Predictors of Suboptimal Virologic Response to Highly Active Antiretroviral Therapy Among Human Immunodeficiency Virus–Infected Adolescents: Analyses of the Reaching for Excellence in Adolescent Care and Health (REACH) Project | Adolescent Medicine | JAMA Pediatrics | JAMA Network
[Skip to Navigation]
Sign In
December 7, 2009

Predictors of Suboptimal Virologic Response to Highly Active Antiretroviral Therapy Among Human Immunodeficiency Virus–Infected Adolescents: Analyses of the Reaching for Excellence in Adolescent Care and Health (REACH) Project

Author Affiliations

Author Affiliations: The Ginn Group, Inc, East Point, and Georgia Prevention Research Branch, Division of HIV/AIDS, National Centers for HIV, