Childhood sexual abuse and depression in Latino men who have sex with men: Does resilience protect against nonadherence to antiretroviral therapy?

Abstract

This study tested depression as a mediator between childhood sexual abuse and adherence to antiretroviral therapy, an effect moderated by resilience. In total, 149 HIV+ Latino men who have sex with men were recruited for this study. Using a regression-based bootstrap approach, depression mediated the relationship between childhood sexual abuse and antiretroviral therapy adherence, with worse adherence for participants at lowest percentiles of the resilience index. The prevalence of childhood sexual abuse and depression in HIV+ men who have sex with men is high and must be addressed to better prevent disease progression and reduce transmission, especially in expanding Latino populations.

Keywords

adherence childhood sexual abuse HIV and AIDS Latino/Hispanic resilience

Introduction

Along the continuum of HIV care, the clinical outcomes for Latinos (especially men who have sex with men (MSM)) continue to lag behind those of their non-Hispanic White counterparts, from diagnosis, to retention in care, to prescribed antiretroviral therapy (ART) and achievement of viral suppression (Centers for Disease Control and Prevention (CDC), 2012; Henry J. Kaiser Family Foundation, 2013). Optimizing health outcomes through sustained medication adherence would not only slow the progression from HIV to AIDS but also reduce transmission, risk of the development of drug resistance, and other health complications that decrease survival (Braithwaite et al., 2005; Cohen et al., 2012; Friedland and Williams, 1999; Sethi et al., 2003).

Abuse, mental health, and adherence

Two mechanisms to explain disproportionate HIV incidence rates and poorer HIV treatment outcomes in Latinos, specifically MSM, are a history of childhood sexual abuse (CSA; Mimiaga et al., 2009) and mental health problems (Stall et al., 2003). First, irrespective of the HIV status, the prevalence of CSA in MSM is high (Rothman et al., 2011). Data suggest that ethnic minority MSM, compared to White MSM, are twice as likely to experience CSA (prevalence estimates ranging from 15.8% to 39%) and younger when the abuse occurs (Arreola et al., 2005, 2009; Doll et al., 1992; Jinich et al., 1998). Meta-analytic and prospective studies have documented an increased risk of physical and mental health problems subsequent to CSA (Hillberg et al., 2011; Widom et al., 2012), but how CSA affects treatment outcomes and HIV disease management is yet to be determined (Maniglio, 2010; O’Cleirigh et al., 2012; Paolucci et al., 2001).

Second, depression affects a large proportion of people living with HIV and AIDS (Bing et al., 2001) and impacts ART adherence (Gonzalez et al., 2011). Longitudinal studies before and after the advent of ART have clearly shown depression and psychological stress to be significant predictors of AIDS progression and mortality (Leserman, 2008a; Page-Shafer et al., 1996; Patterson et al., 1996). Assuming that access to high-quality care is available, AIDS prevention largely depends on adherence levels. However, disparities in adherence rates are present in Latinos receiving care in the United States (Simoni et al., 2012).

Understanding how mediating and moderating psychosocial factors influence health is necessary to advance health disparities research (Mackinnon and Luecken, 2008). For example, CSA is a well-known nonspecific risk factor for mental health problems and risk-taking behaviors in MSM who are HIV− (Mimiaga et al., 2009), but exactly how and when CSA affects HIV treatment outcomes are not clear.

In light of limited empirical evidence for the long-term effect of CSA on HIV-related health outcomes, a plausible direct pathway is through depression (Ferrando and Freyberg, 2008). Depression produces changes in concentration, energy levels, and overall mood that directly impact HIV health-care behaviors (Catz et al., 1999; Safren et al., 2002). However, the question remains as to for whom depression impairs adherence. Only one intervention has targeted depression and adherence issues in Latinos living with HIV (Simoni et al., 2013). Investigating the mechanisms responsible for successful disease outcomes, despite the presence of CSA, may have public health implications as it relates to HIV treatment and prevention.

