Sage Journals: Discover world-class research

Abstract

Socioeconomic status (SES) is a fundamental cause of ill health, but an understudied determinant of health for gay, bisexual, and other men who have sex with men (MSM). Surprisingly, few studies have examined the relations between poverty and depression among young MSM. The aims of this study were to determine the reliability and validity of an 18-item Family Resource Scale (FRS) as a measure of SES among YMSM and examine the relations between SES and depression, while taking into account factors associated with resilience or risk for poor mental health. Reliability of the SES scale was determined with Cronbach’s alpha. Validity was assessed with factor analysis and bivariable comparisons with other SES measures. Multiple logistic regression was used to determine the relations between depressive symptomology (via the Beck Depression Inventory–II), SES, and gay-related psychosocial variables. In this racially/ethnically diverse sample (mean age = 21.8 years, 37.3% Hispanic/Latino, 30.5% White, 14.9% Black, and 17.0% other race/ethnicity), 70.8% reported incomes ≤ $10,000 and the mean FRS score was 4.1 (SD = 0.9, range 0-5). The FRS demonstrated reliability (α = .91) and criterion and construct validity. The Beck Depression Inventory–II rated 17.6% with depressive symptomology. Higher FRS scores were associated with a lower odds of depression (adjusted odds ratio = 0.58; 95% confidence interval = 0.46-0.74) in logistic regression models controlling for gay community affinity and internalized homophobia. This diverse sample of YMSM in New York City reported substantial financial hardship and those who were more gay-identified had fewer material resources. Fewer material resources and internalized homophobia were both associated with higher odds of depression.

Keywords

behavioral issues drug use physiological and endocrine disorders HIV/AIDS research quantitative research

Introduction

In 2014, 46.7 million (14.8%) U.S. residents were living in poverty (DeNavas-Walt & Proctor, 2015). Although census data suggest that male same-sex households have higher median incomes than married opposite sex households (U.S. Census Bureau, 2013b) and are less likely to be living in poverty (Badgett, Durso, & Schneebaum, 2013), studies have demonstrated that gay, bisexual, and other men who have sex with men (MSM) on average make 10% to 30% less than heterosexual men (Carpenter, 2007; Klawitter, 2015; Sabia, 2015). Analyses of the National Survey of Family Growth and the 2012 Gallup Daily Tracking Poll revealed that a higher proportion of gay men were living in poverty relative to their heterosexual counterparts (Badgett et al., 2013). Despite these data, the “myth of gay affluence” (McGarrity, 2014) perpetuates the belief that lesbians, gays, and bisexuals (LGBs) tend to be more educated and wealthier than the general population, in part, because of assumptions that they have dual income households with no children to support (Albelda, Badgett, Schneebaum, & Gates, 2009; McGarrity, 2014). This myth may contribute to the dearth of research concerned with the relations between SES and health among MSM.

Socioeconomic status (SES) is a fundamental cause of ill health (Link & Phelan, 1995), but an understudied determinant of the health of gay, bisexual, and other MSM. Lower SES has been associated with newly diagnosed HIV infections (Mayer et al., 2014), HIV testing (Lauby & Milnamow, 2009), and self-rated health among MSM (Thomeer, 2013). SES affects a variety of health outcomes including mental health, and the relationship between depression and poverty is quite robust (Reiss, 2013). For example, in the National Health and Nutrition Examination Study (NHANES), individuals living below the poverty line were two and a half times more likely to report depression than those who are at or above the poverty line (Pratt & Brody, 2014).

Research suggests that sexual minorities are more likely to experience depression as compared with heterosexuals (Institute of Medicine, 2011). For example, in Wave IV of the National Longitudinal Study of Adolescent Health (Add Health), between 16.7% and 24.7% of men who had sex with men only (MSMO) and men who had sex with men and women (MSWM), respectively, reported depression as compared with only 8.7% of men who had sex with women (MSW; Dyer, Regan, Pacek, Acheampong, & Khan, 2015). There are surprisingly few studies that have examined the relations between poverty and depression among MSM. Moreover, results are mixed. For example, Wohl et al. (2013) examined the relationship between depression and income among Latino and African American MSM and did not identify a significant association. Meanwhile, De Santis, Colin, Provencio Vasquez, and McCain (2008) reported that lower income was associated with more depressive symptoms among Hispanic MSM in South Florida and Mao et al. (2009) reported that lower income was associated with increased odds major depressive disorder among MSM in Australia.

Measurement of SES is fraught with challenges, with inaccurate income reporting a common issue (Besharov & Couch, 2009; Dorling, 1999; Rosenfeld, 2010; Sen, 1976; Turrell, 2000). Income is commonly used as an indicator of SES; however, for emerging adults aged 18 to 24 years(Arnett, 2000) in particular, income may not be a reliable measure of SES as their incomes may be supplemented by parents or significant others. Income may also not be a good measure of SES among the poor because it does not accurately capture variation (Sen, 1976). In a previous study, a material resource scale among drug users in New York City (NYC; Ompad et al., 2012) was a reliable, valid, and a useful measure of SES in low income populations. The overriding goal of this analysis is to disentangle social and psychosocial determinants of health, which are drivers of syndemics—the multiple and overlapping health burdens that negatively affect the well-being on sexual minority populations (Halkitis, 2010; Halkitis, Wolitski, & Millett, 2013). First, the reliability and validity of a material resources scale as a measure of SES among young men who have sex with men (YMSM) is assessed. Next, the relations between material resources and depression is examined in an effort to understand the associations between SES and mental health among YMSM while taking into account gay-related psychosocial variables that may be associated with resilience or risk for poor mental health.

Method

Sample

This analysis uses data from an ongoing prospective study of YMSM in NYC locally known as Project 18 (P18). The study protocols and methodology have been described in detail previously (Halkitis, Kapadia, et al., 2013; Halkitis, Moeller, et al., 2013). Recruitment was conducted via venue-based outreach as well as through flyers and Internet advertisements between 2009 and 2011. In order to be eligible for the study, men had to be aged 18 to 19 years, biologically male, a resident of the NYC metropolitan area, and self-report sex with another man in the past 6 months. Participants also had to be HIV seronegative to qualify for the study, thus, all participants were tested for HIV. Follow-up interviews were conducted every 6 months; 486 men (81%) returned for the 36-month follow-up interview. The study was reviewed and approved by New York University’s Institutional Review Board.

Data Collection

The measures used in this analysis were collected at the 36-month follow-up interviews unless otherwise indicated. The 36-month surveys were conducted between July 2012 and June 2014. The data were collected through audio-computer assisted self-interview and the survey included items on sociodemographics, material resources, substance use, and depression. A total of 482 YMSM were included in this analysis, after exclusion of four respondents who were missing data on SES or depression.

With regard to sociodemographic characteristics, participants self-reported their race/ethnicity (i.e., Hispanic/Latino, Black, White, Asian, Pacific Islander, and multiracial) and sexual orientation. Sexual orientation was assessed with a one-item measure (Kinsey, Pomeroy, & Martin, 1948), which was dichotomized as exclusively homosexual versus not exclusively homosexual.

