Adherence to highly active antiretroviral therapy (HAART) has proven to be a major challenge for individuals diagnosed with HIV infection.1 Successful adherence to HAART has been shown to significantly improve virologic and immunologic outcomes for HIV-infected patients.2-9 Conversely, nonadherence to HAART can lead to treatment failure and lower quality of life due to increased incidence of opportunistic infections.5,8,10 Nonadherence can also result in development of HIV antiretroviral resistance,11,12 limitations on future treatment options, and potential transmission of resistant strains of HIV to others via sexual and/or injecting drug use behavior.13-16 Complicating the prospect of successful adherence to HAART is the level of adherence required for clinical efficacy. Unlike other chronic diseases such as hypertension or diabetes, where 80% adherence is generally considered adequate,17 adherence to HIV antiretroviral medications of ≥95% is considered necessary to achieve and maintain viral suppression.5,18
Numerous barriers to adherence to antiretroviral therapy have been documented, including illicit drug use,19-22 lack of social support,23-26 and the extent and severity of side effects related to antiretroviral medications.27-29 Psychologic factors such as depression, anxiety, and stress are among the strongest and most consistent predictors of nonadherence to antiretroviral medications.19,20,23-25,30-34 These factors have also been independently correlated with higher viral loads,35 declining CD4 cell count,36,37 faster progression to AIDS,38-40 development of AIDS-defining clinical conditions,39 and increased mortality among HIV-infected individuals,37,39 although the immunologic mechanisms that influence these outcomes are not clearly understood.35 In addition, psychologic distress is a major risk factor for illicit drug use and excessive consumption of alcohol,41,42 behaviors that have been independently associated with nonadherence to antiretroviral medications.19,21,43
Several studies have attempted to estimate the prevalence of various forms of psychologic distress among HIV-infected individuals. In a review of depression and HIV, Angelino and Treisman42 estimated that prevalence of major depression among HIV-infected individuals ranged from 15% to 40%. Cohen et al44 assessed the prevalence of distress, anxiety, and depression in a convenience sample of 101 patients selected from the waiting room of a major AIDS center and found that 72.3% had high distress scores, 70.3% had high anxiety scores, and 45.5% had scores indicating depression. Morrison et al45 compared the prevalence of anxiety and depressive disorders between 93 HIV-infected and 62 uninfected women. The rate of depression among the HIV-infected group was 19.4%, 4 times that observed among uninfected women (4.8%). In a review of published literature in the same article, the investigators presented estimates of the prevalence of major depressive disorder among HIV-infected women that ranged from 1.9% to 35% in clinical samples and from 30% to 60% in community samples. Cruess et al,46 in a review of the prevalence of mood disorders in HIV disease, estimated that rates of depression ranged from 5% to 20%, and a meta-analysis of published studies of depression and HIV indicated that HIV-infected individuals were nearly twice as likely to develop major depressive disorder as HIV-negative individuals.47
Several measures of psychologic distress have been used in published research studies investigating adherence to anti-retroviral medications. Examples include the Center for Epidemiological Studies Depression Scale19,20,24,30,32 and the Beck Depression Inventory.23,48 Other research has used a composite of social indicators to estimate stress levels in HIV-infected clients.34 However, few published studies have offered brief psychological scales that could be easily applied in various settings to help identify individuals at higher risk of nonadherence. This research focuses on the relationship between a revised, 4-item version of the Perceived Stress Scale (PSS)49 and nonadherence to HAART among clients enrolled in treatment adherence support programs in New York State.
In 1999, the New York State Department of Health AIDS Institute funded 14 programs to provide multifaceted adherence support to HIV-infected clients. Adherence programs were designed as networks of clinical and non-clinical providers who worked collaboratively to address the multiple needs of HIV-infected individuals receiving HAART. Each program was required to develop individualized treatment plans, which were to be modified according to the changing circumstances of clients. Programs differed according to their theoretical underpinnings but, in general, posited a belief in “meeting the client where he/she is” in terms of readiness to engage in various interventions. Examples of services offered by adherence programs include identification of adherence barriers and risk factors; HIV/antiretroviral treatment education; thorough review of client's antiretroviral regimen; identification and management of side effects; peer education/support; case management/social work; pharmacy education/support services; and one-on-one counseling.
