Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Oct 21.
Published in final edited form as: Hisp J Behav Sci. 2024 Oct 21;46(2):59–78. doi: 10.1177/07399863241292015

Factor Structure and Measurement Invariance of the Cultural Attitudes Toward Healthcare and Mental Illness Questionnaire Among Latino Sexual Minority Men

Alyssa Lozano 1, Tae Kyoung Lee 2, Alejandra Fernandez 3, Guillermo Prado 1, Steven A Safren 1, Daniel E Jimenez 4, Audrey Harkness 1
PMCID: PMC12439767  NIHMSID: NIHMS2073729  PMID: 40964153

Abstract

Mental health disparities among Latino sexual minority men (LSMM) are exacerbated by un(der)treated mental health needs. This study sought to establish the factor structure of the Cultural Attitudes Toward Healthcare and Mental Illness Questionnaire (CAHMIQ) and examine measurement invariance across nativity (born in vs. outside the continental U.S.). Participants included 290 LSMM in South Florida. Confirmatory factor analysis examined the factor structure of CAHMIQ and measurement invariance examined equivalence across LSMM born outside (51.4%) and inside (48.3%) the continental U.S. The second-order factor model showed an acceptable fit (CFI/RMSEA = .91/.03) and consisted of first-ordered factors representing manifest indicators of causes and supports for mental health. Measurement invariance results suggest that the CAHMIQ performs equivalently across nativity groups (Δχ2(18) = 16.941, p = .5272). The CAHMIQ may be used among both nativity groups to understand attitudes toward healthcare and mental illness and inform LSMMs’ engagement in mental health treatment.

Keywords: Latino, sexual minority men, mental health, nativity, psychometrics

Sexual minority men (SMM)1 experience significant mental health disparities compared to heterosexual men, including anxiety, panic, social phobia, depression, and post-traumatic stress disorders (Cohen et al., 2016). According to a nationally representative study, Latino sexual minority men (LSMM) reported a higher prevalence of mental health disorders relative to non-Hispanic White sexual minority individuals (Rodriguez-Seijas et al., 2019). Moreover, in a study of immigrant Latino sexual minorities, more than 60% of the sample reported clinically significant depressive symptoms and a majority were not receiving treatment (Rhodes et al., 2013).

The minority stress framework (Meyer, 2003) posits that stigma-related stressors, including but not limited to rejection, hate crimes, internalized heterosexism, and/or identity concealment, fuel mental health disparities experienced by sexual minority populations. Additionally, sexual minority individuals who also identify with a minoritized racial/ethnic group may experience additional and intersecting stressors related to racism, ethnocentrism, and cultural factors that may also influence their mental health, as posited by intersectionality theory (Crenshaw, 1991). Considering the mental health disparities faced by LSMM and the role racial/ethnic and sexual discrimination play in these disparities, it is important to understand LSMM’s utilization of mental healthcare resources to improve their mental health.

Research examining mental healthcare utilization among LSMM is limited, and more research is needed to understand barriers and facilitators to the utilization of mental health treatments (Harkness et al., 2023). In one study of a racially and ethnically diverse group of sexual minority men between the ages of 16 and 20 (20% Latino), less than half of those who were diagnosed with a major depressive episode and post-traumatic stress disorder received treatment (Burns et al., 2015). Another study of Latino sexual minority men in South Florida (N = 290) found that only 20.4% received at least one mental health treatment session in the past year, despite high rates of clinically significant depressive symptoms, anxiety symptoms, trauma symptoms, and problematic substance use (Harkness et al., 2023).

Reasons for the underutilization of mental healthcare services among LSMM can be attributed to several factors. One is that many clinicians are not trained in, or do not adhere to, practice guidelines for working with sexual minority persons, which may deter LSMM from engaging in mental healthcare services (Alessi, 2013). Moreover, poor integration of mental healthcare in primary and/or specialized care systems may dissuade Latino SMM from seeking it (Funk et al., 2008). A literature review showed that somatic symptoms (e.g., headaches, stomachaches) may be perceived as more culturally appropriate compared to cognitive symptoms of anxiety or depression among Latino youth generally (Varela & Hensley-Maloney, 2009). Thus, LSMM may not seek out mental healthcare for mental health issues but rather for the somatic symptoms (Mayo et al., 2020). These findings might suggest that Latino adults, and potentially LSMM, may not present for treatment until more severe impacts are felt.

In addition to cultural influences on mental healthcare utilization, an individual’s nativity status (i.e., where they are born) may also be associated with mental healthcare utilization. However, little research has focused on the impact nativity status may have on mental healthcare utilization among LSMM specifically. One study showed less mental healthcare utilization among first generation immigrant Latino adults compared to second-generation and non-immigrant Latino adults with mood and anxiety disorders (Bauldry & Szaflarski, 2017). Recently immigrated Latino individuals may face barriers to seeking mental healthcare related to their citizenship status in the U.S. For example, Latino adults who are undocumented may be concerned with confidentiality in healthcare facilities (Sun et al., 2016), a barrier that likely extends to LSMM who are undocumented. Latino immigrants continue to be a majority of the immigrant population in the U.S. (Pew Research Center, 2020), therefore it is important to understand mental healthcare utilization behaviors of heterogeneous LSMM subgroups (e.g., based on nativity) to achieve equitable mental health outcomes.