Resilience

Only a minority of individuals who experience a traumatic life event (e.g. CSA) later develop a mental illness (Boudewyn and Liem, 1995; Felitti et al., 1998; Ozer et al., 2008). Thus, successful adaptation (i.e. “resilience”) to a stressor occurs with high frequency. The full biopsychosocial model is implicated in resilience research, suggesting that “bouncing back” from childhood trauma or other stressors requires adequate prosocial and interpersonal competencies, personality traits such as hardiness and optimism, and positive/adaptive coping responses (Bonanno, 2004; Heller et al., 1999). More importantly, understanding resilience in the context of CSA breaks the tradition of deficit-focused models when studying issues related to vulnerable and disadvantaged populations. As described above, MSM populations disproportionately experience greater numbers of traumatic life events, such as episodes of CSA and HIV infection. What is not clear is how resilience may be linked to better treatment outcomes following an episode of CSA.

Not all HIV+ Latino MSM with a history of CSA have poor treatment outcomes. Investigating how and when CSA affects HIV health outcomes may inform treatment and adherence interventions. In this study, we first hypothesized depression to be the indirect pathway between CSA and adherence. Second, we hypothesized this indirect effect (i.e. mediation) to be moderated by resilience (see Figure 1(a)).

Figure 1.

(a) Hypothesized conceptual conditional process model for childhood sexual abuse and adherence. Indirect effect of M between X and Y at levels of Z. (b) Indirect pathway through depression to adherence levels (k = 10,000 bootstrap samples). Indirect effect of M between X and Y at levels of Z = (a₁ + a₃Z) × b₁. See Table 1 for results including covariates not shown here. Dashed lines represent estimates (a₁, a₃, and b₁) used in the test of statistically significant indirect effects.

Methods

Participants

Data were collected from April 2012 through March 2013 in El Paso, TX, USA, at Centro de Salud Familiar La Fe CARE Center, a federally qualified health center that offers comprehensive HIV and AIDS services to predominately Mexican-American individuals living on the US–Mexico border. Clinic staff identified participants based on the following inclusion criteria: (1) biologically male, (2) HIV+, (3) currently or historically engaged in sex with men, (4) at least 18 years of age, and (5) Latino/Hispanic according to self-report.

The final sample of 149 participants averaged 42.2 years of age (standard deviation (SD) = 12.2 years; median = 43 years) and had been living with HIV for an average of 9.5 years (SD = 7.5 years; median = 8 years). Their median household annual income was US$12,000 (semi-interquartile range = US$6400). Of the total participants, 54 completed the survey in Spanish and 95 in English.

Procedure

Once participants were identified and agreed to be contacted, they were referred to researchers, who then confirmed the eligibility of all recruited patients and scheduled an appointment to complete a 2-hour interview. All interviews were conducted in either English or Spanish (based on participant preference) with a bilingual researcher in a private clinic office. During the interview, consent for participation and extraction of medical records was obtained, and both an open-ended interview and paper-and-pencil survey were completed. After the interview, participants were reimbursed US$30 for their time and transportation costs.

All English-language survey materials were translated to the regional Spanish dialect by a certified translator. The materials were then back-translated into English by a second certified translator, and all discrepancies were resolved during review meetings with both translators and bilingual research staff who were from the US–Mexico border region. The survey was part of a larger longitudinal parent study. Administrative approval was obtained from La Fe CARE Center, and the University of Texas at El Paso Institutional Review Board approved all materials and procedures.

Measures

CSA

We utilized the conceptual and operational definition of CSA provided by two large studies documenting the prevalence rates of CSA in Latino MSM (Arreola et al., 2008, 2009). Based on sexual experiences prior to the age of 16 years, participants reported a history of (1) no sex, (2) consensual sex, or (3) forced sex. Those reporting no sex or consensual sex (i.e. not forced) were collapsed into a single category and given a score of 0. The forced sex group self-reported having at least one instance of being “forced or frightened by someone into doing something sexually that you did not want to do” and was given a score of 1. Data have shown that a history of nonconsensual sex has been associated with greater HIV-related fatigue (Leserman et al., 2008b), nonadherence to ART (Leserman et al., 2008c), and faster progression to AIDS (Leserman et al., 2008a)

Depressive symptoms

Depressive symptoms were assessed with the Patient Health Questionnaire-9 (PHQ-9), a measure shown to be psychometrically sound in Latinos (Wiebe et al., 2013). The PHQ-9 instructions specify a 2-week reporting period, with each item using a 4-point response scale from “not at all” to “nearly every day.”

