26 14.81 15.38 18.66 Not married or partnered 50.37 65.43 74.07 69.23 61.99 Living alone 45.93* 60.61 70.37 73.08 56.87 Children 28.03* 9.32 48.15* 23.08 14.66 Sexually active 48.51* 61.63 59.26 57.69 58.42 HIV/AIDS 0.00 33.18 12.00 28.00 24.61 Note. Student t-tests, Pearson’s χ2 tests, or Fisher’s exact tests were applied to conduct comparisons by sexual orientation and gender identity; *Indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men; bisexuals were compared with lesbians/gay men.

Table 2. Services and Programs: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics

Total Sexual orientation Lesbians Gay Men Bisexualsa Gender identity Transgender Non-transgender Gender Female Male Age 60-69 70-79 80 and older Race and Ethnicity White Hispanic African American Asian American Below 200% Poverty Level Yes No Education Some college or less 4 years of college or more Relationship status Partnered, legally recognized Partnered, not legally recognized Not married or partnered Living arrangement Living alone Living with others HIV/AIDS (men) Yes No

In-home care Need Use % % 18.34 14.45

Home-delivered meals Need Use % % 13.31 11.85

Meal site/free groceries Need Use % % 21.10 18.51

Door-to-door transportation Need Use % % 15.42 11.04

Day programs Need Use % % 19.16 13.96

15.56 17.13

11.11 14.35

5.19*‡ 14.58

4.44*‡ 13.19

11.11*‡ 21.30

8.89*‡ 18.98

14.81 15.51

8.89 11.81

14.07 19.91

11.11 14.58

33.33*

22.22

22.22

18.52

48.15*‡

40.74*

25.93

14.81

14.81

14.81

42.31*‡ 17.44

30.77* 13.85

19.23 12.99

15.38 11.62

61.54*‡ 19.49

57.69*‡ 16.92

15.38 15.38

11.54 10.94

34.62*‡ 18.46

26.92*‡ 13.33

18.06 17.96

12.26 15.08

7.74* 15.08

5.81* 13.75

15.48 22.17

13.55 19.51

15.48 15.74

9.03 11.97

12.90* 20.40

10.32 15.08

17.85 19.44 20.00

13.04 17.36 20.00

12.81* 11.11 28.57

10.98 11.11 25.71

20.59 20.14 31.43

17.62 18.06 31.43

14.65* 13.89 31.43

10.76 9.03 22.86

16.70* 22.92 34.29

12.13 18.06 20.00

17.51 22.22 26.67 8.70

13.92 15.56 20.00 8.70

11.81 17.78 23.33 13.04

10.34 15.56 20.00 13.04

17.93* 26.67 30.00 34.78

15.19* 26.67 26.67 30.43

15.61 22.22 16.67 8.70

10.76 17.78 13.33 8.70

16.24* 37.78 10.00 39.13

11.81* 28.89 6.67 30.43

30.99* 12.94

25.15* 9.95

25.73* 8.21

22.81* 7.21

39.77* 13.93

35.67* 11.69

24.56* 11.69

15.79* 9.20

23.98* 16.92

15.79 13.68

25.00* 15.64

18.89* 12.56

21.67* 9.95

20.00* 8.53

28.33* 18.01

25.00* 15.64

19.44 13.98

14.44 9.72

25.00* 16.82

17.22 12.80

4.76* 21.60 20.69

3.81* 17.60 15.92

2.86* 12.00 16.71

2.86* 10.40 14.85

7.62* 22.40 23.87

6.67* 21.60 20.16

4.76* 13.60 19.10

4.76 11.20 12.73

6.67* 20.80 21.75

5.71* 14.40 15.92

21.69* 13.95

17.46* 10.47

17.46* 7.75

15.49* 6.98

26.76* 13.57

23.10* 12.40

17.18 13.18

11.83 10.08

23.94* 12.40

17.75* 8.91

20.69 16.78

15.86 14.77

17.24 14.09

15.17 13.09

26.90 20.13

25.52* 16.78

16.55 15.44

13.10 11.41

24.14 18.46

17.24 14.09

Note. Pearson’s χ2 tests or Fisher’s exact tests were applied; *Indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity.