The Cultural Attitudes Toward Healthcare and Mental Illness Questionnaire (CAHMIQ) measures cultural attitudes toward healthcare and mental illness, including beliefs about what causes mental health problems, beliefs about effective approaches for addressing mental health problems, and preferences for who to talk to about mental health problems (Jimenez et al., 2012). The CAHMIQ was developed based on the Cultural Influences on Mental Health framework which suggests that an individual’s cultural background may contribute to the development of mental health beliefs, how an individual defines mental health, and preferences an individual has for mental healthcare (Jimenez et al., 2012). Unlike other mental health attitude measures (e.g., The New Inventory of Attitudes Towards Seeking Mental Health Services [Mackenzie et al., 2004]), the CAHMIQ utilizes the Cultural Influences on Mental Health framework to examine attitudes toward mental health using a culturally syntonic lens. Despite the potential utility of the CAHMIQ for measuring LSMM’s mental healthcare beliefs, to our knowledge, the CAHMIQ has only been used among older adults (age 65+). Therefore, there is a need to assess the appropriateness and psychometric functioning of the CAHMIQ among other groups with which it could be useful, including LSMM.

Our team conducted a qualitative study to identify the determinants of LSMM’s engagement in behavioral health services and identified culturally relevant determinants that can impact LSMM’s engagement in behavioral health services (Harkness et al., 2021). Our qualitative findings suggested the utility of quantitatively measuring cultural factors that could influence LSMM’s engagement in behavioral health services. This led us to use the CAHMIQ to quantitatively assess these cultural factors and examine how they relate to LSMM’s use of behavioral health services. However, as noted above, the CAHMIQ had not previously been validated with LSMM. Validating the CAHMIQ as a measure that can be used to assess LSMM’s cultural beliefs about mental health treatment and preferences for treatment may facilitate bridging the gap between unmet mental health needs and mental health service utilization within this population, via research and clinical assessments, given the limited (and in some cases, dated) mental health research with LSMM. This could in turn assist researchers and clinicians to improve assessment of LSMM’s needs and preferences to promote positive mental health and leverage cultural strengths. As such, the current study aimed to examine the factor structure of the CAHMIQ when used with LSMM and its measurement invariance (i.e., configural, metric, and scalar) across LSMM born in and outside the continental U.S. to enhance the measure’s utility in research and clinical contexts.

Methods

Participants and Procedures

Participants included 290 LSMM in South Florida. Participants were recruited online, via social media or word of mouth for an observational longitudinal cohort study of LSMM’s engagement in HIV-prevention and behavioral health services, called the DIMELO study (Harkness et al., 2023). Full details on the recruitment process can be found elsewhere (Harkness et al., 2023). Briefly, participants were recruited from social media, listservs, community venues and events, snowball recruitment, and a “consent-to-contact” database of previous study participants in South Florida. Potential participants completed a pre-screener to determine eligibility. Eligible participants were LSMM who identified as gay, bisexual, or a man who has sex with men and were aged 18 to 60, living in the greater Miami area in Florida. An additional eligibility criterion included self-reported HIV-negative or unknown status, as the parent study was focused on HIV-prevention. If eligible, participants completed a survey lasting 60 to 75 min and were compensated $40 for their participation in the baseline assessment. Demographic information can be found in Table 1. Approximately 48.4% of participants were born in the continental U.S. and 51.6% of participants were born outside the continental U.S. including Colombia, Venezuela, Puerto Rico, and Cuba. Participants unable to provide consent or communicate in English/Spanish were excluded. Data were collected between February 18, 2020, and August 26, 2020. All study procedures were reviewed and approved by the University of Miami Institutional Review Board.

Table 1.

Participant Sociodemographic Characteristics.

Participant characteristics Overall
 Born in the continental U.S.
 Born outside the continental U.S.
N = 290 N = 140 N = 149
Age M (SD) 31.99 (8.32) 30.06 (7.98) 33.83 (8.26)
Race n (%)*
 White  230 (79.3)  112 (80.0)  118 (79.2)
 Black   14 (4.8)   10 (7.1)    3 (2.0)
 Asian    2 (0.7)    0 (0.0)    2 (1.3)
 Indigenous   11 (3.8)    6 (4.3)    5 (3.4)
 Multiracial   26 (9.0)    9 (6.4)   17 (11.4)
 Other    4 (1.4)    0 (0.0)    4 (2.7)
 Decline to answer    3 (1.0)    3 (2.1)    0 (0.0)
Language n (%)**
 English  172 (59.5)  114 (66.7)   51 (33.3)
 Spanish   70 (24.2)   14 (20.0)   56 (80.0)
 Both   47 (16.2)   11 (23.4)   36 (76.6)
Religiosity n (%)
 Not religious/spiritual  110 (37.9)   58 (41.4)   51 (34.2)
 Somewhat religious/spiritual  144 (49.7)   70 (50.0)   74 (49.7)
 Very religious/spiritual   35 (12.1)   11 (7.9)   24 (16.1)
 Decline to answer    1 (0.3)    1 (0.7)    0 (0.0)
Open in a new tab

Note. Totals vary across variables due to non-response.