Material resources were measured with an 18-item modified version (Ompad et al., 2012; see the appendix) of the 30-item Family Resource Scale (FRS; Dunst & Leet, 1987; Van Horn, Bellis, & Snyder, 2001), which included items on physical and human resources and time allocation. The modified FRS excluded 12 items on child- and time-related resources (Ompad et al., 2012). The 18-item FRS was added to the P18 survey at the Month 36 visit. Men were asked whether they (and their family) currently had adequate material resources to meet the specific needs of the family as a whole as well as the needs of individual family members. Items addressed resources such as having food for two meals a day, a house or apartment, money to pay monthly bills, time to get enough sleep/rest, and money to save (all items listed in Table 1) and were roughly ordered from most to least basic needs. Each item contained six answer options: “never,” “rarely,” “less than half the time,” “about half the time,” ‘more than half the time,” and “always.” A sum score was computed with a potential range from 0 (never enough resources for all items) to 5 (always enough resources for all items) with a higher score indicative of more resources. The 18-item FRS was previously validated with a drug-using population (Ompad et al., 2012) and had high internal consistency (α = .91).

Table 1.

Results From an Exploratory Factor Analysis With Oblique Rotation for the Revised 18-Item Family Resource Scale Among 482 YMSM in New York City.

	Factor 1: Basic needs	Factor 2: Health and social services	Factor 3: Economic resources
Food for 2 meals per day	0.81	−0.05	0.03
House or apartment	0.99	−0.01	−0.16
Money to buy necessities	0.69	−0.08	0.34
Enough clothes for you/family	0.72	0.02	0.18
Heat for house/apartment	0.94	0.05	−0.07
Indoor plumbing/water	0.96	0.06	−0.07
Money to pay monthly bills	0.63	0.00	0.31
Good job for self/spouse	0.15	0.10	0.47
Medical care for self/family	0.02	1.06	−0.02
Public assistance	−0.02	0.55	0.01
Dependable transport	0.56	0.17	0.17
Time to get enough sleep/rest	0.59	−0.06	0.19
Furniture for home/apartment	0.76	0.05	0.12
Phone/access to phone	0.76	0.14	0.06
Dental care for self/family	0.09	0.59	0.25
Money buy things for self	0.08	0.04	0.88
Money for entertainment	0.09	0.02	0.90
Money for savings	−0.08	0.09	0.85
Eigenvalue	11.20	1.37	1.21
Cronbach’s α^a	0.93	0.65	0.84

Note. Model fit statistics: χ²(102) = 268.77, p < .001, root mean square error of approximation (RMSEA) = 0.058 (95% confidence interval = 0.050-0.067); comparative fit index (CFI) = 0.996, Tucker–Lewis index (TLI) = 0.995, standardized root mean square residual (SRMR) = 0.034.

Cronbach’s α for all 18 items combined = .91. Bold face type identifies the items that load on each factor.

Educational attainment was categorized as high school or less, associate’s degree or some college, and bachelor’s degree or higher. Perceived familial SES was measured at baseline, which was categorized as lower (lower and lower middle), middle, and upper (upper middle and upper) class. Income was categorized as none, <$5,000, <10,000, <$20,000, and >$20,000. Current living situation was defined as living alone, with parents/ family, with a friend/roommate, with a male lover/partner, or in shelters and/or homeless. Ten income sources in the past 6 months and which gave the most income were assessed (Ompad et al., 2012). The source that gave the most income was collapsed into five categories: employment (i.e., regular full- or part-time employment, owning a business, and temporary work), public assistance (i.e., Public Assistance, Welfare, Supplemental Security Income, food stamps, State Public Aid, disability, unemployment), informal economy (i.e., recycling cans, selling items on the street, returning bottles for deposits, windshield wiping, or panhandling for money), income from someone else (i.e., relative, significant other, or friend), and illegal activities (i.e., theft, robbing, stealing, conning, selling drugs, and sex for money). Hunger was assessed with one item assessing whether someone had gone hungry in the past 12 months.

Because previous research has suggested that MSM have lower incomes (Carpenter, 2007; Klawitter, 2015; Sabia, 2015), the relationship between gay-related psychosocial measures and material resources was examined. Gay community affinity was measured with one item that stated “I feel part of the gay community in New York City.” The variable was categorized as low (i.e., strongly disagree and disagree), medium (i.e., neither agree nor disagree), or high (i.e., strongly agree and agree). Gay-related public and personalized stigma were measured with two subscales adapted from Berger, Ferrans, and Lashley’s (2001) HIV Stigma Scale by changing the reference from “people with HIV” to “people who are gay” and reducing the number of items. Public stigma was measured with two items: “Most people think a person who is gay is disgusting.” and “Most people who are gay are rejected when others find out.” Cronbach’s α = .81 in this sample. The response options were strongly disagree (1), disagree (2), agree (3), and strongly agree (4). A sum score was calculated and a dichotomous variable was created for low (i.e., less than 5) and high (i.e., 5 to 8) gay-related public stigma. Personalized stigma was a three-item scale with the same response options as the Public Stigma Scale and the following items: “I have been hurt by how people reacted to learning I’m gay,” “I have stopped socializing with some people because of their reactions of my being gay,” and “I have lost friends by telling them that I am gay.” Cronbach’s α = .85 in this sample. A sum score was calculated and dichotomized into low (i.e., less than 6) and high (i.e., 6-12). Internalized homophobia was measured with a four-item scale from Thiede et al. (2003); Cronbach’s α = .79 in Thiede et al. and .86 in this sample. A sum score was calculated and dichotomized into low (i.e., less than 12) and high (i.e., 12-20).

The health outcome of interest was self-reported depressive symptomology, as measured by the Beck Depression Inventory–Version II (BDI-II; Beck, Steer, Ball, & Ranieri, 1996). The BDI-II contains 21 items with ordinal response options (range: 0 to 3) with higher scores indicating a higher level of depression; Cronbach’s α = .91 in both Beck et al. (1996) and this sample. The mean sum score was computed for all 21 items; thus, the possible range for the sum scores was 0 to 5. The BDI-II was categorized into a dichotomous variable of no depression (BDI-II ≤ 13) versus mild depression or more (BDI-II ≥14) using established cut-points.

Analysis

Validation of the FRS

To establish validity of the modified FRS, which has not been used in an urban sample of YMSM, an exploratory factor analysis (EFA) was conducted. This was done to determine the true factor structure of all items within this sample and to examine whether items loaded into the distinct constructs from the original measure. EFA was conducted using Mplus software, Version 6 (Muthén & Muthén, 2007). All items were entered simultaneously to examine factor loadings and fit. Ideally, items with good fit load on only a single factor, have a loading of ≥0.4, and do not cross-load onto other factors; each loading onto a factor is ideally 0.4 or higher than a loading on another factor (Stevens, 1992). EFA is used to ensure that items load into distinct concepts and results can also be used to suggest elimination of items with poor fit. Items that loaded together were considered factors if they were associated with eigenvalues greater than 1 to ensure that all values represent substantial variation (Kaiser, 1960). Oblimin rotation was used to rotate clusters of items onto different factors without orthogonal constraints as factors were theoretically correlated. Since items were ordinal, they were treated as categorical data in the model.

In addition to examining factor loadings and eigenvalues, additional measures to ensure that factor structures had good fit were assessed including the root mean square error of approximation (RMSEA), standardized root mean square residual (SRMR), comparative fit index (CFI; Bentler, 1990) and the Tucker–Lewis index (TLI; Tucker & Lewis, 1973). Cronbach’s alphas (Cronbach, 1951) were estimated to determine internal consistency (reliability) of each factor. An α ≥ .90 is considered excellent reliability; α = .70 to .89 represents good reliability, and α = .60 to .69 represents acceptable reliability.