Data were collected by client interview and chart abstraction upon enrollment (baseline) and approximately every 3 months thereafter. The median number of days between baseline and follow-up 1 (F1) and between follow-up 1 and follow-up 2 (F2) was 100 and 99, respectively. Chart abstraction consisted mainly of items easily accessible in the client's medical record. Structured client interviews included demographics, psychosocial data, and several questions related to the client's adherence to antiretroviral medications.
Measurement of Adherence
Adherence was measured by 3-day self-report during the client interview. Clients were asked what HIV antiretroviral medications they were taking, the number of pills and doses prescribed per day for each medication, and the number of missed doses for each antiretroviral in each of the 3 days preceding the interview. Adherence was dichotomized to 100% adherent versus <100% adherent. Although 95% adherence is generally recommended, a 100% cutoff was used to simplify interpretation. Due to the frequency of low-dose regimens, only 12 (2.0%) of 590 clients had mean adherence between 95% and 100% at any time across 3 interviews.
The 14-item PSS was originally introduced by Cohen et al49 in 1983 and was shown to have adequate reliability and to be independently correlated with life-event scores, depressive symptoms, and social anxiety. The shorter 4-item version of the scale was proposed as an effective tool for repeated measures among large groups over time or in situations in which limiting the number of items is an important consideration.49 Cohen and Williamson50 introduced a 10-item version of the PSS in 1988, which Chesney et al51 later applied in adherence research with the Adult AIDS Clinical Trials Groups.
Individual items from the short version of the PSS are displayed in Table 1. Basic wording of the individual questions was not substantially changed from the original version of the scale. However, the Likert scale response options were changed from “never,” “almost never,” “sometimes,” “fairly often,” and “very often” to “never or rarely,” “sometimes,” “often,” and “mostly or always.” Response options were changed because demonstration project and evaluation staff thought that the latter categories provided options that were more distinctly different from one another than previous versions of this scale. It was also believed that reducing the response options from 5 to 4 items would improve client comprehension of, and response to, the perceived stress questions, particularly in an interview setting. These conclusions were subjective, however, and were not based on previous empirical research.
Perceived stress questions were examined individually and as part of an index of responses across all 4 questions. Specifically, responses to individual questions were scored from 1 to 4, with 1 and 4 representing the least and most desirable responses for each question, respectively. Thus, a range of scores from 4 to 16 was possible. Scores were then grouped according to the quartile distribution of responses overall. The range of responses for these quartiles was 0-6 (≤25th percentile), 7-8 (26-50th percentile), 9-10 (51-75th percentile), and 11-16 (>75th percentile). The cutoffs for determining the quartile distribution of responses were the same at baseline, F1, and F2. A 4-category perceived stress variable was then computed based on this distribution, with groupings of ≤6, 7-8, 9-10, and ≥11.
Several other independent and control variables were used in this research. Basic demographic variables included sex, age, and race/ethnicity. Illicit drug use was measured by asking clients whether they had taken cocaine, crack, or heroin in the 3 months before the interview with yes/no response options. Similarly, clients were asked about whether they had “regularly” drunk ≥3 alcoholic drinks per day during the 3 months before the interview. Belief in the efficacy of medications was assessed by asking clients how sure they were that the HIV medications would help them to fight HIV infection. Response options were “not sure,” “pretty sure,” and “very sure,” which were then dichotomized to “not sure” versus “pretty/very sure.” Clients were asked what housing they currently resided in, with initial response categories being “homeless,” “transitional,” “institutional,” and “stable.” Responses were subsequently grouped according to whether clients reported “stable” housing versus all other housing categories. Analyses also controlled for the length of time since diagnosis of HIV infection, number of days between interviews, and number of pills in the antiretroviral regimen.