*

p < .05.

**

p < .001.

Measures

The Cultural Attitudes Toward Healthcare and Mental Illness Questionnaire.

The Cultural Attitudes toward Healthcare and Mental Illness Questionnaire (CAHMIQ) was developed to measure cultural attitudes toward healthcare and mental illness. The CAHMIQ was developed by the investigators of The Primary Care Research in Substance Abuse and Mental Health for Elderly (PRISM-E) study, a group of 37 researchers across 10 different sites (Bartels et al., 2004). Each item was developed and reviewed using a consensus process amongst all investigators. The measure asks participants the following three questions to assess their attitudes toward healthcare and mental illness: (1) “Nobody knows for sure what causes mental health problems such as depression, but people have many different ideas about what the causes might be. What do you think causes depression?” (2) “If you had a mental health problem, what do you think would help you get better?” and (3) “Who would you talk to if you had a mental health problem?” The current study did not utilize the fourth item of the CAHMIQ: “Who makes most of the decisions about your health care?” due to the lack of relevance in this younger sample. Following each question, participants were presented with a checklist of potential responses to which they could respond “yes” or “no.” Question 1 included 18 yes/no responses, question 2 included 6 yes/no responses, and question 3 included 15 yes/no responses. For each question, respondents were able to select all response options that applied to them. The complete list of potential response options for each question is listed in Table 2. Full measure information can be found elsewhere (Jimenez et al., 2012). The measure was translated by bilingual team members using best practices in instrument translation: forward translation (English to Spanish), back translation (Spanish back to English), and an evaluation of the original and back-translated versions to ensure meaning was retained (Kurtines & Szapocznik, 1995).

Table 2.

CAHMIQ Items, Response Options, Endorsement, and Factor Loadings.

Second-ordered factors Item(s) First-ordered factors Response options % Endorsed Standardized factor loadings
Causes of mental health problems Item 1 Stress/loss • Loss (e.g., family, friends) 93.8 0.692
Nobody knows for sure what causes mental health problems such as depression, but people have different ideas about what the causes might be. What do you think causes mental health problems (check all that apply)? • Loss or lack of pleasurable activities 65.2 0.684
• Family issues 91.4 0.781
• Money issues 92.4 0.708
• Political stress 52.1 0.661
• Safety issues 69.7 0.717
• Stress or worry 89.0 0.713
Medical • Medical illness 78.3 0.773
• Infectious disease 60.3 0.722
• Nutritional deficiency 55.9 0.760
• Chemical imbalance 72.4 0.713
• Genetics 72.1 0.657
Spirit/psyche • Disturbance of the body, mind, and spirit 67.9 0.796
• Something you did wrong in the past 60.7 0.837
• Supernatural (e.g., witchcraft, hexes) 18.6 0.568
Environment/culture • Moving to a different place 55.2 0.869
• Cultural differences 52.1 0.866
• Adjusting to a different culture 60.3 0.941
Services/supports for mental health Item 2 Therapy • Private counseling 91.4 0.458
If you had a mental health problem, what do you think would help you get better? • Group counseling 50.0 0.504
• Social worker 25.2 0.543
• Psychologist 66.9 0.611
Item 3 Medical • Pills or medications 49.0 0.567
Who would you talk to if you had a mental health problem? • Psychiatrist 52.8 0.740
• Medical doctor 37.2 0.798
Family/friend • Romantic partner/significant other 67.2 0.722
• Family member living with you 35.5 0.546
• Family member not living with you 41.4 0.683
• LGBT friends 65.2 0.698
• Heterosexual friends 41.0 0.707
Alternative medicine provider • Alternative care provider (e.g., massage, acupuncture) 14.1 0.701
• Healera 10.7 —
• Herbal remedies 27.9 0.704
• Alternative therapies (acupuncture, massage, etc.) 54.5 0.650
• Spiritual advice 43.4 0.723
• Religious/spiritual leadera 14.5 —
• Someone from churcha 3.1 —
12-step program • 12-step programa 8.6 —
Open in a new tab
a

Response option excluded from analyses.