Table 2. Services and Programs: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics, continued Case manager/social worker

Total Sexual orientation Lesbians Gay Men Bisexualsa Gender identity Transgender Non-transgender Gender Female Male Age 60-69 70-79 80 and older Race and Ethnicity White Hispanic African American Asian American Below 200% Poverty Level Yes No Education Some college or less 4 years of college or more Relationship status Partnered, legally recognized Partnered, not legally recognized Not married or partnered Living arrangement Living alone Living with others HIV/AIDS (men) Yes No

Mental health services

Alcohol/substance abuse program

Caregiver support

Telephone/online referrals

Need

Use

Need

Use

Need

Use

Need

Use

Need

Use

%

%

%

%

%

%

%

%

%

%

22.24

19.64

27.27

23.05

6.17

5.52

10.06

7.31

21.10

17.05

14.81 22.45 40.74*

12.59 19.44 40.74*

22.22 25.00 59.26*‡

19.26 21.06 51.85*‡

3.70 6.02 11.11

3.70 5.56 7.41

8.89 8.10 29.63*‡

4.44 6.48 25.93*‡

20.74 19.44 29.63

19.26 15.51 22.22

42.31* 21.20

42.31* 18.46

80.77*‡ 24.79

65.38*‡ 21.03

15.38 5.81

11.54 5.30

42.31*‡ 8.72

34.62*‡ 6.15

34.62 20.34

26.92 16.41

18.06 23.50

16.13 20.62

27.10 26.39

23.87 22.39

4.52 6.65

4.52 5.99

12.26 8.65

7.74 7.10

21.29 20.62

19.35 16.41

23.57 16.67 28.57

21.05 14.58 22.86

29.98* 22.22 14.29

25.63 17.36 14.29

7.55 3.47 0.00

6.86 2.78 0.00

10.98 6.94 11.43

8.01 5.56 5.71

22.88 16.67 17.14

18.76 12.50 14.29

20.04 35.56 30.00 17.39

16.88* 35.56 30.00 17.39

23.42* 44.44 36.67 30.43

20.04 35.56 30.00 21.74

4.01* 13.33 16.67 8.70

3.80* 11.11 13.33 8.70

7.38* 20.00 13.33 39.13

5.06* 15.56 10.00 30.43

18.57 31.11 30.00 30.43

15.40 24.44 26.67 21.74

47.37* 11.94

41.52* 10.70

41.52* 21.89

35.09* 18.91

7.60 5.72

7.02 4.98

10.53 8.71

9.36 5.72

33.92* 17.41

26.90* 14.43

35.56* 16.59

30.00* 15.17

30.00 25.83

25.00 21.80

6.67 5.92

5.00 5.69

12.22 8.77

7.78 6.87

27.22* 19.19

21.11 15.88

4.76* 18.40 28.12

4.76* 16.00 24.67

14.29* 26.40 31.03

11.43* 22.40 26.26

4.76 8.00 5.84

2.86 7.20 5.57

4.76 10.40 10.88

1.90 7.20 8.49

12.38 22.40 22.81

7.62* 19.20 18.57

26.76* 16.28

24.51* 13.18

31.27* 21.71

26.76* 17.83

7.32 4.65

6.76 3.88

10.99 8.91

8.73 5.43

23.10 18.22

18.59 14.73

33.79* 18.46

29.66* 16.11

38.62* 20.13

32.41* 17.11

6.21 6.71

4.83 6.38

8.28 8.39

6.90 7.05

20.00 21.48

17.24 16.44

Note. Pearson’s χ2 tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity.

Table 2. Services and Programs: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics, continued

Total Sexual orientation Lesbians Gay Men Bisexualsa Gender identity Transgender Non-transgender Gender Female Male Age 60-69 70-79 80 and older Race and ethnicity White Hispanic African American Asian American Below 200% poverty level Yes No Education Some college or less 4 years of college or more Relationship status Partnered, legally recognized Partnered, not legally recognized Not married or partnered Living arrangement Living alone Living with others HIV/AIDS (men) Yes No