Criterion and construct validity for the full FRS and subscales was assessed by examining whether there were significant differences between subgroups on key SES variables. Criterion validity (i.e., an association with a “gold standard”) was assessed by examining the mean full and factor scores within strata of income. Construct validity (i.e., the scale behaves as expected) was determined by comparing FRS scores with variables such as educational attainment, perceived baseline familial SES, and hunger in the past 12 months. In addition to SES variables, selling drugs and sex, living situation, and gym membership were used to assess construct validity. Those who sold drugs or sex were hypothesized to have fewer material resources while those living with their parents or those having a gym membership would have more. Parametric tests (i.e., independent samples t tests, analyses of variance) were used for all pairwise comparisons of continuous variables. For pairwise comparisons that were highly skewed and/or heterogeneous, nonparametric tests were computed to confirm that p values were in fact similar. A Bonferroni correction was applied to adjust for multiple comparisons between subscales (α = .05/3 subscales = .017) to reduce the likelihood of Type I error.

Associations Between FRS and Depression

Potential differences among those with and without depression symptomology were examined, along the key SES and gay-related psychosocial variables. Categorical variables were compared with χ² tests unless a cell had fewer than 5 participants, in which case Fisher’s exact test was used. Multivariable logistic regression models were then constructed to determine the associations between the full-modified FRS and factor scores and depression, while controlling for potential confounders identified in the bivariable analyses. Effect modification was assessed with interaction terms in the models. Postestimation goodness-of-fit tests, Pearson’s χ² and Hosmer–Lemeshow, were calculated for each model. These analyses were conducted with STATA 12.1 (StataCorp LP, 2012).

Results

In the current sample of 482 respondents, the mean age was 21.8 years (SD = 0.7). With respect to race, 37.3% were Hispanic/Latino, 30.5% were White, 14.9% were Black, and 17.0% were Asian, Native American, multiracial, or another race. Approximately half (50.8%) identified as exclusively heterosexual. Most (70.8%) reported incomes less than $10,000 annually.

The EFA for the modified FRS is presented in Table 1. The EFA model had very good fit, χ²(102) = 268.77, p < .001; RMSEA = 0.058 (95% confidence interval [CI] = 0.050-0.067); SRMR = 0.034; CFI = 0.996; and TLI = 0.995, high internal consistency (full scale α = .91), and each item cleanly loaded onto separate factors. Factor 1 items were basic needs, including food, housing, money for necessities and monthly bills, clothes, heat, indoor plumbing/water, transport, sleep, furniture, and phone access. Factor 2 items include health and social services such as medical care, dental care, and public assistance. Factor 3 items were mainly economic resources such as a good job and money for savings, entertainment, and buying things for one’s self. Internal consistency for Factors 1 and 3 was excellent or good (α = .93 and .84, respectively). Internal consistency for Factor 2 was substantially lower, but acceptable (α = .65).

The criterion and construct validity results for the full FRS and subscales are presented in Table 2. There was a dose response relationship between income and the full FRS scores (p = .016) as well as scores for Factors 1 (basic needs, p = .013) and 3 (economic resources, p = .061) such that FRS scores increased as income increased, providing evidence of criterion validity. There was significant variability within income strata. For those with no income, the mean full-scale score was 4.00 (SD = 1.09, range = 0-5.00). The mean full-scale scores for incomes less than $5,000 were 4.08 (SD = 0.92, range = 0.23-5.00), 4.17 (SD = 0.90, range = 0.06-5.00) for less than $10,000, 4.13 (SD = 0.82, range = 0.50-5.00) for less than $20,000, and 4.52 (SD = 0.55, range = 2.67-5.00) for greater than or equal to $20,000.

Table 2.

Criterion and Construct Validity Analysis of the Revised 18-Item Family Resource Scale Among 482 YMSM in New York City.

	Full scale: Mean score (SD)	p value^a	Factor 1: Mean score (SD)	p value^a	Factor 2: Mean score (SD)	p value^a	Factor 3: Mean score (SD)	p Value^a
Race/ethnicity		<.001^c		<.001^c		<.001^c		.191
Black	4.01 (0.97)		4.32 (0.95)		3.36 (1.52)		3.66 (1.32)
Hispanic/Latino	3.96 (1.07)		4.34 (1.06)		3.29 (1.62)		3.42 (1.53)
White	4.39 (0.60)		4.75 (0.43)		3.98 (1.21)		3.73 (1.26)
Asian, Pacific Islander, or other race/ethnicity	4.24 (0.82)		4.59 (0.76)		3.67 (1.44)		3.70 (1.38)
Educational attainment		.062		.068		.334		.062
High school or less	4.10 (0.93)		4.46 (0.89)		3.57 (1.51)		3.50 (1.42)
Associate’s degree/some college	4.13 (0.96)		4.47 (0.96)		3.39 (1.51)		3.75 (1.47)
Bachelor’s degree or higher	4.35 (0.76)		4.70 (0.66)		3.75 (1.34)		3.87 (1.24)
Perceived baseline familial SES		<.001^c		.002^c		<.001^c		<.001^c
Lower	3.82 (0.96)		4.30 (0.92)		2.91 (1.66)		3.19 (1.44)
Middle	4.28 (0.87)		4.58 (0.83)		3.78 (1.32)		3.83 (1.36)
Upper	4.33 (0.80)		4.62 (0.79)		4.07 (1.19)		3.75 (1.30)
Income		.016^c		.013^c		.152		.061
None	4.00 (1.09)		4.38 (1.08)		3.63 (1.46)		3.24 (1.57)
<$5,000	4.08 (0.92)		4.42 (0.91)		3.66 (1.47)		3.47 (1.46)
<$10,000	4.17 (0.90)		4.55 (0.85)		3.53 (1.56)		3.59 (1.29)
<$20,000	4.13 (0.82)		4.51 (0.77)		3.29 (1.51)		3.73 (1.28)
>$20,000	4.52 (0.55)		4.76 (0.35)		3.83 (1.37)		4.39 (0.91)
Missing	4.19 (0.74)		4.64 (0.46)		3.42 (1.27)		3.54 (1.52)
Living situation^b		.375		.354		.160		.465
Alone	4.25 (0.80)		4.50 (0.75)		3.94 (1.24)		3.78 (1.29)
With my parents/family	4.11 (0.97)		4.46 (0.96)		3.45 (1.53)		3.64 (1.43)
With a friend/roommate	4.16 (0.89)		4.55 (0.79)		3.63 (1.53)		3.50 (1.42)
With a male lover/partner	4.17 (0.80)		4.56 (0.74)		3.51 (1.36)		3.57 (1.30)
Main income source		.005^c		<.001^c		.076		.011^c
Regular job, temporary job, or own business	4.13 (0.92)		4.48 (0.89)		3.49 (1.53)		3.66 (1.36)
Public assistance	3.47 (1.41)		3.53 (1.56)		4.27 (1.59)		2.73 (1.52)
Parent, friend, relative, or spouse	4.30 (0.69)		4.70 (0.52)		3.83 (1.31)		3.53 (1.38)
Illegal activities	3.63 (1.28)		4.12 (1.23)		3.08 (1.44)		2.71 (2.01)
Other	4.34 (0.71)		4.68 (0.46)		3.80 (1.45)		3.90 (1.15)
Hungry in the past 12 months		<.001^c,d		<.001^{b, c}		<.001^{b, c}		<.001^{b, c}
No	4.31 (0.81)		4.63 (0.77)		3.76 (1.37)		3.86 (1.25)
Yes	3.35 (0.10)		3.88 (0.98)		2.73 (1.71)		2.37 (1.39)
Sexual orientation		.016^{b, c}		.006^{b, c}		.228^b		.162^b
Exclusively homosexual	4.05 (1.01)		4.39 (0.99)		3.50 (1.53)		3.51 (1.46)
Not exclusively homosexual	4.25 (0.78)		4.61 (0.69)		3.66 (1.44)		3.69 (1.32)
Gay community affinity		.830		.983		.851		.279
Low	4.13 (0.83)		4.51 (0.71)		3.60 (1.40)		3.47 (1.46)
Medium	4.12 (1.00)		4.49 (0.97)		3.51 (1.54)		3.57 (1.41)
High	4.18 (0.91)		4.50 (0.89)		3.60 (1.51)		3.70 (1.34)
Gay-related stigma (personal)		.137^b		.332^b		.265^b		.070^b
Low	4.19 (0.90)		4.53 (0.86)		3.64 (1.51)		3.69 (1.37)
High	4.06 (0.91)		4.45 (0.86)		3.48 (1.43)		3.44 (1.42)
Gay-related stigma (public)		<.001^{b, c}		<.001^{b, c}		<.001^{b, c}		.006^{b, c}
Low	4.24 (0.84)		4.59 (0.75)		3.70 (1.41)		3.69 (1.38)
High	3.83 (1.06)		4.21 (1.12)		3.16 (1.67)		3.28 (1.41)
Internalized homophobia		.010^{b, c}		.017^b		.922^b		<.001^{b, c}
Low	4.19 (0.89)		4.54 (0.84)		3.59 (1.49)		3.69 (1.36)
High	3.87 (0.95)		4.25 (0.94)		3.56 (1.44)		3.03 (1.45)
Sold drugs		.092^b		.750^b		.057^b		.004^{b, c}
No	4.16 (0.91)		4.50 (0.87)		3.61 (1.47)		3.65 (1.37)
Yes	3.87 (0.84)		4.45 (0.66)		3.06 (1.66)		2.87 (1.61)
Sold sex		.010^{b, c}		.021^b		<.001^{b, c}		.026^b
No	4.18 (0.89)		4.53 (0.84)		3.64 (1.44)		3.64 (1.37)
Yes	3.72 (1.03)		4.19 (1.02)		2.79 (1.78)		3.11 (1.62)
Gym membership		.012^{b, c}		.411^b		.001^{b, c}		<.001^{b, c}
No	4.08 (0.91)		4.48 (0.85)		3.43 (1.56)		3.45 (1.42)
Yes	4.29 (0.88)		4.55 (0.88)		3.89 (1.27)		3.90 (1.30)
Depression		<.001^{b, c}		<.001^{b, c}		.031^b		<.001^{b, c}
No	4.24 (0.86)		4.59 (0.82)		3.65 (1.45)		3.76 (1.34)
Yes	3.69 (0.98)		4.11 (0.95)		3.26 (1.62)		2.88 (1.45)