Data Analytic Plan

The analytic plan consisted of two steps. To examine the hypothesized second-ordered factor structure of the CAHMIQ (see Figure 1) among a sample of 290 LSMM, second-order confirmatory factor analyses (CFA) was performed (Brown, 2015). CFA is used to verify the number of underlying dimensions of the instrument (i.e., factors) and the relationship between observed measures (i.e., item responses) and factors (Brown, 2015). Model specification of the CFA was based on previous findings (Jimenez et al., 2012) and allowed for examination of the second-order factor structure of the CAHMIQ. Because all response options were binary, the hypothesized first- and second-order CFA models were estimated using the mean- and variance-adjusted weighted least squares (WLSMV) estimator (Liu et al., 2017). To account for respondents who selected more than one response option, two summed variables that represent the cumulative effects of item responses for the corresponding two second-ordered factors (i.e., mental health causes and mental health supports) were created. For example, if a participant selected family issues, cultural differences, and stress or worry as causes of mental health problems, this would be analytically accounted for by the summed variable for causes of mental health problems (question 1). The two second-ordered factors were regressed on their respective sum score variable. Therefore, all estimated parameters (i.e., item loadings and intercepts) were adjusted after accounting for the cumulative effects of respondents selecting more than one option. Model fit was examined using the Comparative Fit Index (CFI) with values >.90 indicating good fit, Tucker-Lewis Index (TLI) with values >.90 indicating good fit, and Root Mean Square Error of Approximation (RMSEA) with values <.08 indicating good fit (Hu & Bentler, 1999; Yu, 2002).

Figure 1.

Figure 1.

Open in a new tab

Factor structure of the CAHMIQ.

After identifying the factor structure of the CAHMIQ, measurement invariance was examined for the CAHMIQ between LSMM based on nativity status (i.e., born in the continental U.S. and born outside the continental U.S.) by comparing two competing invariant models with increasing constraints of parameters in the identified CFA model described in the first step (Figure 1; Vandenberg & Lance, 2000). The first invariance model, configural invariance, is the baseline model and ensures that the number of factors and the specific indicators that pattern onto each factor are the same across groups. If there is configural invariance, this implies that the overall factor structure of the CAHMIQ is equivalent across groups. The next invariance model that was tested was metric invariance which is tested against the configural invariance model to examine whether the factor loadings are similar between the two nativity groups. However, in the current study, model estimation problems occurred when conducting metric invariance. In this case, Muthén suggests that metric and scalar invariance models can be examined simultaneously (Muthen, 2013). Therefore, the next and final level of invariance was scalar invariance, which constrains item intercepts (e.g., thresholds) to be equal across groups and is tested against the configural invariance model (Meredith & Horn, 2001). If there is scalar invariance, we can infer that not only did items contribute equally to the latent constructs, but that they also had similar endorsements across nativity groups. If decrement in model fit indices (Δχ2) between the two nested (constrained) models were not significant at the 0.05 level, measurement invariance was achieved and the CAHMIQ was equivalent among nativity groups (Chen, 2007; Cheung & Rensvold, 2002; Marsh et al., 2013). There were no missing cases across items. Because surveys were offered in English and Spanish, preferred language was included as a covariate to adjust for a possible language effect. All analyses were performed using MPlus v8.3 (Muthén & Muthén, 2017) which utilized full information maximum likelihood (Enders & Bandalos, 2001).

Results

Second-Order Factor Structures

For the CFA, a two-factor model (see Figure 1) was selected based on model fit and the original structure of the CAHMIQ (Jimenez et al., 2012). This model retained the original structure of item one of the CAHMIQ, which assessed participants’ beliefs about the causes of mental health problems (e.g., cultural differences, family issues). In contrast, items 2 and 3, which focus on what (e.g., services, medication) and who (e.g., family, medical, or psychological service providers) would help address a mental health problem, were collapsed because these items were conceptually overlapping (i.e., both were about potential mental health supports). This resulted in two correlated second-order factors representing the three items of the CAHMIQ, which were made up of eight first-ordered factors representing different conceptual groupings of the manifest indicators that represented causes of mental health (from item 1) and supports for mental health (from items 2 and 3; see Figure 1). Table 2 displays the questions and responses in their respective groupings. The second-order factors were moderately and positively correlated (r = .647, p < .001), suggesting these constructs, although distinct, complement each other.

Results showed that four response options had low response variability (i.e., a healer, someone from church, or a religious/spiritual leader or attend a 12-step program) such that these response options were not frequently endorsed by participants (ranging from 3.8% to 14.1%; see Table 2), which increased model convergence problems (no standard errors of CFA model parameters). Thus, these response options were excluded from subsequent analyses. Overall, the model fit for the entire sample was acceptable χ2(582) = 694.09, p < .001; RMSEA = .026 (90% CI [0.017, 0.033]), CFI = .910, TLI = .903. This indicates that the second-order factor structure of the CAHMIQ fits the data reported by the current LSMM sample. All factor loadings across the eight constructs (subscales) ranged from .504 to .984.