Health services

Health promotion

Housing assistance

Veterans services

Need

Use

Need

Use

Need

Use

Need

Use

%

%

%

%

%

%

%

49.84

45.13

27.92

19.48

24.19

19.16

50.37 47.45 77.78*‡

43.06 47.41 66.67*

24.44 27.08 55.56*‡

17.78 18.06 44.44*‡

22.96 22.69 44.44*

76.92* 48.89

61.54 44.62

53.85*‡ 27.01

46.15*‡ 18.46

54.19 48.12

49.03 43.90

26.45 28.38

52.17 45.14 40.00

47.37 41.67 31.43

49.37 60.00 46.67 47.83

Unmet need

LGBT comfort level using services

%

%

M (SD)

8.28

7.31

35.67

3.40 (.82)

17.78 18.29 33.33

0.74*‡ 9.95 11.11

0.74*‡ 9.26 3.70

31.18 36.05 32.00

3.42 (.89) 3.41 (.82) 3.35 (.70)

50.00* 22.91

42.31* 17.95

15.38 7.86

11.54 7.01

48.00 35.20

3.37 (.64) 3.40 (.83)

20.00 19.29

26.45 23.06

20.00 18.63

2.58* 10.20

1.94* 9.31

33.04 35.42

3.46 (.84) 3.40 (.81)

28.38 28.47 20.00

21.51* 17.36 2.86

27.23* 17.36 14.29

22.88* 10.42 8.57

7.32* 8.33 20.00

6.41 7.64 17.14

35.40 34.91 41.38

3.34 (.84)* 3.53 (.79) 3.71 (.56)

45.15 55.56 43.33 43.48

25.74* 42.22 33.33 47.83

17.51* 31.11 23.33 39.13

21.10* 40.00 33.33 39.13

16.03* 28.89 30.00 34.78

7.17 13.33 13.33 4.35

6.12 11.11 13.33 4.35

34.97 44.74 31.82 35.29

3.45 (.82) 3.33 (.85) 3.13 (.89) 3.32 (.59)

60.23* 46.77

54.39* 42.29

33.92 26.37

25.15* 17.91

42.11* 17.16

35.67* 12.44

11.11 6.97

10.53 5.97

41.40* 31.60

3.39 (.77) 3.45 (.81)

48.89 50.95

44.44 45.97

26.67 29.15

18.89 20.14

35.56* 19.67

28.33* 15.40

13.89* 5.92

12.22* 5.21

42.45 33.01

3.45 (.69) 3.40 (.86)

45.71 51.20 50.13

43.81 45.60 44.83

20.95 28.80 29.44

13.33 23.20 19.63

11.43* 22.40 27.85

5.71* 16.80 23.08

5.71 4.80 9.81

4.76 4.80 8.49

30.65 35.96 37.04

3.31 (.98) 3.37 (.81) 3.44 (.78)

49.58 50.39

44.51 46.51

29.86 25.19

20.00 18.60

26.20 21.32

21.41 15.89

10.14 5.81

9.01 5.04

37.68 31.84

3.44 (.75) 3.36 (.91)

53.79 45.97

49.66 41.61

31.72 27.18

22.07 18.12

29.66* 20.13

23.45 16.44

8.97 10.74

8.28 9.73

40.18 33.03

3.36 (.83) 3.43 (.81)

Note. Student’s t-test, one-way ANOVA, Pearson’s χ2 tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity.

Table 3. Housing: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics Household size M (SD)