Resilience

Resilience was assessed by the 25-item Connor–Davidson Resilience Scale (CD-RISC). The CD-RISC defines resilience as “personal qualities that enable one to thrive in the face of adversity” (Connor and Davidson, 2003: 76). The original CD-RISC assessed the psychometric properties in six samples, including a large general population (non-help-seeking) sample, a psychiatric outpatient sample, as well as participants enrolled in a randomized controlled trial (RCT) for the treatment of post-traumatic stress disorder (PTSD). Internal consistency (α = .89) and test–retest reliability (between two consecutive medical visits; intraclass correlation (ICC) = .87) estimates in all samples were good.

Adherence to ART

Past 30-day adherence to ART through the visual analog scale (VAS) has shown to be one of the most accurate time frames and measures for self-reported adherence (Amico et al., 2006; Giordano et al., 2004; Lu et al., 2007; Oyugi et al., 2004). The VAS is a 10-cm line on which participants indicate the percentage of doses of all HIV medications taken in the past 30 days. In order to obtain more accurate responses, a short paragraph that normalizes adherence difficulties prefaced the administration of the VAS (Chesney et al., 2000).

Statistical analyses

All statistical analyses were conducted with IBM SPSS V.19 statistical software (IBM Corp., 2012). We used a moderated mediation regression approach (also known as a conditional process model; Hayes, 2013) with 10,000 bootstrap samples (k) for the assessment of indirect effects for continuous dependent variables. The specific SPSS program employs heteroskedastic-consistent standard errors (SEs) for our outcome variable to correct for any violations of distributional and error variance assumptions. This correction allows us to preserve the natural units of measurement and interpretation of results.

Although a test of an “indirect effect” is often considered synonymous with a “mediation” test, this is not entirely accurate as it depends on the nature of the hypothesis (for a complete discussion, see Mathieu and Taylor (2006)). No direct effect is required (see Hayes, 2009) as we predicted CSA to ultimately transmit an effect on adherence through depression, with the effect becoming weaker as scores on the resilience index increased (see Morera and Castro, 2013).

This specific test (i.e. indirect effect of X on Y through M [a × b] at levels of Z is solved by [a_1j + a₃_jZ] × b₁) is now prioritized when testing distal indirect relationships as it estimates the indirect effect directly, rather than assuming its presence through a series of regression analyses (i.e. the causal steps approach; see Morera and Castro, 2013). In Figure 1(a), the indirect effect (M) at levels of Z (i.e. resilience) is solved by (a_1j + a₃_jZ) × b₁. The value of Z represents CD-RISC scores ranging from the 10th percentile to the 90th percentile, which are derived from the creation of bootstrap confidence intervals. Additionally, years living with HIV, age at the time of interview, language of interview, and annual household income were modeled as covariates of both depression and adherence to demonstrate the unique effects of our primary variables in the model.

Results

Descriptive data

The average resilience index score was 96.0 (SD = 21.29), with a possible minimum and maximum score ranging from 25 to 125 on the CD-RISC (the pattern of the moderated indirect effect is shown across percentiles of CD-RISC scores).

A total of 66.4 percent (n = 99) of participants reported at least one sexual experience prior to the age of 16 years, with 34 percent of those (n = 34) reporting the sexual experience as being nonconsensual. Overall, 22.8 percent of the sample, over one in five, reported an episode of nonconsensual sex.

The average depression score for the no sex/consensual sex group fell into the “mild” category (criterion range from 5 to 9, M = 5.9, SD = 6.2). The average depression score for the nonconsensual sex group approached the “moderate” category (M = 9.8, SD = 8.2, “moderate” cutoff score = 10).