Note. SES = socioeconomic status; YMSM = young men who have sex with men.

Analysis of variance unless otherwise specified. ^bExcluded two respondents who lived in a shelter or were homeless. ^ct test. ^dSignificant after Bonferroni adjustment.

With respect to construct validity, the full FRS, Factor 1, and Factor 3 generally behaved as expected when compared with educational attainment (p = .062, p = .068, and p = .062, respectively), perceived baseline familial SES (p < .001, p = .002, and p < .001, respectively), main income source (p = .005, p < .001, and p = .011, respectively), and hunger (all p < .001)—lower SES positions were associated with lower FRS scores. Those who sold drugs (p = .092 and p = .004 for full FRS and Factor 3, respectively) or sex (p = .010, p = .021, and p = .026 for full FRS, Factor 1, and Factor 3, respectively) generally had lower FRS scores whereas those with gym memberships had significantly higher scores for full FRS (p = .012) and Factor 3 (p < .001). Living situation was not significantly associated with either the full scale or any of the factors. These findings suggest that there is construct validity for the full FRS as well as Factors 1 and 3.

Factor 2, health and social services, demonstrated different patterns of association with the SES variables (Table 2). Educational attainment was not significantly associated with the Factor 2 FRS scores. A significant dose response relationship was observed between the Factor 2 score and perceived baseline familial SES (p < .001), but not income (p = .152). With respect to income sources, those on public assistance had higher factor scores as compared with all other groups (p = .076). Although this finding was not statistically significant, these findings also suggest construct validity, as these associations are in accordance with health and social service eligibility (i.e., those on public assistance are likely eligible for Medicaid). Finally, those who sold drugs and sex had lower Factor 2 scores (p = .057 and p < .001, respectively) whereas those with gym memberships had higher scores (p = .001).

The associations between the FRS, sexual orientation, and gay-related psychosocial variables are presented in Table 2. For the full FRS and Factor 1, those who identified as exclusively homosexual had significantly lower FRS scores relative to those who did not (p = .016 and p = .006, respectively). Gay community affinity and personal gay-related stigma were not associated with material resources, however high public gay-related stigma (all p ≤ .006) and internalized homophobia (p = .010, p = .014, and p < .001 for full FRS, Factor 1, and Factor 3, respectively) were associated with lower FRS scores.

The BDI-II rated 17.6% of the sample as meeting criteria for mild or greater depressive symptomology. In bivariable analyses (Table 2), those who reported depression had significantly lower material scores for the whole FRS (p < .001) as well as each factor (p < .001, p = .031, and p < .001 for Factor 1, and Factor 3, respectively). Table 3 presents the bivariable analyses of sociodemographic and gay-related psychosocial correlates of depression. YMSM who were depressed were significantly more likely to report being hungry in the past 12 months (p < .001), lower gay community affinity (p = .042), higher gay-related public stigma (p = .005), and higher internalized homophobia (p < .001). Income was not significantly associated with depression.

Table 3.

Bivariable Correlates of Current Depression Among 482 YMSM in New York City.