The χ2 test of independence was used to investigate differences in nonadherence according to categorical variables, such as illicit drug use and perceived stress. Tukey's multiple comparison procedure was used to assess significant differences between categories for multiple category variables. Independent samples t test procedures were used to examine mean differences in continuous variables, such as age, according to adherence status. Multiple logistic regression analyses were performed to identify variables significantly associated with nonadherence while controlling for the potentially confounding effects of other variables. All statistical analyses were performed using SPSS version 11.5 (Chicago, IL).
Comparison of Included and Excluded Clients
Data regarding clients included and excluded from analyses are shown in Table 2. Fifteen hundred two clients were enrolled in the adherence programs and receiving HAART at the time of their baseline interview, of whom 590 (39.3%) completed 2 quarterly follow-up interviews and had complete data regarding adherence and perceived stress. This subgroup of 590 clients represented 10 of the 14 participating treatment adherence programs. Clients who were excluded from the analyses were significantly less likely to have stable housing (P < 0.05) and were significantly more likely to have been diagnosed with HIV infection >1 year before the baseline interview (P < 0.05). Clients with a diagnosis of mental illness were more likely to have been included in the analysis (P = 0.05), although this relationship was only marginally significant. Otherwise, no significant differences were found between included and excluded groups (Table 2).
Bivariate Relationship Between Perceived Stress and Nonadherence
Table 3 displays the proportion of clients who reported being nonadherent to their antiretroviral regimen at baseline, F1, and F2 according to the quartile score for each individual perceived stress question. At baseline, 2 of the perceived stress questions (c and d in Table 1) showed a statistically significant relationship with nonadherence. At F1 and F2, all 4 perceived stress questions were significantly associated with nonadherence. α reliability coefficients were 0.72 at baseline, 0.69 at F1, and 0.72 at F2-marginally high considering the limited number of items in the scale. Factor analysis indicated the presence of only 1 factor, explaining 54.0%, 51.7%, and 54.3% of the variance at baseline, F1, and F2, respectively. These results provided support for the development of 1 overall index of perceived stress (at each time period), which was used for the remainder of the analysis.
Table 4 displays the percentage of nonadherence at baseline, F1, and F2 according to the quartile distribution of the combined, overall perceived stress score. This index of overall perceived stress was significantly associated with nonadherence across all interview periods. At baseline, clients with overall perceived stress scores in the highest quartile were significantly more likely to be nonadherent (42.7%) than clients in the lowest quartile (25.0%). At F1, clients who scored in the highest quartile were significantly more likely to be nonadherent (39.4%) than clients in both the lowest (16.7%) and the next lowest (22.4%) quartiles. At F2, clients who scored in the highest quartile were significantly more likely to be nonadherent (47.7%) than clients in all lower categories (31.0%, quartile 3; 24.5%, quartile 2; 11.5%, quartile 1). Further, clients who scored in the lowest quartile were significantly less likely to be nonadherent (11.5%) than clients in all other categories.
Multiple Logistic Regression Models Predicting Nonadherence
Three multivariate models were constructed to determine whether perceived stress was a statistically significant predictor of nonadherence at baseline, F1, and F2 while controlling for the potentially confounding effects of other variables entered into the models. Independent variables that were entered into the models consisted of variables that were significantly associated with nonadherence at the bivariate level (ie, perceived stress, illicit drug use, frequent alcohol use, housing status, medication efficacy, and when the client tested HIV positive) as well as a number of demographic control variables (ie, sex, age, race/ethnicity, and number of days between interviews).
Table 5 displays the results of the multiple logistic regression model predicting nonadherence at baseline. Perceived stress was a strong and significant predictor of nonadherence, controlling for other variables entered into the model. Specifically, clients in the highest quartile of perceived stress were approximately twice as likely to be nonadherent as were clients who scored in the lowest quartile (odds ratio [OR] = 2.12). Clients in the 35- to 44-year-old age group were also significantly more likely to be nonadherent than clients aged 45 years and older (OR = 1.53). Clients who tested positive for HIV within the past year were significantly less likely to be nonadherent than clients who tested positive >1 year ago (OR = 0.40). Lastly, clients who were prescribed 1-5 pills (OR = 0.50), 6-10 pills (OR = 0.53), and 11-15 pills (OR = 0.54) in their regimen were significantly less likely to be nonadherent than clients who reported a regimen of ≥16 pills each day.