Nine participants were not on ART. Those on ART (n = 140) self-reported taking 91.6 percent of all HIV medications in the past month. CD4 cell count (available for only 129 participants) averaged 545 cells/mL³ of blood (SD = 279; CD4 cell counts in nonclinical samples typically range from 700 to 900). Viral Load (VL) data (available for only 124 participants) indicated that 74.2 percent were virally suppressed at most recent blood draw.

Moderated mediation model

Preliminary bivariate analyses supported the hypothesized associations between the PHQ-9 and CD-RISC scores with adherence by the VAS (rs = −.21 and .17, ps < .05), as well as PHQ-9 and CD-RISC with each other (r = −.34, p < .01). Figure 1(b) shows the primary model estimates. Table 1 shows all model estimates.

Table 1.

Conditional process model results by outcome variable.

Unstandardized regression coefficients (SE)
Outcome: past 30-day adherence
Predictors
1. Depression (b₁ path)	−.62 (.30)*
2. Childhood sexual abuse (c path)	−1.40 (3.64)
3. Years living with HIV (covariate)	−.67 (.26)*
4. Age (covariate)	.24 (.11)*
5. Language (covariate)	−2.75 (2.26)
6. Annual income (covariate)	.01 (.01)
Outcome: depression
Predictors
1. Childhood sexual abuse (a₁ path)	24.77 (5.72)**
2. CD-RISC	−.06 (.02)**
3. Childhood sexual abuse × CD-RISC (a₃ path)	−.21 (.05)**
4. Years living with HIV (covariate)	.03 (.08)
5. Age (covariate)	−.08 (.05)
6. Language (covariate)	3.22 (1.06)**
7. Annual income (covariate)	−.01 (.01)

CD-RISC: Connor–Davidson Resilience Scale; SE: standard error.

Partially standardized effect size estimate for a₁ × b₁ = −.84.

p < .05; **p < .01.

In the first stage of conditional process model (see Figure 1(b)), CSA was associated with greater self-reported depression symptoms (a₁ = 24.77, SE = 5.72, p < .01). Our index of resilience as a simple effect (a₂) and the interaction with CSA (a₃) were statistically significant in predicting self-reported depression symptoms (a₃ = −.21, SE = .05, p < .01), which was then negatively associated with adherence (b₁ = −.62, SE = .30, p < .05). Covariates are listed in Table 1.

Table 2 shows the indirect effect of depression on diminishing adherence as scores on the resilience index increase. For example, the effect of depression on adherence was strongest for participants at the lowest percentile of the resilience index (10th percentile score = 67, indirect effect = −6.77, SE = 3.44, p < .05). The moderated indirect effect remained statistically significant up to the 50th percentile score (CD-RISC score = 97, indirect effect = −2.96, SE = 1.71, p < .05), that is, no differences in adherence through depression (with or without a history of CSA) resulted if participants scored above the 50th percentile on the CD-RISC measure.

Table 2.

Moderated indirect effects: childhood sexual abuse impacts adherence through depression at percentile scores of resilience.

CD-RISC percentiles	CD-RISC scores (min = 25, max = 125)	Indirect effects of depression on adherence	SE	95% CIs
10th	67	−6.77	3.44	(−15.22 to −1.31)
25th	85	−4.48	2.36	(−10.41 to −.82)
50th	97	−2.96	1.71	(−7.55 to −.54)
75th	114	−.81	1.15	(−4.39 to .51)
90th	120	−.05	1.15	(−3.12 to 1.76)

CD-RISC: Connor–Davidson Resilience Scale scores; SE: standard error; CI: confidence interval.

Unstandardized indirect effect estimates represent mean difference in adherence by predictor variable (1 = childhood sexual abuse, 0 = no abuse) through depression at percentiles of CD-RISC scores. Bold font indicates statistically significant indirect effects.

The partially standardized indirect effect was −.84, interpreted as the number of SDs that adherence decreased for every change in depression of size a₁ (see Figure 1(b); MacKinnon, 2008; Preacher and Kelley, 2011). All regression coefficients otherwise are reported as unstandardized estimates as they represent the mean difference in adherence between the CSA groups (i.e. 1 = forced sex group, 0 = no sex/consensual sex group). For example, the indirect effect of depression on adherence equal to −6.77 (at the 10th percentile) represents a nearly 7 percent point difference in all ART medications taken in the past month between those with and without a history of CSA.