Measurement Invariance by Nativity

Findings from the test of the configural invariance indicated that the model fit the data well in both groups χ2(998) = 1,065.21, p = .068; RMSEA = .022 (90% CI [0.000, 0.032]), CFI = .979, TLI = .976, suggesting that first-order latent constructs of CAHMIQ (i.e., stress/loss, medical, spirit/psyche, environment/culture, therapy, medical, family/friend, and alternative medicine provider) are similar across the two nativity groups and that the factor structure is supported in both samples. See Table 3 for a summary of fit statistics for each level of invariance testing. A χ2 difference test was used to compare the fit of the more constrained scalar invariance model with that of the configural invariance model. Findings (Δχ2(18) = 16.941, p = .527) indicated that the fit of the scalar invariance model was not significantly worse than the configural invariance model. This suggests that participants had similar factor structures and similar endorsement patterns of each item in CAHMIQ regardless of nativity.

Table 3.

Summary of Fit Statistics for Tests of Measurement Invariance.

Model χ2 df RMSEA CFI TLI Δχ2 Δdf p
Configural (unconstrained model) 1,065.213  998 0.022 0.979 0.979
Scalar (constrained loadings + thresholds) 1,018.674 1,016 0.021 0.980 0.977 16.941 18 .5272
Open in a new tab

Discussion

The current study aimed to examine the psychometric performance of the CAHMIQ among LSMM in South Florida, as this measure has not previously been validated psychometrically in this population but has shown to be associated with LSMM’s engagement in behavioral health services in prior work (Harkness et al., 2023) and could potentially be useful in other research and clinical contexts with LSMM. Further, this study helps to meet a need for culturally relevant validated measures of cultural attitudes toward mental health to assess for and in turn, eliminate mental health related disparities among LSMM. We found evidence for two distinct factors that assess participants’ beliefs about causes of mental health problems and mental health services/supports that could address mental health problems, both of which are grounded in cultural specificity (e.g., values and beliefs).

Study results also suggest that the CAHMIQ performs equivalently across LSMM with different nativity statuses, which is important for determining the appropriateness of using this measure and making comparisons of scores across these important subgroups of LSMM. These findings provide psychometric support for the CAHMIQ and indicate that it can be used among LSMM across nativity statuses to measure cultural influences on mental health and treatment.

Although the results of the confirmatory factor analysis supported a model with two second-order factors representing causes of mental health problems and supports for mental health and eight first-ordered factors representing different conceptual groupings of causes of mental health and supports for mental health (Figure 1), there were four response options that were excluded from the analyses due to low response variability which increased model convergence problems. The four response options removed were related to supports for mental health and suggest that LSMM were unlikely to talk to a healer, someone from church, or a religious/spiritual leader or attend a 12-step program to address their mental health concerns. To contextualize this finding, among the entire sample of LSMM, 37.9% reported they were not religious/spiritual, 49.7% were somewhat religious/spiritual, and 12.1% were very religious/spiritual (see Table I). Given that more than a third of the sample was not religious or spiritual, it is not surprising that they were also unlikely to seek mental health support from a healer, someone from church, or a religious/ spiritual leader. Some LSMM who attend religious services report concealing their sexual orientation to individuals affiliated with their places of worship due to these religious environments being non-affirming (Wright & Stern, 2016), which could further explain why LSMM did not seek mental health support in these contexts. In fact, some minority stress experiences contributing to LSMM’s mental health disparities may be attributable to non-affirming religious and spiritual contexts (Barnes & Meyer, 2012; Wright & Stern, 2016). Our finding that these religious/spiritual response options on the CAHMIQ did not fit within the factor structure of the measure for LSMM underscores the importance of our analysis in identifying the most appropriate and relevant way to use this measure with a population that is distinct from the older adult population for which the measure was initially developed.

Due to the binary nature of the data, metric invariance was not examined and, therefore, invariance of factor loadings across nativity status was not assessed. However, with the findings related to configural and scalar invariance, it can be inferred that the same basic factor structure can be applied across LSMM born in and outside the continental U.S. and that these groups can be meaningfully compared on latent means. In other words, when using the CAHMIQ among LSMM with different nativity statuses, the same scoring guidelines may be used.

Implications

Given mental health disparities and stigma surrounding mental health in the Latino community (Fripp & Carlson, 2017) and in particular for LSMM (Breslau et al., 2017), the findings of this study have important implications for efforts to increase mental health treatment uptake, and in turn eliminate mental health disparities, among LSMM. The CAHMIQ may be used in primary care or behavioral health settings in conjunction with standard mental health (e.g., depression) screenings. Using the measure in screening contexts could inform appropriate referrals and best practices for engaging LSMM in mental health treatment by identifying what services/supports would be acceptable and appropriate to the individual (e.g., peer support, psychologist, medication) and facilitate culturally congruent and strengths-based case conceptualizations.