Table 6 displays the results of the multiple logistic regression model predicting nonadherence at F1. A number of variables were significantly associated with nonadherence at this time. First, perceived stress continued to remain a strong and significant predictor of nonadherence. Clients who scored in the top quartile of perceived stress were about twice as likely as clients in the lowest quartile to be nonadherent at F1 (OR = 2.16). Other variables that were associated with a greater likelihood of nonadherence included female sex (OR = 1.76), lack of confidence in the efficacy of HIV medications (OR = 1.95), frequent alcohol use (OR = 2.56), illicit drug use (OR = 1.79), and nonadherence at baseline (OR = 2.59). Clients who lived in stable housing were significantly less likely to be nonadherent at F1 (OR = 0.56), and clients who were prescribed 1-5 pills per day were significantly less likely to be nonadherent than the reference group of clients who were prescribed ≥16 pills per day (OR = 0.38).
Results of the multiple logistic regression model predicting nonadherence at F2 are presented in Table 7. As in previous models predicting nonadherence at baseline and F1, perceived stress was a significant predictor of nonadherence at F2. In this case, clients who scored in all higher quartile ranges of perceived stress scores were significantly more likely to be nonadherent than clients in the lowest quartile of perceived stress scores. Clients who scored in the highest quartile of perceived stress were >5 times as likely to report being nonadherent than clients in the lowest quartile (OR = 5.12). Those clients who scored in the 2nd and 3rd quartiles of perceived stress were between 2 (OR = 2.33) and 3 (OR = 2.72) times as likely to be nonadherent at F2 (OR = 2.33 and 2.72, respectively). Clients who reported being nonadherent at baseline were almost twice as likely to be nonadherent at F2 (OR = 1.63), and clients who were nonadherent at F1 were nearly 4 times as likely to be nonadherent at F2 (OR = 3.74). Those clients who tested positive for HIV within the past year were significantly less likely to be nonadherent at F2 (OR = 0.29).
Some limitations of this research should be described. Our measure of perceived stress was derived from a validated measure49 but was revised based on the judgment of demonstration project and evaluation staff. Response options were shortened from 5 to 4, and the responses themselves were altered slightly with the intention of making them more distinctly different from each other and more comprehensible to clients. Language and literacy issues precluded the project from utilizing self-administered surveys to collect these data, which may have made it more practical to use the longer versions of the scale with the original response options. Furthermore, it is possible that the 14- or 10-item versions of the PSS may be more effective than the 4-item version at distinguishing clients with adherence problems. However, part of the rationale for using the abbreviated version is its convenience and ease of use, attributes that may be compromised by use of the longer versions of the scale.
Adherence was measured by self-report, which has been shown to overestimate adherence52 or, in the context of this research, to underestimate nonadherence. However, self-report may represent the most realistic method of assessing adherence to antiretroviral medications in nonresearch settings, because it is inexpensive and relatively easy to implement. Adherence questions were administered by a variety of clinical and non-clinical staff across a range of institutional and community-based settings where adherence support and medical care were being provided. Self-report has also been shown to correlate well with other more objective measures of adherence52,53 as well as with clinical markers such as viral load.2,3,54 With this in mind, it is important for clinical and non-clinical providers to be aware of the social desirability of being adherent to treatment, so they can create an atmosphere of tolerance and understanding that will encourage clients to be honest regarding missed doses of antiretroviral medications and other instances of nonadherence to their treatment regimen.
Because the demonstration project was primarily funded as a service delivery model, it was necessary to minimize the frequency and content of interview questions to avoid burdening providers and clients/patients. Although this research benefits from the fact that data were collected in relatively “real life” circumstances-during interactions with clinical and community-based providers-many dimensions were assessed using relatively crude measures. For example, substance use may have been better measured using a more in-depth, validated scale.
Last, this research reflects the experiences of clients attending adherence support programs in New York State. This may limit the generalizability of results to the HIV-infected population receiving care in other settings in New York, the United States, and elsewhere.