Discussion

In this cross-sectional study of HIV+ Latino MSM, we provide evidence for one pathway that links CSA to current ART adherence levels utilizing a statistical approach that simultaneously tests mediators and moderators. Specifically, we found that individuals with a history of CSA report greater depressive symptoms and worse adherence, especially when not well equipped with stress-coping abilities (i.e. not resilient).

On a positive note, it is encouraging to find that CSA, depression, and their association with adherence diminished as stress-coping ability increased (i.e. an essential part of resilience). We utilized an index of resilience provided by Connor and Davidson (2003), which represents skills present in individuals that allow them to successfully deal with adversity. Utilizing this conceptualization, individuals who successfully adapted to a stressor (e.g. CSA) appear to be at nonsignificantly greater risk for nonadherence to ART.

Typically, researchers conceptualize resilience as an outcome (i.e. success or failure), as opposed to an individual trait, achieved through a “dynamic process” of successfully dealing with a stressor (Luthar et al., 2000; Masten, 2001). This implies an innate or learned set of coping skills that protect an individual from the mental sequelae that are comorbid with CSA. However, conceptualizations of resilience in HIV are variable. Some articulations of resilience include successful disease management globally (Van Eeden-Moorefield, 2008), engagement in advocacy and prevention efforts to decrease stigma (Herrick et al., 2011), and adoption of serosorting sexual behaviors (i.e. selecting partners with concordant HIV status), which reduce new infections, experiences of stigma, and risk of criminalization due to transmission (Kurtz et al., 2012). Moving forward, future studies should aim to understand how to best learn from these resilient individuals to successfully manage HIV.

Outside of HIV research, conceptualizations of resilience are multileveled. These levels include biological and genetic factors (Cicchetti, 2010), individual traits (e.g. hardiness), “systems” such as the protective effect of family and community cohesion (Hermann et al., 2011), and context. For example, resilience following CSA would be less likely within the context of a poverty- and crime-ridden environment (Williams and Nelson-Gardell, 2012). However, to date, existing interventions to enhance resilience are typically geared toward an individual’s interpretations of stressors and regulation of negative affect (Kent et al., 2011; Loprinzi et al., 2011).

Implications

MSM are disproportionately represented in the HIV epidemic in certain regions of the world and suffer high rates of CSA. Interventions designed for HIV+ MSM who have suffered severe trauma or stress are necessary for making large steps toward reducing AIDS cases (i.e. preventing disease progression) and reducing the number of new infections indirectly. From a resilience perspective, one might instead examine MSM who, despite a history of CSA, are adherent to ART and achieve viral suppression. This approach will allow resilient individuals to reveal what allows them to manage HIV successfully, despite a history of trauma.

Limitations

The findings presented in this study must be interpreted in light of some methodological limitations. First, the cross-sectional design does not allow us to imply causation, and the design may inherently bias our estimates of the indirect effect (Maxwell and Cole, 2007). Second, we must also recognize the limitations from the temporal assumption we are making with regard to the order of our variables in our model (Mathieu and Taylor, 2006). Third, this sample was relatively healthy and engaged in care (recruited by clinical staff during appointments, mean CD4 cell count >500, and nearly all on ART). Therefore, individuals in our sample may have been more resilient to begin with. Also, data show that individuals, if consistent with ART (with ART initiated at CD4 cell counts >350), can achieve CD4 cell count levels in the normal range (Gras et al., 2007). However, the mean depression for all participants was in the mild symptom category, suggesting mental illness is still a concern. Last, we are limited by the use of only self-report measures in our model. Still, all estimates of the prevalence of CSA, depressive symptomatology, and adherence levels were in the range found in other studies.

In conclusion, we successfully tested a model outlining a path from which CSA may exert a long-term health effect on HIV adherence. The pathway supported was through depression, but only for Latino HIV+ MSM whose index of resilience was below the 50th percentile. The data support the continual use and development of coping interventions, as well as further investigation into the conceptual and operational definitions of what it means to be resilient while living with HIV.