Additionally, the CAHMIQ can be used in research settings to identify potential areas of focus within mental health treatment/prevention to ensure that treatments being developed and tested account for LSMM’s culturally embedded understandings of what drives mental health concerns. For example, some treatments may focus more on social/environmental causes of distress, whereas others may have a more biological underpinning of explaining causes of mental health. The CAHMIQ can inform (a) which of these different types of treatments to refer LSMM to and (b) the development and testing of strengths-based treatments that are aligned with the view of mental health that is consistent with LSMM’s beliefs as measured by the CAHMIQ. For example, if the CAHMIQ revealed that most LSMM in a given setting believe that adjusting to a new culture is one of the main causes of mental health problems, a tailored outreach campaign could focus on linking LSMM to mental health services that focus on adjusting to a new culture and highlight this as a potential stressor. As another example, if the CAHMIQ revealed that family and friend support was their preferred mode of support for a mental health concern, this could inform the development and testing of a treatment that leverages cultural strengths around family and relational support.

In addition to identifying potential areas of focus within mental health treatment/prevention, results provide insight as to what areas of focus are not applicable and/or relevant to LSMM. Study findings suggest that LSMM are unlikely to seek mental health support from a healer, someone from church, or a religious/spiritual leader, therefore these response options were not included in the final measure. This finding highlights the importance of adapting measures to the unique experiences of different populations. It may be that adapted versions of the CAHMIQ are needed to ensure its cultural relevance to different populations.

Our findings can also inform standardized scoring guidelines when using the CAHMIQ with LSMM. The CAHMIQ, when used with LSMM, should use a two-factor scoring procedure with separate subscales for causes of mental health problems and services/supports for mental health. Depending on the research question(s), the measure may be scored as binary for each subscale (i.e., “yes” if the participant endorsed any response option on the subscale or “no” if the participant did not endorse any response option on the subscale). For example, if someone responded affirmatively to any response option on the environment/culture subscale, this subscale would be scored “yes.” Alternatively, the measure could be scored as an additive score within each subscale (i.e., summing the total number of response options endorsed for each subscale).

Limitations

Our findings should be interpreted while considering the following limitations. First, findings are limited by the demographics of the participants in the current study. This study did not include women or non-binary individuals, nor did it include LSMM who were living with HIV due to the focus on HIV-prevention in the parent study. Further, participants in this study were from one geographic area in South Florida and this study does not account for the full heterogeneity of Latino communities. One analytic decision we made was deciding to examine nativity as two groups: those born in the continental U.S. compared to those born outside the continental U.S. based on prior research showing that Puerto Rican born individuals—who are not foreign born—are more culturally aligned with cultural values/beliefs of neighboring countries in the Caribbean (e.g., Cuba) than cultural values/beliefs of the U.S. Although we felt that this was the best analytic decision for these data, we also recognize that an alternative approach could have been to compare U.S. and foreign born LSMM, which may have yielded different information. Similarly, although there was measurement invariance across nativity status, participants’ specific country of origin was not examined for measure invariance, which may have provided information regarding specific cultural influences/attitudes by country of origin. It is unclear how the length of stay in the U.S. may have contributed to the study findings, however, most participants born outside the continental U.S. had lived in the U.S. for more than 10 years. Moreover, we did not measure nor evaluate acculturation in the analysis and how it may impact cultural influences/attitudes toward mental health. Further, although we followed best practices for translation to Spanish, we did not evaluate content validity prior to administering the Spanish translations. Future research with this measure could evaluate the content validity of the Spanish CAHMIQ translation using strategies such as cognitive interviewing.

This study contributes to the psychometric literature for LSMM and, to our knowledge, provides the first data related to measurement invariance of the CAHMIQ among LSMM of different nativity statuses. Establishing factor structure and measurement invariance can inform the use of the CAHMIQ with LSMM, as well as help researchers and clinicians by providing standardized scoring guidelines for this measure when used with LSMM. Continued use of the CAHMIQ with LSMM will advance research related to why mental health services are underutilized among LSMM and how to increase the uptake of services, and in turn, equitable mental health outcomes.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Center for AIDS Research (CFAR) at the University of Miami under Award Number P30AI073961 (Pawha), the National Institute on Minority Health and Health Disparities under Award Number U54MD002266 (Behar-Zusman) and K23MD015690 (Harkness), and the National Institute of Mental Health under Award Number P30MH133399 (Safren).

Biographies

Alyssa Lozano is a Research Assistant Professor at the University of Miami School of Nursing and Health Studies. Her research interests include addressing inequities in substance use and mental health outcomes among Hispanic sexual minority youth.

Tae Kyoung Lee is an Assistant Professor at Sungkyunkwan University. His research focuses on (a) disparities in adolescent health and development, (b) minority health, and (c) the effectiveness of intervention programs.

Alejandra Fernandez is an Assistant Professor at the Peter O’Donnell School of Public Health at UT Southwestern Medical Center. Her research focuses on the influence of family functioning behaviors on Hispanic adolescent health behaviors, including substance use, sexual risk behaviors, mental health outcomes, and obesity.

Guillermo Prado is a Professor in Nursing and Health Studies, Public Health Sciences, and Psychology at the University of Miami. His research focuses on developing, evaluating, and translating preventive interventions for addressing smoking, alcohol, drug abuse, HIV, and obesity health disparities among Hispanic youth.