Estimates of the prevalence of psychologic distress among HIV-infected individuals have varied considerably based on several factors, including the construct being measured (ie, depression, anxiety, coping), the method used to assess these constructs (ie, Center for Epidemiological Studies Depression Scale, Beck Depression Inventory, Diagnostic and Statistical Manual of Mental Disorders criteria), the population being studied (ie, men who have sex with men, injecting drug users, women), and the study methodology (ie, sample design, cross-sectional vs. longitudinal). Regardless of the variability in these estimates, it is clear that psychologic distress is often higher in HIV-infected than uninfected individuals.46 The impact of psychological distress on adherence to antiretroviral medications is also well documented. In fact, a meta-analysis of the effects of depression on adherence to medical treatment in general found that depressed patients were ~3 times more likely to be nonadherent to treatment recommendations.55 Psychological distress is also a risk factor for substance use, which is likewise associated with poor adherence to antiretroviral medications.19-21 Despite these findings, few psychological scales used in published adherence research are brief and convenient enough to be easily applied in clinical and non-clinical treatment settings. Our findings suggest that-in addition to routine measurement of adherence over time-this revised, 4-item version of the PSS may serve as an effective, additional tool to assist in identifying those at higher risk of nonadherence to target these individuals for enhanced adherence support and referral to mental health, substance use, and other necessary services. In multivariate analysis, clients who scored in the highest quartile of perceived stress scores were significantly more likely to be nonadherent at baseline and over time. In 2 (baseline and F2) of the 3 models, perceived stress was a stronger predictor of nonadherence than all other variables entered into the model.
This scale may represent a viable alternative to other more labor-intensive psychologic measures, especially in institutions and settings hampered by limited health care resources, where time spent with clients and patients is already constrained. This revised, 4-item perceived stress measure takes ~1-2 minutes to complete and can easily be used in various clinical and non-clinical settings without taking excess time away from other necessary activities. In addition, the measure is straightforward and requires no specialized training for its use.
The PSS can also be used to identify the need for a short-term intervention to relieve an acutely stressful situation, refer the client for a more in-depth mental health assessment, and make long-range plans to address any identified issues as part of the adherence service program and/or in conjunction with the overall health care treatment plan. Future research should investigate the applicability of this measure among specific subgroups and in studies of other medical and behavioral conditions, because its usefulness is likely to extend beyond just the field of HIV treatment adherence.
The authors gratefully acknowledge the assistance of the staff of the following adherence support demonstration programs: AIDS Community Resources, Albany Medical Center, Bellevue Hospital Center, Brooklyn AIDS Task Force, Harlem Hospital Center, Montefiore Medical Center, Mount Sinai Medical Center, New York Presbyterian Hospital, North Shore University Hospital, Village Center for Care, and Westchester Medical Center and all of their network sites. In addition the authors thank Donna O'Connor of the New York State Department of Health AIDS Institute for data entry and administrative support on the project.
1. Bartlett JA. Addressing the challenges of adherence. J Acquir Immune Defic Syndr
. 2002;29(Suppl 1):S2-S10.
2. Knobel H, Carmona A, Grau S, et al. Adherence and effectiveness of highly active antiretroviral therapy. Arch Intern Med
3. Bangsberg D, Hecht F, Charlebois E, et al. Adherence to protease inhibitors, HIV-1 viral load, and development of drug resistance in an indigent population. AIDS
4. Maher K, Klimas N, Fletcher MA, et al. Disease progression, adherence, and response to protease inhibitor therapy for HIV infection in an urban Veterans Affairs medical center. J Acquir Immune Defic Syndr
5. Paterson DL, Swindells S, Mohr J. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med
6. McNabb J, Ross JW, Abriola K, et al. Adherence to highly active antiret-roviral therapy predicts virologic outcome at an inner-city human immunodeficiency virus clinic. Clin Infect Dis
7. Le Moing V, Chene G, Carrieri MP, et al. Clinical, biologic, and behavioral predictors of early immunologic and virologic response in HIV-infected patients initiating protease inhibitors. J Acquir Immune Defic Syndr
8. Bangsberg DR, Perry S, Charlebois ED, et al. Non-adherence to highly active antiretroviral therapy predicts progression to AIDS. AIDS
9. Knobel H, Guelar A, Carmona A, et al. Virologic outcome and predictors of virologic failure of highly active antiretroviral therapy containing protease inhibitors. AIDS Patient Care STDS
10. Deeks SG. Determinants of virological response to antiretroviral therapy: implications for long-term strategies. Clin Infect Dis
. 2000;30(Suppl 2):S177-S184.