Footnotes

Acknowledgements

We are grateful to the patients and staff of Centro de Salud Familiar la Fe CARE Center, Inc. for their participation and facilitation of this study. We would like to thank Jessica Armendariz, Saul Dueñas, Jessica Garcia, Carolina Lara, Olga Khonyakina, Miriam Pando, and Yvette Venezuela, who assisted with data collection and entry.

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 present study was supported by a grant from the National Institute on Drug Abuse (5R24DA029989-03). The first author completed the manuscript with postdoctoral support by grant T32 MH19105 from the National Institute of Mental Health of the U.S. Public Health Service. Support to the third author was provided by K24MH093243 and University of Washington Center for AIDS Research (CFAR), an NIH-funded program (P30AI27757) which is supported by the following NIH Institutes and Centers (NIAID, NCI, NIMH, NIDA, NICHD, NHLBI, NIA).

References

Amico

Fisher

Cornman

. (2006) Visual analog scale of ART adherence: Associations with 3-day self-report and adherence barriers. Journal of Acquired Immune Deficiency Syndromes 42: 455–459.

Arreola

Neilands

Díaz

(2009) Childhood sexual abuse and the sociocultural context of sexual risk among adult Latino gay and bisexual men. American Journal of Public Health 99: S432–S438.

Arreola

Neilands

Pollack

. (2005) Higher prevalence of childhood sexual abuse among Latino men who have sex with men than non-Latino men who have sex with men: Data from the Urban Men’s Health Study. Child Abuse & Neglect 29: 285–290.

Arreola

Neilands

Pollack

. (2008) Childhood sexual experiences and adult health sequelae among gay and bisexual men: Defining childhood sexual abuse. Journal of Sex Research 45: 246–252.

Bing

Burnam

Longshore

. (2001) Psychiatric disorders and drug use among human immunodeficiency virus-infected adults in the United States. Archives of General Psychiatry 58: 721–728.

Bonanno

(2004) Loss, trauma, and human resilience: Have we underestimated the human capacity to thrive after extremely aversive events? American Psychologist 59: 20–28.

Boudewyn

Liem

(1995) Childhood sexual abuse as a precursor to depression and self-destructive behavior in adulthood. Journal of Traumatic Stress 8: 445–459.

Braithwaite

Justice

Chang

. (2005) Estimating the probabilities of patients infected with HIV who will die from comorbid diseases. American Journal of Medicine 118: 890–898.

Catz

Mcclure

Jones

. (1999) Predictors of outpatient medical appointment attendance among persons with HIV. AIDS Care 11: 361–373.

10.

Centers for Disease Control and Prevention (CDC) (2012) HIV in the United States: The stages of care. Available at: http://www.cdc.gov/hiv/pdf/research_mmp_StagesofCare.pdf (accessed 10 November 2013).

11.

Chesney

Ickovics

Chambers

. (2000) Self-reported adherence to antiretroviral medications among participants in HIV clinical trials: The AACTG adherence instruments. AIDS Care 12: 255–266.

12.

Cicchetti

(2010) Resilience under conditions of extreme stress: A multilevel perspective. World Psychiatry 9: 145–154.

13.

Cohen

McCauley

Gamble

(2012) HIV treatment as prevention and HPTN 052. Current Opinions in HIV/AIDS 7: 99–105.

14.

Connor

Davidson

JRT

(2003) Development of a new resilience scale: The Connor–Davidson Resilience Scale (CD-RISC). Depression and Anxiety 18: 76–82.

15.

Doll

Joy

Bartholow

. (1992) Self-reported childhood and adolescent sexual abuse among adult homosexual and bisexual men. Child Abuse & Neglect 16: 855–864.

16.

Felitti

Anda

Nordenberg

. (1998) Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. American Journal of Preventive Medicine 14: 245–258.

17.

Ferrando

Freyberg

(2008) Treatment of depression in HIV positive individuals: A critical review. International Review of Psychiatry 20: 61–71.

18.

Friedland

Williams

(1999) Attaining higher goals in HIV treatment: The central importance of adherence. AIDS 13(Suppl. 1): S61–S72.