	n (%)	Depression
	n (%)	No (n = 397)	Yes (n = 85)	p value^a
Race/ethnicity				.234
Black	72 (14.9)	58 (80.6)	14 (19.4)
Hispanic/Latino	180 (37.3)	145 (80.6)	35 (19.4)
White	147 (30.5)	119 (81.0)	28 (19.1)
Asian, Pacific Islander or other race/ethnicity	82 (17.0)	74 (90.2)	8 (9.8)
Educational attainment				.447
High school or less	330 (68.5)	267 (80.9)	63 (19.1)
Associate’s degree/some college	62 (12.9)	54 (87.1)	8 (12.9)
Bachelor’s degree or higher	88 (18.3)	74 (84.1)	14 (15.9)
Perceived baseline familial SES				.162
Lower	157 (32.6)	122 (77.7)	35 (22.3)
Middle	184 (38.3)	154 (83.7)	30 (16.3)
Upper	140 (29.1)	120 (85.7)	20 (14.3)
Income^c				.247^b
None	91 (18.9)	74 (81.3)	17 (18.7)
<$5,000	130 (27.0)	99 (76.2)	31 (23.9)
<$10,000	120 (24.9)	106 (88.3)	14 (11.7)
<$20,000	75 (15.6)	61 (82.4)	13 (17.6)
>$20,000	22 (4.6)	19 (86.4)	3 (13.6)
Living situation^e				.608
Alone	61 (12.7)	49 (80.3)	12 (19.7)
With my parents/family	206 (42.9)	171 (83.0)	35 (17.0)
With a friend/roommate	169 (35.2)	136 (80.5)	33 (19.5)
With a male lover/partner	44 (9.2)	39 (88.6)	5 (11.4)
Main income source				.153^b
Regular job, temporary job, or own business	332 (68.9)	279 (84.3)	52 (15.7)
Public assistance	10 (2.1)	7 (70.0)	3 (30.0)
Parent, friend, relative, or spouse	96 (19.9)	76 (79.2)	20 (20.8)
Illegal activities	17 (3.5)	11 (64.7)	6 (35.3)
Other	24 (5.0)	20 (83.3)	4 (16.7)
Hungry in the past 12 months				<.001^d
No	397 (82.4)	341 (85.9)	56 (14.1)
Yes	84 (17.4)	55 (65.5)	29 (34.5)
Mean material resources score (SD)	4.1 (0.9)	4.2 (0.9)	3.7 (1.0)	<.001^d
Sexual orientation				.138
Exclusively homosexual	245 (50.8)	208 (84.9)	37 (15.1)
Not exclusively homosexual	237 (49.2)	189 (79.8)	48 (20.3)
Gay community affinity				.042
Low	144 (29.9)	109 (75.7)	35 (24.3)
Medium	121 (25.1)	104 (86.0)	17 (14.1)
High	217 (45.0)	184 (84.8)	33 (15.2)
Gay-related stigma (personal)				.424
Low	322 (67.2)	268 (83.2)	54 (16.8)
High	157 (32.8)	126 (80.3)	31 (19.8)
Gay-related stigma (public)				.005^d
Low	373 (77.4)	317 (85.0)	56 (15.0)
High	109 (22.6)	80 (73.4)	29 (26.6)
Internalized homophobia				<.001^d
Low	422 (87.7)	360 (85.3)	62 (14.7)
High	59 (12.3)	36 (61.0)	23 (39.0)

Note. SES = socioeconomic status; YMSM = young men who have sex with men.

χ² unless otherwise specified. ^bFisher’s exact test. ^cIncome was missing for 32 (6.6%) respondents. ^dSignificant after Bonferroni adjustment. ^eExcluded two respondents who lived in a shelter or were homeless.

Multivariable logistic regression models examining conditional associations between FRS measures and depression are presented in Table 4. The relationship between depression and the full FRS scale as well as each factor individually (Models 1-4) was examined, controlling for gay-related psychosocial variables that remained significant in the models. Material resources were significantly and negatively associated with depression, such that a one-point increase in the full FRS score was associated with decreased odds of depression (Model 1; adjusted odds ratio [AOR] = 0.58; 95% CI = 0.46-0.74) after controlling for gay community affinity and internalized homophobia. Similarly, each factor of the FRS was also associated with a 16% to 36% decreased odds for depression, after controlling for gay community affinity (Model 2; Factor 1 only) and internalized homophobia (Models 2-4; all three factors). Model 5 includes all three factors in the model to determine whether one factor was particularly associated with depression, over and above other factors—only Factor 3 (economic resources) remained significant in the model after controlling for internalized homophobia (AOR = 0.72, 95% CI = 0.58-0.89).

Table 4.

Multivariable Logistic Regression Models of the Correlates of Current Depression Among 482 YMSM in New York City.

	Crude Odds Ratios (95% CI)	Model 1: AOR (95% CI)	Model 2: AOR (95% CI)	Model 3: AOR (95% CI)	Model 4: AOR (95% CI)	Model 5: AOR (95% CI)
Full FRS score	0.58 (0.46, 0.73)	0.58 (0.46, 0.74)
Factor 1 score: Basic needs	0.61 (0.49, 0.77)		0.62 (0.49, 0.78)			0.78 (0.57, 1.07)
Factor 2 score: Health and social services	0.85 (0.73, 0.99)			0.84 (0.72, 0.98)		1.06 (0.87, 1.29)
Factor 3 score: Economic resources	0.66 (0.56, 0.77)				0.67 (0.57, 0.80)	0.72 (0.58, 0.89)
Gay community affinity
Low	1.00	1.00	1.00
Medium	0.51 (0.27, 0.97)	0.51 (0.26, 1.00)	0.51 (0.26, 1.00)
High	0.56 (0.33, 0.95)	0.63 (0.36, 1.10)	0.61 (0.35, 1.08)
Gay-related stigma (public)
Low	1.00
High	2.05 (1.23, 3.41)
Internalized homophobia
Low	1.00	1.00	1.00	1.00	1.00	1.00
High	3.70 (2.05, 6.66)	3.06 (1.65, 5.67)	3.10 (1.68, 5.73)	3.77 (2.08, 6.82)	3.07 (1.67, 5.65)	3.02 (1.63, 5.58)

Note. AOR = adjusted odds ratio; FRS = Family Resource Scale; YMSM = young men who have sex with men.

Higher gay community affinity was also associated with decreased odds of depression, after controlling for internalized homophobia and material resources for the full FRS and Factor 1 models only. Internalized homophobia was consistently associated with a 3.02 to 3.77 increased odds for depression in all models. Effect modification was assessed between the full FRS and individual factor scores and internalized homophobia; significant interactions were not identified.

Both the postestimation Pearson’s χ² and Hosmer–Lemeshow tests demonstrated good fit for Models 3 to 4, with p < .1. Model 1 demonstrated good fit as assessed by Pearson’s χ² (p = .151) but not the Hosmer–Lemeshow test (p = .006). Model 2 demonstrated good fit as assessed by the Hosmer–Lemeshow test (p = .108) but not Pearson’s χ² (p = .005). Given that there were fewer covariate patterns as compared with observations for both Model 1 (171 and 481, respectively) and Model 2 (107 and 481, respectively), the Pearson’s χ² is the most appropriate test. Thus, Models 3 to 5 fit the data well whereas Model 1, which included the full modified FRS scale, did not.

Discussion

There are three key findings from this analysis of the modified FRS: (1) the modified FRS demonstrated reliability and validity as a measure of material resources among YMSM, (2) YMSM who identified as exclusively homosexual and experienced high levels of gay-related public stigma and internalized homophobia had fewer material resources, and (3) depressed YMSM had significantly fewer material resources and were more likely to experience internalized homophobia. These results and their implications are discussed below and demonstrate the critical role that social conditions hold in shaping the health of sexual minority men.

These analyses support the reliability of the modified FRS among urban YMSM. The full FRS, Factor 1 (basic needs), and Factor 3 (economic resources) demonstrated good-to-excellent reliability. Factor 2 (health and social services) had acceptable reliability. Cronbach’s alpha for the modified FRS was .91, which was exactly the same reliability estimate for the FRS among economically disadvantaged current, former, and nondrug users in NYC (Ompad et al., 2012). The factors identified among this sample of YMSM were also similar to those identified in a previous validation study (Ompad et al., 2012). There was also support for criterion and construct validity of the modified FRS among YMSM. For criterion validity, the overall, Factor 1, and Factor 3 scores increased as income increased which is consistent with previous validity studies (Brannan, Manteuffel, Holden, & Heflinger, 2006; Ompad et al., 2012). With respect to construct validity, the overall FRS and its subscales performed as expected with other SES variables, illegal activities, and gym membership.