11. Pomerantz RJ. Primary HIV-1 resistance: a new phase in the epidemic? JAMA
12. D'Aquila RT, Johnson VA, Welles SL, et al. Zidovudine resistance and HIV-1 disease progression during antiretroviral therapy. Ann Intern Med
13. Erice A, Mayers DL, Strike DG, et al. Primary infection with zidovudine-resistant human immunodeficiency virus type 1. N Engl J Med
14. Imrie A, Beveridge A, Genn W, et al. Transmission of human immuno-deficiency virus type 1 resistant to nevirapine and zidovudine. J Infect Dis
15. Hecht FM, Grant RM, PetroPoulos CJ, et al. Sexual transmission of an HIV-1 variant resistant to multiple reverse-transcriptase and protease inhibitors. N Engl J Med
16. Boden D, Hurley A, Zhang L, et al. HIV-1 drug resistance in newly infected individuals. JAMA
17. Sackett D, Snow J. The magnitude of compliance and non-compliance. In: Haynes R, Taylor D, Sackett D, eds. Compliance in Health Care
. Baltimore, MD: Johns Hopkins University Press; 1979:11-12.
18. Low-Beer S, Yip B, O'Shaughnessy MV, et al. Adherence to triple therapy and viral load response. J Acquir Immune Defic Syndr
19. Arnsten JH, Demas PA, Grant RW, et al. Impact of active drug use on antiretroviral therapy adherence and viral suppression in HIV-infected drug users. J Gen Intern Med
20. Bouhnik D-A, Chesney M, Carrieri P, et al. Nonadherence among HIV-infected injecting drug users: the impact of social instability. J Acquir Immune Defic Syndr
. 2002;31(Suppl 3):S149-S153.
21. Lucas GM, Gebo KA, Chaisson RE, et al. Longitudinal assessment of the effects of drug and alcohol abuse on HIV-1 treatment outcomes in an urban clinic. AIDS
22. Stein MD, Rich JD, Maksad J, et al. Adherence to antiretroviral therapy among HIV-infected methadone patients: effect of ongoing illicit drug use. Am J Drug Alcohol Abuse
23. Safren SA, Otto MW, Worth JL, et al. Two strategies to increase adherence to HIV antiretroviral medication: life-steps and medication monitoring. Behav Res Ther
24. Catz SL, Kelly JA, Bogart LM, et al. Patterns, correlates, and barriers to medication adherence among persons prescribed new treatments for HIV disease. Health Psychol
25. Spire B, Duran S, Souville M, et al. Adherence to highly active antiretro-viral therapies (HAART) in HIV-infected patients: from a predictive to a dynamic approach. Soc Sci Med
26. Singh N, Berman SM, Swindells S, et al. Adherence of human immuno-deficiency virus-infected patients to antiretroviral therapy. Clin Infect Dis
27. Roca B, Gomez CJ, Arnedo A. Adherence, side effects, and efficacy of stavudine plus lamivudine plus nelfinavir in treatment-experienced HIV-infected patients. J Infect
28. Max B, Sherer R. Management of the adverse effects of antiretroviral therapy and medication adherence. Clin Infect Dis
. 2000;30(Suppl 2):S96-S116.