19.

Giordano

Guzman

Clark

. (2004) Measuring adherence to antiretroviral therapy in a diverse population using a visual analog scale. HIV Clinical Trials 5: 74–79.

20.

Gonzalez

Batchelder

Psaros

. (2011) Depression and HIV/AIDS treatment nonadherence: A review and meta-analysis. Journal of Acquired Immune Deficiency Syndrome 58: 181–187.

21.

Gras

Kesselring

Griffin

. (2007) CD4 cell counts of 800 cells/mm³ or greater after 7 years of highly active antiretroviral therapy are feasible in most patients starting with 350 cells/mm³ or greater. Journal of Acquired Immune Deficiency Syndrome 45: 183–192.

22.

Hayes

(2009) Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs 76: 408–420.

23.

Hayes

(2013) Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York: Guilford Press.

24.

Heller

Larrieu

D’Imperio

. (1999) Research on resilience to child maltreatment: Empirical considerations. Child Abuse & Neglect 23: 321–338.

25.

Henry

. Kaiser Family Foundation (2013) HIV/AIDS policy fact sheet: Latinos and HIV. Available at: http://www.kff.org/hivaids/6007.cfm (accessed 3 October 2013).

26.

Hermann

Stewart

Diaz-Granado

. (2011) What is resilience? Canadian Journal of Psychiatry 56: 258–265.

27.

Herrick

Lim

Wei

. (2011) Resilience as an untapped resource in behavioral intervention design for gay men. AIDS and Behavior 15: 25–29.

28.

Hillberg

Hamilton-Giachritsis

Dixon

(2011) Review of meta-analyses on the association between child sexual abuse and adult mental health difficulties: A systematic approach. Trauma, Violence, & Abuse 12: 38–49.

29.

IBM Corp. (2012) IBM SPSS Statistics for Windows (version 19.0). Armonk, NY: IBM Corp.

30.

Jinich

Paul

Stall

. (1998) Childhood sexual abuse and HIV risk-taking behavior among gay and bisexual men. AIDS and Behavior 2: 41–51.

31.

Kent

Davis

Stark

. (2011) A resilience-oriented treatment for posttraumatic stress disorder: Results of a preliminary randomized clinical trial. Journal of Traumatic Stress 24: 591–595.

32.

Kurtz

Buttram

Surratt

. (2012) Resilience, syndemic factors, and serosorting behaviors among HIV-positive and HIV-negative substance-using MSM. AIDS Education and Prevention 24: 193–205.

33.

Leserman

(2008a) Role of depression, stress, and trauma in HIV disease progression. Psychosomatic Medicine 70: 539–545.

34.

Leserman

Barroso

Pence

. (2008b) Trauma, stressful life events and depression predict HIV-related fatigue. AIDS Care 20: 1258–1265.

35.

Leserman

Ironson

O’Cleirigh

. (2008c) Stressful life events and adherence in HIV. AIDS Patient Care and STDs 22: 403–411.

36.

Loprinzi

Prasad

Schroeder

. (2011) Stress Management and Resilience Training (SMART) program to decrease stress and enhance resilience among breast cancer survivors: A pilot randomized clinical trial. Clinical Breast Cancer 11: 364–368.

37.

Safren

Skolnik

. (2007) Optimal recall period and response task for self-reported HIV medication adherence. AIDS and Behavior 12: 86–94.

38.

Luthar

Cicchetti

Becker

(2000) The construct of resilience: A critical evaluation and guidelines for future work. Child Development 71: 543–562.

39.

MacKinnon

(2008) Introduction to Statistical Mediation Analysis. Mahwah, NJ: Lawrence Erlbaum.

40.

Mackinnon

Luecken

(2008) How and for whom? Mediation and moderation in health psychology. Health Psychology 72: S99–S100.

41.

Maniglio

(2010) Child sexual abuse in the etiology of depression: A systematic review of reviews. Depression and Anxiety 27: 631–642.

42.

Masten

(2001) Ordinary magic: Resilience processes in development. American Psychologist 56: 227–238.

43.