Factor 2 is unlike the other factors as it reflects, in part, income and services provided to low-income individuals by the state or federal government to alleviate some of the effects of poverty. Factor 2 has been retained in the modified FRS for three reasons. First, the scale was kept intact so that the performance of the scale could be compared with different populations and settings. Second, although both the official and supplemental U.S. poverty threshold calculations exclude Medicaid and Medicare benefits from their calculations, public assistance income is included in the calculations (DeNavas-Walt & Proctor, 2015; Short, 2015) and is a part of Factor 2. Third, excluding important resources such as health and dental services would result in the underestimation of material resources.

Based on self-reported income, 70.8% of the P18 sample reportedly earned ≤$10,000 per year and were thus living below the federal poverty threshold, which was $11,490 in 2013 (U.S. Census Bureau, 2013a). YMSM who identified as exclusively homosexual had significantly fewer material resources as measured by the full FRS and Factor 1 (basic needs) scores as compared with those who did not. This is consistent with analyses of the Add Health where consistently identifying as gay over time was associated with decreased earnings among gay men (Sabia, 2015). Several researchers have suggested that income inequality between LGB and heterosexual workers may be due to discrimination (Badgett, 1995; Klawitter, 2011), occupational clustering (Blandford, 2003), or marital status (Allegretto & Arthur, 2001), and run counter to commonly held beliefs that sexual minority individuals are more affluent than their heterosexual peers (Gates & Newport, 2012). Others have suggested that stress related to discrimination affects job performance and thus earnings (Badgett, 1995). In this study, YMSM who reported high levels of public gay-related stigma and internalized homophobia had fewer material resources. These findings suggest that the underlying causes of these resources gaps could reflect both external and internal processes and the critical role that social and psychosocial factors may play in the well-being of young sexual minority men. Previous research has clearly documented the detrimental effects of state-sanctioned discrimination on the health of sexual minorities (Halkitis, 2012a, 2012b; Hatzenbuehler, McLaughlin, Keyes, & Hasin, 2010; Hatzenbuehler, Phelan, & Link, 2013) and these conditions have been noted as drivers of the overlapping health problems (i.e., the syndemic) faced by sexual minority men across the age spectrum (Halkitis et al., 2012; Halkitis et al., 2014; Halkitis & Figueroa, 2013; Halkitis, Wolitski, et al., 2013). These findings provide evidence that SES may be socially determined by sexual orientation.

Depression was significantly associated with material resources such that for each one point increase in the FRS score there was a 42% decrease in the odds of depression after controlling for gay community affinity and internalized homophobia. When each FRS factor (i.e., basic needs, health and social services, and economic resources) was examined individually, decreased odds for depression with higher material resources was observed. When all three factors were placed in the same model, only Factor 3 (economic resources) remained significant after controlling for internalized homophobia. These findings suggest that SES is likely a key determinant of depression among YMSM, even after controlling for factors associated with resilience (i.e., gay community affinity) and risk (i.e., internalized homophobia).

Gay-related psychosocial variables were associated with depression, even after controlling for SES. Internalized homophobia was a strong and consistent correlate of depression among YMSM in the P18 study, even after adjustment for SES. Newcomb and Mustanski (2010) conducted a meta-analysis of the association between internalized homophobia and internalizing mental health outcomes and observed small-to-moderate associations, with stronger associations among LGB older samples. The relationship between internalized homophobia and depression is modified by geography, such that the relationship was only significant for urban MSM in a recent Australian study (McLaren, 2015). These findings are consistent with the growing body of literature that indicates that internalized homophobia is an important psychosocial correlate of health. Higher gay community affinity was associated with decreased odds of depression. Mao et al. (2009) reported that lower gay community involvement was associated with increased odds major depressive disorder. This finding suggests that gay community involvement may be a factor that contributes to resilience related to mental health outcomes.

It is important to note that depression was not associated with income in this sample. This finding is similar to findings from Gamarel, Reisner, Parsons, and Golub (2012) who did not find an association between income, measured as <$12,000 versus ≥$12,000, and depression among gay and bisexual men in NYC. This is inconsistent with other studies among MSM (De Santis et al., 2008; Mao et al., 2009) where lower income was associated with depression. Income data are highly susceptible to reporting biases, which may partially explain these inconsistencies and justify the use of nonmonetary measures of SES. In this sample, 66.2% reported receiving any money from a parent, friend, relative, boyfriend, spouse, or sugar daddy’s income and 14.3% received money from theft, robbing, stealing, selling drugs, or sex for money regardless of their main income source (data not reported). For populations that may receive substantial income from people in their lives or illegal sources, income may not be the best measure of SES (Arnett, 2000; Ompad et al., 2012).

The limitations of the study should be considered when interpreting these results. First, the men in this study lived in NYC—a city with a high cost of living, a high median income, and likely a more gay-friendly environment than some other U.S. cities. The extent to which the sample of YMSM and their experiences are generalizable to other YMSM may be limited. Of particular salience to this analysis, the distress associated with similar incomes in settings with lower costs of living may not be as severe. Personal incomes may not reflect the entirety of the men’s material resources—although 19.9% of the P18 men reported getting most of their money from a parent, friend, relative, or spouse, 66.2% reported getting any money from someone else (data not reported). As such, monetary income may be underestimated. The FRS is not an income-based measure of SES and is less likely to significantly underestimate material resources.

Despite these limitations, the modified 18-item FRS scale shows promise as an additional measure of socioeconomic disadvantage, particularly among racial and sexual minority youth. Nonincome measures of SES, such as the modified FRS scale, may provide a more accurate and reliable measurement approach. There was notable variability of the full FRS scale scores within income strata, suggesting that income may underestimate economic resources among youth and providing additional evidence for the limited utility of income as an SES measure.

In summary, this racially and ethnically diverse sample of YMSM in NYC experienced substantial financial hardship and those who were more gay-identified had fewer material resources. Fewer material resources and internalized homophobia were both associated with higher odds of depression. The modified FRS has been validated among YMSM and is a good nonincome based measure SES. These results suggest that SES may be a fundamental cause of health among gay, bisexual, and other MSM, and provide evidence for the detrimental effects that social conditions, both in terms of finances and discrimination, may play in shaping the health of gay men. This research supports the need for structural interventions to ameliorate multiple health problems faced by sexual minority individuals and to more broadly address the multiple health disparities noted in the lesbian, gay, bisexual, and transgender population (2011).

Footnotes

Appendix

		Never	Rarely	Less than ½ the time	About ½ the time	More than ½ the time	Always
READ: I would like to start off by asking about whether or not you and your family have adequate resources to meet the needs of the family as a whole as well as the needs of individual family members. Right now, how often are the following needs adequately met? PROBE: By adequate I mean meets your needs
1.	Food for two meals a day	0	1	2	3	4	5
2.	House or apartment	0	1	2	3	4	5
3.	Money to buy necessities	0	1	2	3	4	5
4.	Enough clothes for you or your family	0	1	2	3	4	5
5.	Heat for your house or apartment	0	1	2	3	4	5
6.	Indoor plumbing or water	0	1	2	3	4	5
7.	Money to pay monthly bills	0	1	2	3	4	5
8.	Good job for yourself or your spouse	0	1	2	3	4	5
9.	Medical care for you or your family	0	1	2	3	4	5
10.	Public assistance (SSI, AFDC, Medicaid, etc.)	0	1	2	3	4	5
11.	Dependable transportation	0	1	2	3	4	5
12.	Time to get enough sleep/rest	0	1	2	3	4	5
13.	Furniture for your home or apartment	0	1	2	3	4	5
14.	Telephone or access to phone	0	1	2	3	4	5
15.	Dental care for you or your family	0	1	2	3	4	5
16.	Money to buy things for yourself	0	1	2	3	4	5
17.	Money for entertainment	0	1	2	3	4	5
18.	Money to save	0	1	2	3	4	5

Acknowledgements

The authors would like to thank the P18 study staff and participants.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was funded by the National Institute on Drug Abuse (NIDA; R01DA025537). DCO was supported by the Center for Drug Use and HIV Research (CDUHR–P30 DA011041). JJP was supported by a NIDA career development award (K01 DA038800).

References

Albelda

Badgett

M. V. L.

Schneebaum

Gates

(2009). Poverty in the lesbian, gay, and bisexual community. Los Angeles, CA: Williams Institute.

Allegretto

S. A.

Arthur

M. M.

(2001). An empirical analysis of homosexual/heterosexual male earnings differentials: Unmarried and unequal? Industrial & Labor Relations Review, 54, 631-646. doi:10.2307/2695994

Arnett

J. J.

(2000). Emerging adulthood—A theory of development from the late teens through the twenties. American Psychologist, 55, 469-480. doi:10.1037//0003-066x.55.5.469

Badgett

M. V. L.

(1995). The wage effects of sexual orientation discrimination. Industrial & Labor Relations Review, 48, 726-739. doi:10.2307/2524353

Badgett

M. V. L.

Durso

L. E.

Schneebaum

(2013). New patterns of poverty in the lesbian, gay, and bisexual community. Retrieved from http://williamsinstitute.law.ucla.edu/wp-content/uploads/LGB-Poverty-Update-Jun-2013.pdf

Beck

A. T.

Steer

R. A.

Ball

Ranieri

W. F.

(1996). Comparison of Beck Depression Inventories–IA and –II in psychiatric outpatients. Journal of Personality Assessment, 67, 588-597. doi:10.1207/s15327752jpa6703_13

Bentler

P. M.

(1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238-246. doi:10.1037/0033-2909.107.2.238

Berger

B. E.

Ferrans

C. E.

Lashley

F. R.

(2001). Measuring stigma in people with HIV: Psychometric assessment of the HIV stigma scale. Research in Nursing & Health, 24, 518-529. doi:10.1002/nur.10011

Besharov

D. J.

Couch

(2009). European measures of income, poverty, and social exclusion: Recent developments and lessons for U.S. poverty measurement. Journal of Policy Analysis and Management, 28, 713-715.

10.

Blandford

J. M.

(2003). The nexus of sexual orientation and gender in the determination of earnings. Industrial & Labor Relations Review, 56, 622-642. doi:10.2307/3590960

11.

Brannan

A. M.

Manteuffel

Holden

E. W.

Heflinger

C. A.

(2006). Use of the Family Resource Scale in children’s mental health: Reliability and validity among economically diverse samples. Administration and Policy in Mental Health, 33, 182-197. doi:10.1007/s10488-006-0032-8

12.

Carpenter

C. S.

(2007). Revisiting the income penalty for behaviorally gay men: Evidence from NHANES III. Labour Economics, 14, 25-34. doi:10.1016/j.labeco.2005.06.001

13.

Cronbach

L. J.

(1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297-334. doi:10.1007/BF02310555

14.

De Santis

J. P.

Colin

J. M.

Provencio Vasquez

McCain

G. C

. (2008). The relationship of depressive symptoms, self-esteem, and sexual behaviors in a predominantly Hispanic sample of men who have sex with men. American Journal of Men’s Health, 2, 314-321. doi:10.1177/1557988307312883

15.

DeNavas-Walt

Proctor

B. D.

(2015). Income and poverty in the United States: 2014. Current Population Reports. Retrieved from https://www.census.gov/content/dam/Census/library/publications/2015/demo/p60-252.pdf

16.

Dorling

(1999). Who’s afraid of income inequality? Environment and Planning A, 31, 571-574.

17.

Dunst

C. J.

Leet

H. E.

(1987). Measuring the adequacy of resources in households with young children. Child: Care, Health and Development, 13, 111-125.

18.

Dyer

T. P.

Regan

Pacek

L. R.

Acheampong

Khan

M. R.

(2015). Psychosocial vulnerability and HIV-related sexual risk among men who have sex with men and women in the United States. Archives of Sexual Behavior, 44, 429-441. doi:10.1007/s10508-014-0346-7

19.

Gamarel

K. E.

Reisner

S. L.

Parsons

J. T.

Golub

S. A.

(2012). Association between socioeconomic position discrimination and psychological distress: Findings from a community-based sample of gay and bisexual men in New York City. American Journal of Public Health, 102, 2094-2101. doi:10.2105/AJPH.2012.300668

20.

Gates

G. J.

Newport

(2012). Special report: 3.4% of U.S. adults identify as LGBT: Inaugural Gallup findings based on more than 120,000 interviews. Retrieved from http://www.gallup.com/poll/158066/special-report-adults-identify-lgbt.aspx

21.

Halkitis

P. N.

(2010). Reframing HIV prevention for gay men in the United States. American Psychologist, 65, 752-763. doi:10.1037/0003-066X.65.8.752

22.

Halkitis

P. N.

(2012a, April). Discrimination and homophobia fuel the HIV epidemic in gay and bisexual men. Psychology AIDS Exchange Newsletter, pp. 1, 4-11.

23.

Halkitis

P. N.

(2012b). Obama, marriage equality, and the health of gay men. American Journal of Public Health, 102, 1628-1629. doi:10.2105/AJPH.2012.300940

24.

Halkitis

P. N.

Figueroa

R. P.

(2013). Sociodemographic characteristics explain differences in unprotected sexual behavior among young HIV-negative gay, bisexual, and other YMSM in New York City. AIDS Patient Care and STDs, 27, 181-190. doi:10.1089/apc.2012.0415

25.

Halkitis

P. N.

Kapadia

Siconolfi

D. E.

Moeller

R. W.

Figueroa

R. P.

Barton

S. C.

Blachman-Forshay

(2013). Individual, psychosocial, and social correlates of unprotected anal intercourse in a new generation of young men who have sex with men in New York City. American Journal of Public Health, 103, 889-895. doi:10.2105/AJPH.2012.300963

26.

Halkitis

P. N.

Kupprat

S. A.

Hampton

M. B.

Perez-Figueroa

Kingdon

Eddy

J. A.

Ompad

D. C.

(2012). Evidence for a syndemic in aging HIV-positive gay, bisexual, and other MSM: Implications for a holistic approach to prevention and healthcare. Annals of Anthropological Practice, 36, 365-386. doi:10.1111/napa.12009

27.

Halkitis

P. N.

Moeller

R. W.

Siconolfi

D. E.

Storholm

E. D.

Solomon

T. M.

Bub

K. L.

(2013). Measurement model exploring a syndemic in emerging adult gay and bisexual men. AIDS and Behavior, 17, 662-673. doi:10.1007/s10461-012-0273-3

28.

Halkitis

P. N.

Perez-Figueroa

R. E.

Carreiro

Kingdon

M. J.

Kupprat

S. A.

Eddy

(2014). Psychosocial burdens negatively impact HIV antiretroviral adherence in gay, bisexual, and other men who have sex with men aged 50 and older. AIDS Care, 26, 1426-1434. doi:10.1080/09540121.2014.921276

29.