Steven A. Safren is a Professor of Psychology at the University of Miami. His research focuses on health behavior change, with a particular emphasis on mental health and substance use components of HIV prevention and treatment domestically and internationally.

Daniel E. Jimenez is an Associate Professor at the University of Miami Miller School of Medicine. His research interests include geriatric mental health services research, health promotion, multicultural mental health, and mental illness prevention.

Audrey Harkness is an Assistant Professor at the University of Miami School of Nursing and Health Studies. Her program of research is focused on developing, adapting, and evaluating interventions and implementation strategies to achieve health equity among key populations affected by HIV and mental health disparities.

Footnotes

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Dr. Safren receives royalties from Oxford University Press, Gilford Publications, and Springer/Humana Press for books on cognitive behavioral therapy. The other authors have no relevant financial or non-financial interested to disclose.

1.

Consistent with guidance from the National Institutes of Health, the term “sexual minority” men (SMM) is used when referring to participants in the current study as well as other studies in which all participants identified as gay, bisexual, or another sexual minority identity (e.g., pansexual).

References

  1. Alessi EJ (2013). Acknowledging the impact of social forces on sexual minority clients: Introduction to the special issue on clinical practice with LGBTQ populations. Clinical Social Work Journal, 41(3), 223–227. 10.1007/s10615-013-0458-x [DOI] [Google Scholar]
  2. Barnes DM, & Meyer IH (2012). Religious affiliation, internalized homophobia, and mental health in lesbians, gay men, and bisexuals. American Journal of Orthopsychiatry, 82(4), 505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bartels SJ, Coakley EH, Zubritsky C, Ware JH, Miles KM, Areán PA, Chen H, Oslin DW, Llorente MD, Costantino G, Quijano L, McIntyre JS, Linkins KW, Oxman TE, Maxwell J, Levkoff SE, & PRISM-E Investigators. (2004). Improving access to geriatric mental health services: A randomized trial comparing treatment engagement with integrated versus enhanced referral care for depression, anxiety, and at-risk alcohol use. The American Journal of Psychiatry, 161(8), 1455–1462. 10.1176/appi.ajp.161.8.1455 [DOI] [PubMed] [Google Scholar]
  4. Bauldry S, & Szaflarski M (2017). Immigrant-based disparities in mental health care utilization. Socius, 3, 2378023116685718. [Google Scholar]
  5. Breslau J, Cefalu M, Wong EC, Burnam MA, Hunter GP, Florez KR, & Collins RL (2017). Racial/ethnic differences in perception of need for mental health treatment in a US national sample. Social Psychiatry and Psychiatric Epidemiology, 52(8), 929–937. 10.1007/s00127-017-1400-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brown TA (2015). Confirmatory factor analysis for applied research. Guilford Publications. [Google Scholar]
  7. Burns MN, Ryan DT, Garofalo R, Newcomb ME, & Mustanski B (2015). Mental health disorders in young urban sexual minority men. Journal of Adolescent Health, 56(1), 52–58. 10.1016/j.jadohealth.2014.07.018 [DOI] [Google Scholar]
  8. Chen FF (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 14(3), 464–504. [Google Scholar]
  9. Cheung GW, & Rensvold RB (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 9(2), 233–255. [Google Scholar]
  10. Cohen JM, Blasey C, Barr Taylor C, Weiss BJ, & Newman MG (2016). Anxiety and related disorders and concealment in sexual minority young adults. Behavior Therapy, 47(1), 91–101. 10.1016/j.beth.2015.09.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Crenshaw K (1991). Mapping the margins: Intersectionality, identity politics, and violence against women of color. Stanford Law Review, 43(6), 1241–1299. 10.2307/1229039 [DOI] [Google Scholar]
  12. Enders CK, & Bandalos DL (2001). The relative performance of full information maximum likelihood estimation for missing data in structural equation models. Structural Equation Modeling: A Multidisciplinary Journal, 8(3), 430–457. [Google Scholar]
  13. Fripp JA, & Carlson RG (2017). Exploring the influence of attitude and stigma on participation of African American and Latino populations in mental health services. Journal of Multicultural Counseling and Development, 45(2), 80–94. [Google Scholar]
  14. Funk M, Saraceno B, Drew N, & Faydi E (2008). Integrating mental health into primary healthcare. Mental Health in Family Medicine, 5(1), 5–8. [PMC free article] [PubMed] [Google Scholar]