Own, paid off mortgage %

Housing ownership Own, paying Rent mortgage % %

Other: rent free, nursing home, etc. %

House/apartment %

Types of housing Senior housing/assisted living/nursing home %

SRO/homeless

Housing stability

%

%

Total 1.48 (.65) 13.16 27.96 53.95 4.93 88.06 6.97 4.98 31.91 Sexual orientation Lesbians 1.65 (.74)*‡ 15.56 39.26* 36.30*‡ 8.89* 91.73 6.02 2.26 32.59 Gay Men 1.44 (.61) 13.21 26.65 56.13 4.01 87.65 7.36 4.99 32.78 Bisexualsa 1.37 (.69) 0.00* 7.41* 88.89*‡ 3.70 77.78 7.41 14.81* 14.81 Gender identity Transgender 1.38 (.64) 7.69 7.69* 76.92* 7.69 80.77 11.54 7.69 26.92 Non-transgender 1.49 (.65) 13.52 29.12 52.51 4.85 88.29 6.82 4.90 32.24 Gender Female 1.61 (.76)* 13.55 35.48* 42.58* 8.39* 90.85 6.54 2.61 30.32 Male 1.44 (.60) 12.87 25.73 57.56 3.84 86.82 7.27 5.91 32.73 Age 60-69 1.52 (.68)* 11.78 28.64 54.04 5.54 89.07 5.12* 5.81 29.86* 70-79 1.44 (.56) 16.31 26.95 54.61 2.13 87.05 9.35 3.60 33.80 80 and older 1.23 (.43) 17.65 23.53 50.00 8.82 79.41 20.59 0.00 50.00 Race and Ethnicity White 1.52 (.67) 14.83* 29.03 52.33 3.81* 89.27 7.08 3.65 31.63 Hispanic 1.30 (.64) 2.22 26.67 60.00 11.11 80.00 11.11 8.89 28.89 African American 1.47 (.57) 0.00 23.33 66.67 10.00 90.00 0.00 10.00 33.33 Asian American 1.35 (.57) 17.39 34.78 34.78 13.04 86.96 8.70 4.35 34.78 Below 200% Poverty Level Yes 1.46 (.75) 5.88* 4.71* 82.35* 7.06 76.47* 11.76* 11.76* 26.04 No 1.51 (.61) 15.96 38.15 42.39 3.49 94.70 3.79 1.52 32.84 Education Some college or less 1.39 (.62)* 6.18* 14.61* 69.66* 9.55* 82.58* 8.43 8.99* 29.78 4 years of college or more 1.53 (.66) 16.15 33.25 47.51 3.09 90.87 6.01 3.13 32.54 Relationship status Partnered, legally recognized 2.08 (.50)* 20.95* 52.38* 21.90* 4.76 97.00* 3.00 0.00* 43.81* Partnered, not legally recognized 1.82 (.63) 16.00 34.40 41.60 8.00 88.80 9.60 1.60 39.20 Not married or partnered 1.20 (.51) 10.03 18.70 67.48 3.79 85.37 7.05 7.59 26.29 Living arrangement Living alone 1.03 (.19)* 10.09* 20.17* 67.15* 2.59* 84.15* 8.93* 6.92* 26.22* Living with others 2.14 (.50) 17.44 38.37 36.05 8.14 93.28 4.35 2.37 39.92 HIV/AIDS (men) Yes 1.40 (.58) 10.34 26.21 61.38 2.07 92.36* 4.86 2.78 29.17 No 1.52 (.67) 14.24 25.76 55.25 4.75 84.30 8.53 7.17 34.80 Note. Student’s t-test, one-way ANOVA, Pearson’s χ2 tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity; adjusted analyses were not applied to “Other: rent free, nursing home, etc” and “SRO/homeless” due to small sample size.

Table 3. Housing: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics, continued Top 6 Reasons to move among those who are not competent about their housing stability

Total Sexual orientation Lesbians Gay Men Bisexualsa Gender identity Transgender Non-transgender Gender Female Male Age 60-69 70-79 80 and older Race and Ethnicity White Hispanic African American Asian American Below 200% Poverty Level Yes No Education Some college or less 4 years of college or more Relationship status Partnered, legally recognized Partnered, not legally recognized Not married or partnered Living arrangement Living alone Living with others HIV/AIDS (men) Yes No