Halkitis

P. N.

Wolitski

R. J.

Millett

G. A.

(2013). A holistic approach to addressing HIV infection disparities in gay, bisexual, and other men who have sex with men. American Psychologist, 68, 261-273. doi:10.1037/a0032746

30.

Hatzenbuehler

M. L.

McLaughlin

K. A.

Keyes

K. M.

Hasin

D. S.

(2010). The impact of institutional discrimination on psychiatric disorders in lesbian, gay, and bisexual populations: A prospective study. American Journal of Public Health, 100, 452-459. doi:10.2105/AJPH.2009.168815

31.

Hatzenbuehler

M. L.

Phelan

J. C.

Link

B. G.

(2013). Stigma as a fundamental cause of population health inequalities. American Journal of Public Health, 103, 813-821. doi:10.2105/Ajph.2012.301069

32.

Institute of Medicine, Committee on Lesbian Gay Bisexual and Transgender Health Issues and Research Gaps and Opportunities. (2011). The health of lesbian, gay, bisexual, and transgender people: Building a foundation for better understanding. Washington, DC: National Academies Press.

33.

Kaiser

H. F.

(1960). The application of electronic-computers to factor-analysis. Educational and Psychological Measurement, 20, 141-151. doi:10.1177/001316446002000116

34.

Kinsey

A. C.

Pomeroy

W. B.

Martin

C. E.

(1948). Sexual behavior in the human male. London, England: W. B. Saunders.

35.

Klawitter

(2011). Multilevel analysis of the effects of antidiscrimination policies on earnings by sexual orientation. Journal of Policy Analysis and Management, 30, 334-358. doi:10.1002/Pam.20563

36.

Klawitter

(2015). Meta-analysis of the effects of sexual orientation on earnings. Industrial Relations, 54, 4-32. doi:10.1111/Irel.12075

37.

Lauby

J. L.

Milnamow

(2009). Where MSM have their first HIV test: Differences by race, income, and sexual identity. American Journal of Men’s Health, 3, 50-59. doi:10.1177/1557988307313465

38.

Link

B. G.

Phelan

(1995). Social conditions as fundamental causes of disease. Journal of Health and Social Behavior, Spec No., 80-94.

39.

Mao

Kidd

M. R.

Rogers

Andrews

Newman

C. E.

Booth

. . . Kippax

S. C.

(2009). Social factors associated with major depressive disorder in homosexually active, gay men attending general practices in urban Australia. Australian and New Zealand Journal of Public Health, 33, 83-86. doi:10.1111/j.1753-6405.2009.00344.x

40.

Mayer

K. H.

Wang

Koblin

Mannheimer

Magnus

del Rio

. . . Wheeler

(2014). Concomitant socioeconomic, behavioral, and biological factors associated with the disproportionate HIV infection burden among Black men who have sex with men in 6 U.S. cities. PLoS One, 9(1), e87298. doi:10.1371/journal.pone.0087298

41.

McGarrity

L. A.

(2014). Socioeconomic status as context for minority stress and health disparities among lesbian, gay, and bisexual individuals. Psychology of Sexual Orientation and Gender Diversity, 1, 383-397. doi:10.1037/sgd0000067

42.

McLaren

(2015). Gender, age, and place of residence as moderators of the internalized homophobia-depressive symptoms relation among Australian gay men and lesbians. Journal of Homosexuality, 62, 463-480. doi:10.1080/00918369.2014.983376

43.

Muthén

L. K.

Muthén

B. O.

(2007). Mplus user’s guide (6th ed.). Los Angeles, CA: Muthén & Muthén.

44.

Newcomb

M. E.

Mustanski

(2010). Internalized homophobia and internalizing mental health problems: A meta-analytic review. Clinical Psychology Review, 30, 1019-1029. doi:10.1016/j.cpr.2010.07.003

45.

Ompad

D. C.

Nandi

Cerda

Crawford

Galea

Vlahov

(2012). Beyond income: Material resources among drug users in economically-disadvantaged New York City neighborhoods. Drug and Alcohol Dependence, 120, 127-134. doi:10.1016/j.drugalcdep.2011.07.008

46.

Pratt

L. A.

Brody

D. J.

(2014). Depression in the U.S. household population, 2009-2012. NCHS Data Brief, 12(172), 1-8.

47.

Reiss

(2013). Socioeconomic inequalities and mental health problems in children and adolescents: A systematic review. Social Science & Medicine, 90, 24-31. doi:10.1016/j.socscimed.2013.04.026

48.

Rosenfeld

(2010). “The meaning of poverty” and contemporary quantitative poverty research. British Journal of Sociology, 61(Suppl.), 103-110.

49.

Sabia

J. J.

(2015). Fluidity in sexual identity, unmeasured heterogeneity, and the earnings effects of sexual orientation. Industrial Relations, 54, 33-58. doi:10.1111/Irel.12076

50.

Sen

(1976). Poverty: An ordinal approach to measurement. Econometrica, 44, 219-231.

51.

Short

(2015, September). The Supplemental Poverty Measure: 2014. Current Population Reports. Retrieved from https://www.census.gov/content/dam/Census/library/publications/2015/demo/p60-254.pdf

52.

StataCorp

. (2012). Stata/SE 12.1 for Windows. College Station, TX: Author.

53.

Stevens

J. P.

(1992). Applied multivariate statistics for the social sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

54.

Thiede

Valleroy

L. A.

MacKellar

D. A.

Celentano

D. D.

Ford

W. L.

Hagan

. . . Torian

L. V.

(2003). Regional patterns and correlates of substance use among young men who have sex with men in 7 US urban areas. American Journal of Public Health, 93, 1915-1921.

55.

Thomeer

M. B.

(2013). Sexual minority status and self-rated health: The importance of socioeconomic status, age, and sex. American Journal of Public Health, 103, 881-888. doi:10.2105/AJPH.2012.301040

56.

Tucker

L. R.

Lewis

(1973). Reliability coefficient for maximum likelihood factor-analysis. Psychometrika, 38, 1-10. doi:10.1007/Bf02291170

57.

Turrell

(2000). Income non-reporting: Implications for health inequalities research. Journal of Epidemiology & Community Health, 54, 207-214.

58.

U.S. Census Bureau. (2013a). Poverty thresholds for 2013 by size of family and number of related children under 18 years. Washington, DC: Author.

59.

U.S. Census Bureau. (2013b). Table 1. Household characteristics of opposite-sex and same-sex couple households: ACS 2013. Washington, DC: Author.

60.

Van Horn

M. L.

Bellis

J. M.

Snyder

S. W

. (2001). Family resource scale-revised: Psychometrics and validation of a measure of family resources in a sample of low-income families. Journal of Psychoeducational Assessment, 19, 54-68. doi:10.1177/073428290101900104

61.

Wohl

A. R.

Galvan

F. H.

Carlos

J. A.

Myers

H. F.

Garland

Witt

M. D.

. . . George

(2013). A comparison of MSM stigma, HIV stigma and depression in HIV-positive Latino and African American men who have sex with men (MSM). AIDS and Behavior, 17, 1454-1464. doi:10.1007/s10461-012-0385-9

Reliability and Validity of a Material Resources Scale and Its Association With Depression Among Young Men Who Have Sex With Men: The P18 Cohort Study

Abstract

Keywords

Introduction

Method

Sample

Data Collection

Analysis

Validation of the FRS

Associations Between FRS and Depression

Results

Discussion

Footnotes

Appendix

Acknowledgements

Declaration of Conflicting Interests

Funding

References