  15. Harkness A, Lozano A, Bainter S, Mayo D, Hernandez Altamirano D, Rogers BG, Prado G, & Safren SA (2023). Engaging Latino sexual minority men in PrEP and behavioral health care: Multilevel barriers, facilitators, and potential implementation strategies. Journal of Behavioral Medicine, 46(4), 655–667. 10.1007/s10865-022-00371-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Harkness A, Satyanarayana S, Mayo D, Smith-Alvarez R, Rogers BG, Prado G, & Safren SA (2021). Scaling up and out HIV-prevention and behavioral health services to Latino sexual minority men in South Florida: Multilevel implementation barriers, facilitators, and strategies. AIDS Patient Care & STDs, 35(5), 167–179. 10.1089/apc.2021.0018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hu L, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  18. Jimenez DE, Bartels SJ, Cardenas V, Dhaliwal SS, & Alegría M (2012). Cultural beliefs and mental health treatment preferences of ethnically diverse older adult consumers in primary care. The American Journal of Geriatric Psychiatry, 20(6), 533–542. 10.1097/JGP.0b013e318227f876 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kurtines WM, & Szapocznik J (1995). Cultural competence in assessing Hispanic youths and families: Challenges in the assessment of treatment needs and treatment evaluation for Hispanic drug-abusing adolescents. NIDA Research Monograph, 156, 172–189. [PubMed] [Google Scholar]
  20. Liu Y, Millsap RE, West SG, Tein J-Y, Tanaka R, & Grimm KJ (2017). Testing measurement invariance in longitudinal data with ordered-categorical measures. Psychological Methods, 22(3), 486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Mackenzie CS, Knox VJ, Gekoski WL, & Macaulay HL (2004). An adaptation and extension of the Attitudes Toward Seeking Professional Psychological Help Scale. Journal of Applied Social Psychology, 34(11), 2410–2433. 10.1111/j.1559-1816.2004.tb01984.x [DOI] [Google Scholar]
  22. Marsh HW, Nagengast B, & Morin AJ (2013). Measurement invariance of big-five factors over the life span: ESEM tests of gender, age, plasticity, maturity, and la dolce vita effects. Developmental Psychology, 49(6), 1194. [DOI] [PubMed] [Google Scholar]
  23. Mayo D, Hernandez Altamirano D, Chan B, Safren S, & Harkness A (2020). Childhood adversity and adult somatic symptoms among Latinx sexual minority men: The moderating role of stigma-related mental health knowledge [Poster presentation]. Association for Behavioral and Cognitive Therapies 54th Annual Convention, Sexual and Gender Minority Special Interest Group Poster Session, Philadelphia, PA, United States. [Google Scholar]
  24. Meredith W, & Horn J (2001). The role of factorial invariance in modeling growth and change. In Collins LM & Sayer AG (Eds.), New methods for the analysis of change (pp. 203–240). American Psychological Association. [Google Scholar]
  25. Meyer IH (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychological Bulletin, 129(5), 674–697. 10.1037/0033-2909.129.5.674 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Muthen B (2013). Measurement invariance with multigroups. http://www.stat-model.com/discussion/messages/9/11980.html?1530051917 [Google Scholar]
  27. Muthén LK, & Muthén B (2017). Mplus user’s guide: Statistical analysis with latent variables, user’s guide. Muthén & Muthén. [Google Scholar]
  28. Pew Research Center. (2020, August 20). Immigrants in America: Key charts and facts. Pew Research Center’s Hispanic Trends Project. https://www.pewresearch.org/hispanic/2020/08/20/facts-on-u-s-immigrants/ [Google Scholar]
  29. Rhodes SD, Martinez O, Song E-Y, Daniel J, Alonzo J, Eng E, Duck S, Downs M, Bloom FR, Allen AB, Miller C, & Reboussin B (2013). Depressive symptoms among immigrant Latino sexual minorities. American Journal of Health Behavior, 37(3), 404–413. 10.5993/AJHB.37.3.13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Rodriguez-Seijas C, Eaton NR, & Pachankis JE (2019). Prevalence of psychiatric disorders at the intersection of race and sexual orientation: Results from the National Epidemiologic Survey of Alcohol and Related Conditions-III. Journal of Consulting and Clinical Psychology, 87(4), 321–331. 10.1037/ccp0000377 [DOI] [PubMed] [Google Scholar]
  31. Sun CJ, Ma A, Tanner AE, Mann L, Reboussin BA, Garcia M, Alonzo J, & Rhodes SD (2016). Depressive symptoms among Latino sexual minority men and Latina transgender women in a new settlement state: The role of perceived discrimination. Depression Research and Treatment, 2016(1), 4972854. 10.1155/2016/4972854 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Vandenberg RJ, & Lance CE (2000). A review and synthesis of the measurement invariance literature: Suggestions, practices, and recommendations for organizational research. Organizational Research Methods, 3(1), 4–70. [Google Scholar]
  33. Varela RE, & Hensley-Maloney L (2009). The influence of culture on anxiety in Latino youth: A review. Clinical Child and Family Psychology Review, 12(3), 217–233. 10.1007/s10567-009-0044-5 [DOI] [PubMed] [Google Scholar]
  34. Wright AJ, & Stern S (2016). The role of spirituality in sexual minority identity. Psychology of Sexual Orientation and Gender Diversity, 3(1), 71. [Google Scholar]
  35. Yu C-Y (2002). Evaluating cutoff criteria of model fit indices for latent variable models with binary and continuous outcomes. University of California, Los Angeles. [Google Scholar]

RESOURCES