Economic reasons

Health reasons

Different needs with aging

Lack of stability

Need to move out of SF

Rising crime rate

%

%

%

%

%

%

54.41

44.36

39.95

20.59

12.01

11.03

50.00 55.71 52.17

47.78 43.93 30.43

46.67 38.93 26.09

14.44 20.00 47.83*‡

14.44 11.43 13.04

4.44* 12.86 4.35

78.95* 53.25

52.63 44.16

36.84 40.26

42.11* 19.48

15.79 11.95

21.05 10.39

54.21 54.95

44.86 44.03

40.19 39.25

19.63 21.16

14.02 11.26

4.67* 13.31

56.67 45.05 64.71

39.33* 57.14 64.71

38.33 45.05 41.18

20.67 23.08 5.88

14.67* 3.30 11.76

12.00 8.79 5.88

56.88 46.88 36.84 42.86

45.00 37.50 26.32 42.86

41.56 37.50 21.05 35.71

20.31 12.50 15.79 28.57

12.19 9.38 10.53 14.29

9.38* 25.00 10.53 0.00

52.00 56.44

42.40 45.83

31.20* 44.70

32.00* 15.53

8.80 13.64

19.20* 6.44

53.28 54.96

32.79* 48.94

22.95* 46.81

23.77 19.15

9.84 13.12

20.49* 7.09

51.72 57.89 53.18

43.10 46.05 44.57

50.00 38.16 37.83

12.07 19.74 22.47

18.97 14.47 9.74

8.62 11.84 11.24

53.17 56.86

47.22 40.52

39.29 41.18

22.22 18.30

9.13* 16.34

13.49 7.19

58.00 53.40

39.00 46.60

36.00 40.84

18.00 23.04

10.00 12.04

15.00 12.57

Note. Pearson’s χ2 tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity.

Table 4. Resources: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics Social Support M (SD)

No one %

Partner or spouse %

Close friend %

Family %

Social support network Faith Neighbor community % %

Paid caregiver %

Social services %

Therapist %

Religious activities %

Total 3.68 (.96) 8.60 35.87 71.57 17.02 22.15 12.40 3.64 12.40 23.31 30.83 Sexual orientation Lesbians 3.83 (.72) ‡ 3.76*‡ 48.87*‡ 80.45* 25.56*‡ 30.08*‡ 15.79 0.75 8.27 26.32 33.33 Gay Men 3.65 (1.01) 9.88 33.41 69.65 15.53 18.82 10.59 4.47 12.00 20.24 28.57 Bisexualsa 3.64 (.91) 16.00 16.00* 72.00 12.00 32.00 20.00 4.00 32.00* 40.00* 36.00 Gender identity Transgender 3.52 (1.08) 8.00 24.00 60.00 4.00 28.00 28.00*‡ 4.00 28.00* 48.00* 58.33* Non-transgender 3.69 (.96) 8.35 36.70 72.17 17.74 21.91 11.65 3.65 11.65 22.26 29.34 Gender Female 3.79 (.79) 4.61* 43.42* 78.29* 23.03* 29.61* 17.11* 1.32 9.87 27.63 33.33 Male 3.65 (1.01) 9.93 33.41 69.30 15.35 18.96 10.38 4.29 12.87 21.22 29.23 Age 60-69 3.62 (.96) 8.84 36.28 71.16 16.28 21.16 13.26 3.49 13.49 25.35* 33.09 70-79 3.80 (.94) 8.51 36.88 70.92 19.86 24.11 11.35 2.84 9.22 21.99 27.54 80 and older 3.88 (1.04) 5.88 26.47 79.41 14.71 26.47 5.88 8.82 11.76 2.94 15.63 Race and Ethnicity White 3.72 (.93) 8.86 39.03* 73.00 17.51 23.63 10.76 3.59 10.55 21.31 28.10* Hispanic 3.55 (1.15) 11.36 18.18 68.18 15.91 13.64 11.36 2.27 20.45 27.27 32.56 African American 3.57 (1.04) 6.67 26.67 56.67 6.67 6.67 20.00 6.67 20.00 20.00 34.48 Asian American 3.39 (1.16) 4.35 34.78 69.57 13.04 13.04 26.09 8.70 8.70 34.78 60.87 Below 200% Poverty Level Yes 3.48 (1.05)* 11.11 17.54* 64.33* 12.28* 23.39 14.62 3.51 24.56* 28.07 33.93 No 3.77 (.89) 7.46 44.03 75.87 19.40 22.14 11.19 3.23 7.46 21.39 29.59 Education Some college or less 3.50 (1.14)* 13.89* 26.67* 61.11* 5.00* 18.33 10.56 2.78 19.44* 17.78* 27.33 4 years of college or more 3.75 (.87) 6.40 39.57 76.30 22.27 23.93 13.27 4.03 9.48 25.83 32.28 Relationship status Partnered, legally recognized 4.05 (.88)* 1.92* 95.19* 72.12 21.15 23.08 7.69 0.96 2.88* 19.23 23.76 Partnered, not legally recognized 4.01 (.76) 2.40 88.80 65.60 20.80 24.00 16.80 1.60 12.00 23.20 38.02 Not married or partnered 3.47 (.98) 12.53 1.09 74.11 14.71 21.25 11.72 4.90 14.99 24.52 30.34 Living arrangement Living alone 3.48 (1.00)* 12.43* 10.12* 73.70 12.72* 21.97 12.72 4.62 15.90* 23.70 30.18 Living with others 3.95 (.83) 3.52 71.09 68.75 22.66 22.27 11.33 2.34 7.03 22.66 31.30 HIV/AIDS (men) Yes 3.60 (1.00) 10.42 27.78 65.28 15.97 19.44 9.72 0.69* 19.44* 29.86* 32.62 No 3.70 (.95) 9.73 36.24 71.14 14.77 18.46 10.74 6.04 9.73 16.78 27.34 Note. Student’s t-test, one-way ANOVA, Pearson’s χ2 tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity; adjusted analyses were not applied to “paid caregiver” due to small sample size.

Table 4. Resources: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics, continued Identity disclosure

Total Sexual orientation Lesbians Gay Men Bisexualsa Gender identity Transgender Non-transgender Gender Female Male Age 60-69 70-79 80 and older Race and Ethnicity White Hispanic African American Asian American Below 200% Poverty Level Yes No Education Some college or less 4 years of college or more Relationship status Partnered, legally recognized Partnered, not legally recognized Not married or partnered Living arrangement Living alone Living with others HIV/AIDS (men) Yes No

M (SD)

%

3.52 (.68)

52.33

61.13

28.74

29.57

23.26

15.61

11.13

8.47

3.51 (.69) 3.57 (.64) 3.07 (.83)*‡

57.14 53.79 12.00*‡

72.18*‡ 60.43 36.00*

37.59*‡ 27.96 4.00*

37.59‡ 28.67 16.00

20.30 25.12 12.00

21.80* 14.22 4.00

12.03 11.61 4.00

9.02 8.29 4.00

3.23 (.91)*‡ 3.53 (.66)

32.00* 53.40

48.00 62.13

12.00 29.67

16.00 30.37

16.00 23.73

12.00 15.88

4.00 11.52

12.00 8.38

3.50 (.71) 3.54 (.66)

53.29 52.05

68.42* 59.09

33.55 27.05

34.87 27.73

18.42 24.77

20.39 13.86

11.18 11.14

8.55 7.95

3.55 (.69)* 3.49 (.63) 3.26 (.70)

48.24* 60.28 70.59

58.78* 63.83 79.41

24.59* 39.72 35.29

26.70 36.17 38.24

18.97* 30.50 47.06

14.29 21.28 8.82

9.37 14.18 20.59

8.20 8.51 11.76

3.56 (.64)* 3.30 (.80) 3.30 (.92) 3.39 (.78)

57.63* 32.56 23.33 43.48

66.53* 51.16 30.00 39.13

32.20* 16.28 10.00 21.74

33.26* 20.93 13.33 17.39

24.36 30.23 6.67 21.74

15.47 13.95 20.00 17.39

13.14 4.65 3.33 4.35

8.69 6.98 10.00 4.35

3.49 (.70) 3.55 (.66)

27.81* 60.85

41.42* 68.08

9.47* 36.91

14.20* 35.16

18.93 23.94

1.78* 21.95

1.78* 15.46

8.28 8.73

3.44 (.73) 3.55 (.65)

31.46* 60.71

50.56* 65.24

12.92* 35.48

15.73* 35.24

19.10 24.76

6.74* 19.29

4.49* 14.05

7.87 8.57

3.70 (.56)* 3.57 (.61) 3.47 (.70)

72.12* 56.00 45.60

83.65* 68.00 51.92

54.81* 32.00 19.78

46.15* 35.20 22.25

20.19 19.20 26.10

24.04* 16.80 12.91

14.42 12.80 9.34

6.73 9.60 8.24

3.45 (.70)* 3.63 (.63)

48.69* 57.81

53.94* 71.09

21.57* 38.67

23.91* 37.50

25.95 19.92

13.70 18.36

9.62 13.28

7.58 9.77

3.61 (.65) 3.49 (.69)

51.41 52.19

66.20* 55.56

24.65 28.28

26.06 28.62

25.35 24.58

9.15* 16.16

12.68 10.44

7.75 7.74

Will

Revocable/ irrevocable trust %

Future Planning Power of attorney Funeral plans for finance % %

Power of attorney for health care %

Long-term care insurance %

Charitable legacy gift %

Informal caregiving arrangement %

Note. Student’s t-test, one-way ANOVA, Pearson’s χ2 tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity.