Mathieu

Taylor

(2006) Clarifying conditions and decision points for meditational type inferences in organizational behavior. Journal of Organizational Behavior 27: 1031–1056.

44.

Maxwell

Cole

(2007) Bias in cross-sectional analyses of longitudinal mediation. Psychological Methods 12: 23–44.

45.

Mimiaga

Noonan

Donnell

. (2009) Childhood sexual abuse is highly associated with HIV risk-taking behavior and infection among MSM in the EXPLORE study. Journal of Acquired Immune Deficiency Syndromes 51: 340–348.

46.

Morera

Castro

(2013) Important considerations in conducting statistical mediation analyses. American Journal of Public Health 103: 394–396.

47.

O’Cleirigh

Safren

Mayer

(2012) The pervasive effects of childhood sexual abuse: Challenges for improving HIV prevention and treatment interventions. Journal of Acquired Immune Deficiency Syndromes 59: 331–334.

48.

Oyugi

Byakika-Tusiime

Charlebois

. (2004) Multiple validated measures of adherence indicate high levels of adherence to generic HIV antiretroviral therapy in a resource-limited setting. Journal of Acquired Immune Deficiency Syndromes 36: 1100–1102.

49.

Ozer

Best

Lipsey

. (2008) Predictors of posttraumatic stress disorder and symptoms in adults: A meta-analysis. Psychological Trauma: Theory, Research, Practice, and Policy 1: 3–36.

50.

Page-Shafer

Delorenze

Satariano

. (1996) Comorbidity and survival in HIV-infected men in the San Francisco Men’s Health Survey. Annals of Epidemiology 6: 420–430.

51.

Paolucci

Genuis

Violato

(2001) A meta-analysis of the published research on the effects of child sexual abuse. Journal of Psychology 135: 17–36.

52.

Patterson

Shaw

Semple

. (1996) Relationship of psychosocial factors to HIV disease progression. Annals of Behavioral Medicine 18: 30–39.

53.

Preacher

Kelley

(2011) Effect size measures for mediation models: Quantitative strategies for communicating indirect effects. Psychological Methods 16: 93–115.

54.

Rothman

Exner

Baughman

(2011) The prevalence of sexual assault against people who identify as gay, lesbian, or bisexual in the United States: A systematic review. Trauma, Violence, & Abuse 12: 55–66.

55.

Safren

Radomsky

Otto

. (2002) Predictors of psychological well-being in a diverse sample of HIV-positive patients receiving highly active antiretroviral therapy. Psychosomatics 43: 478–485.

56.

Sethi

Celentano

Gange

. (2003) Association between adherence to antiretroviral therapy and human immunodeficiency virus drug resistance. Clinical Infectious Disease 37: 1112–1118.

57.

Simoni

Huh

Wilson

. (2012) Racial/ethnic disparities in ART adherence in the United States: Findings from the MACH14 study. Journal of Acquired Immune Deficiency Syndromes 60: 466–472.

58.

Simoni

Wiebe

Sauceda

. (2013) A preliminary RCT of CBT-AD for adherence and depression among HIV-positive Latinos on the U.S.-Mexico border: The Nuevo Dia study. AIDS and Behavior 17: 2816–2829.

59.

Stall

Mills

Williamson

(2003) Association of co-occurring psychosocial health problems and increased vulnerability to HIV/AIDS among urban men who have sex with men. American Journal of Public Health 93: 939–942.

60.

Van Eeden-Moorefield

(2008) A theory of resilience applied to couples affected by HIV/AIDS. Journal of LGBT Family Studies 4: 433–452.

61.

Widom

Spatz

Czaja

. (2012) A prospective investigation of physical health outcomes in abused and neglected children: New findings from a 30-year follow-up. American Journal of Public Health 102: 1135–1144.

62.

Wiebe

Sauceda

Lara

(2013) Assessing mood disorders and suicidality in Hispanics. In: Benuto

(ed.) Guide to Psychological Assessment with Hispanics. New York: Springer, pp. 113–128.

63.

Williams

Nelson-Gardell

(2012) Predicting resilience in sexually abused adolescents. Child Abuse & Neglect 36: 53–63.