29. Ammassari A, Murri R, Pezzotti P, et al. Self-reported symptoms and medication side effects influence adherence to highly active antiretroviral therapy in persons with HIV infection. J Acquir Immune Defic Syndr
30. Kleeberger CA, Phair JP, Strathdee SA, et al. Determinants of heterogeneous adherence to HIV-antiretroviral therapies in the Multicenter AIDS Cohort Study. J Acquir Immune Defic Syndr
31. Tesoriero J, French T, Weiss L, et al. Stability of adherence to highly active antiretroviral therapy over time among clients enrolled in the Treatment Adherence Demonstration Project. J Acquir Immune Defic Syndr
32. Malcolm S, Ng J, Rosen R, et al. An examination of HIV/AIDS patients who have excellent adherence to HAART. AIDS Care
33. Ammassari A, Trotta MP, Murri R, et al. Correlates and predictors of adherence to highly active antiretroviral therapy: overview of published literature. J Acquir Immune Defic Syndr
34. Gebo KA, Keruly J, Moore RD. Association of social stress, illicit drug use, and health beliefs with nonadherence to antiretroviral therapy. J Gen Intern Med
35. Evans DL, Ten Have TR, Douglas SD, et al. Association of depression with viral load, CD8 T lymphocytes, and natural killer cells in women with HIV infection. Am J Psychiatry
36. Burack J, Barrett D, Stall R, et al. Depressive symptoms and CD4 lymphocyte decline among HIV-infected men. JAMA
37. Ickovics JR, Hamburger ME, Vlahov D, et al. Mortality, CD4 cell count decline, and depressive symptoms among HIV-seropositive women: longitudinal analysis from the HIV Epidemiology Research Study. JAMA
38. Leserman J, Jackson ED, Petitto JM, et al. Progression to AIDS: the effects of stress, depressive symptoms, and social support. Psychosom Med
39. Leserman J, Petitto J, Gu H, et al. Progression to AIDS, a clinical AIDS condition and mortality: psychosocial and physiological predictors. Psychol Med
40. Golub ET, Astemborski JA, Hoover DR, et al. Psychological distress and progression to AIDS in a cohort of injection drug users. J Acquir Immune Defic Syndr
41. Bing E, Burnam A, Longshore D, et al. Psychiatric disorders and drug use among human immunodeficiency virus-infected adults in the United States. Arch Gen Psychiatry
42. Angelino AF, Treisman GJ. Management of psychiatric disorders in patients infected with human immunodeficiency virus. Clin Infect Dis
43. Golin CE, Liu H, Hays RD, et al. A prospective study of predictors of adherence to combination antiretroviral medication. J Gen Intern Med
44. Cohen M, Hoffman R, Cromwell C, et al. The prevalence of distress in persons with human immunodeficiency virus infection. Psychosomatics
45. Morrison M, Petitto J, Ten Have T, et al. Depressive and anxiety disorders in women with HIV infection. Am J Psychiatry
46. Cruess DG, Evans DL, Repetto MJ, et al. Prevalence, diagnosis, and pharmacological treatment of mood disorders in HIV disease. Biol Psychiatry
47. Ciesla JA, Roberts JE. Meta-analysis of the relationship between HIV infection and risk for depressive disorders. Am J Psychiatry
48. Gifford AL, Bormann JE, Shively MJ, et al. Predictors of self-reported adherence and plasma HIV concentrations in patients on multidrug anti-retroviral regimens. J Acquir Immune Defic Syndr
49. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Social Behav
50. Cohen S, Williamson G. Perceived stress in a probability sample of the United States. In: Spacapan S, ed. The Social Psychology of Health: Claremont Symposium on Applied Social Psychology
. Newbury Park: Sage; 1988:31-67.
51. Chesney MA, Ickovics JR, Chambers DB, et al. Self-reported adherence to antiretroviral medications among participants in HIV clinical trials: The AACTG Adherence Instruments. AIDS Care
52. Liu H, Golin CE, Miller LG, et al. A comparison study of multiple measures of adherence to HIV protease inhibitors. Ann Intern Med
53. Wagner JH, Justice AC, Chesney MA, et al. Patient- and provider-reported adherence: toward a clinically useful approach to measuring antiretroviral adherence. J Clin Epidemiol
54. Sethi AK, Celentano DD, Gange SJ, et al. Association between adherence to antiretroviral therapy and human immunodeficiency virus drug resistance. Clin Infect Dis
55. DiMatteo RM, Lepper HS, Croghan TW. Depression is a risk factor for noncompliance with medical treatment. Arch Intern Med