Table 5. Risks: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics

Total Sexual orientation Lesbians Gay Men Bisexualsa Gender identity Transgender Non-transgender Gender Female Male Age 60-69 70-79 80 and older Race and Ethnicity White Hispanic African American Asian American Below 200% Poverty Level Yes No Education Some college or less 4 years of college or more Relationship status Partnered, legally recognized Partnered, not legally recognized Not married or partnered Living arrangement Living alone Living with others HIV/AIDS (men) Yes No

Discrimination Experienced - Past 12 Months Race Age % %

Abuse

Victimization

10.34

% 5.10

% 20.58

18.18 13.69 14.81

10.61 8.07 22.22*

6.25 4.34 8.33

12.50* 20.96 41.67*

46.15 33.39

38.46*‡ 14.64

26.92* 9.64

12.00 4.65

32.00 20.04

11.84 10.28

40.79* 31.31

19.08 14.02

13.16 9.11

6.16 4.62

14.38* 22.40

14.89 9.56 3.23

13.95* 2.94 0.00

36.64 29.41 19.35

17.73 10.29 12.90

13.00* 3.68 3.23

5.48 5.15 0.00

24.52* 11.76 6.25

40.38* 53.33 70.00 52.17

10.00* 18.18 33.33 9.52

3.26* 25.00 60.00 47.62

31.30 31.82 50.00 38.10

15.00 13.64 23.33 4.76

7.83* 9.09 23.33 9.52

3.71* 11.36 13.33 13.04

17.90 29.55 26.67 26.09

52.35* 40.30

15.79 11.92

14.04* 8.55

42.11* 30.05

26.32* 11.92

21.64* 5.44

7.14 3.58

32.74* 15.35

46.11 43.71

14.37 12.20

14.37* 8.54

32.76 34.15

18.39 14.63

13.22 8.78

8.52* 3.67

25.00 18.58

40.00 44.00 45.14

11.88 13.22 12.81

5.94 10.74 11.98

29.70 38.02 33.98

8.91 16.53 17.55

4.95 8.26 12.26

0.99 6.56 5.60

7.92* 22.13 23.53

44.13 43.97

13.91 11.65

11.83 8.84

34.32 33.73

16.86 14.46

12.43* 7.23

5.65 4.42

22.62 18.07

44.83 40.40

5.00 5.94

12.14 9.09

32.86 30.42

22.86* 9.79

13.57* 6.64

6.99 3.46

27.97 19.72

Orientation/Identity %

Gender %

Disability %

Poverty %

44.33

13.05

10.68

34.07

15.76

47.41 40.94 66.67*

27.27*‡ 5.13 29.63*‡

9.09 9.29 25.93*

40.91*‡ 30.81 37.04

80.77*‡ 42.21

69.23*‡ 10.54

38.46*‡ 9.46

50.32 41.89

31.58* 5.61

50.81* 28.67 27.27

Note. Pearson’s χ2 tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity.

Table 6. Health: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics

Total Sexual orientation Lesbians Gay Men Bisexualsa Gender identity Transgender Non-transgender Gender Female Male Age 60-69 70-79 80 and older Race and Ethnicity White Hispanic African American Asian American Below 200% Poverty Level Yes No Education Some college or less 4 years of college or more Relationship status Partnered, legally recognized Partnered, not legally recognized Not married or partnered Living arrangement Living alone Living with others HIV/AIDS (men) Yes No

Poor General health

Physical disability

Poor mental health

HIV/AIDS

Suicidal ideation

%

%

%

%

%

28.60

42.40

8.30

25.04

14.55

20.45* 30.28 24.00

37.40 42.18 50.00

5.65 8.27 16.67

0.00* 33.18 12.00

12.88 13.84 16.00

52.00*‡ 27.65

76.00*‡ 41.12

26.09* 7.44

28.00 24.61

32.00* 13.71

22.52 29.95

39.60 42.27

6.99 8.64

1.33* 32.73

15.89 13.50

28.14 29.79 29.41

40.94 45.71 47.06

9.49 5.19 6.25

28.84* 18.71 2.94

15.76 11.51 11.76

28.27 35.56 30.00 21.74

42.13 40.91 43.33 27.27

7.93 7.14 10.34 4.55

22.88* 51.11 33.33 8.70

14.50 4.55 25.00 8.70

43.86* 22.14

61.18* 33.92

12.12* 6.20

27.49 24.50

20.96* 12.47

37.78* 24.41

56.50* 36.52

12.21* 6.42

32.22* 21.43

16.38 13.88

17.31* 17.60 35.69

29.13* 39.52 47.53

5.10 6.72 9.60

12.50* 27.42 27.87

4.95* 14.40 17.36

34.97* 19.92

48.54* 34.51

10.57* 5.31

28.32* 20.47

17.54* 10.67

35.17 27.52

52.11* 37.71

9.63 8.22

-----

14.79 12.93

Note. Pearson’s χ2 tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity.

Table 6. Health: Comparisons by Sexual Orientation, Gender Identity, and Background Characteristics, continued Health services used Regular doctor office visits

Emergency room visits

Community health centers

Chronic illness management

Medication management

Insurance counseling

%

%

%

%

%

%

HIV services (those with HIV) %

Total 82.63 22.08 20.62 17.05 13.96 3.90 59.60 Sexual orientation Lesbians 82.96 17.78 11.11*‡ 12.59 10.37 0.74* --Gay Men 83.56 22.92 19.68 17.13 13.66 4.86 58.87 Bisexualsa 70.37 22.22 59.26*‡ 22.22 14.81 3.70 33.33 Gender identity Transgender 88.46 34.62 53.85*‡ 42.31*‡ 46.15*‡ 11.54 100.00* Non-transgender 82.56 21.71 19.32 15.73 12.48 3.59 58.16 Gender Female 82.58 20.00 18.06 14.19 12.90 1.29* 100.00 Male 82.71 22.39 20.84 17.52 13.75 4.88 58.62 Age 60-69 81.92 18.99* 24.49* 18.54 13.50 5.03* 63.71* 70-79 83.33 28.47 11.81 14.58 14.58 0.69 38.46 80 and older 88.57 34.29 8.57 8.57 17.14 2.86 100.00 Race and Ethnicity White 84.81 23.00 17.30* 16.88 12.87 3.80 55.56 Hispanic 82.22 22.22 35.56 17.78 13.33 6.67 78.26 African American 76.67 13.33 33.33 6.67 20.00 3.33 50.00 Asian American 86.96 21.74 26.09 17.39 21.74 4.35 100.00 Below 200% Poverty Level Yes 84.21 28.65* 40.94* 28.65* 24.56* 3.51 87.23* No 84.58 19.40 13.18 13.18 9.95 3.98 44.90 Education Some college or less 83.89 23.33 27.22* 21.67 19.44* 4.44 68.97 4 years of college or more 84.12 21.80 18.25 15.40 12.09 3.79 54.44 Relationship status Partnered, legally recognized 86.67 17.14 9.52* 9.52 7.62 3.81 46.15 Partnered, not legally recognized 80.80 28.80 20.00 18.40 13.60 4.00 58.82 Not married or partnered 82.49 21.22 23.87 19.10 16.18 3.98 60.78 Living arrangement Living alone 84.23 23.94 23.10 19.15 15.49 4.23 61.22 Living with others 80.62 19.77 17.05 14.34 12.02 3.10 55.77 HIV/AIDS (men) Yes 86.90 28.28 29.66* 30.34* 20.00* 8.28* --No 82.55 20.13 17.11 11.41 11.07 3.36 --2 Note. Pearson’s χ tests or Fisher’s exact tests were applied; * indicates statistically significant relationship between variables at p < .05. a Due to small sample size, we were unable to distinguish between bisexual women and men in comparisons with lesbians and gay men. ‡ Indicates that a p value remains <.05 in adjusted logistic regressions after controlling for age, income, and education; adjusted tests were applied to comparisons of sexual orientation and gender identity; adjusted analyses were not applied to “insurance counseling” due to small sample size.

Figure 1

Figure 2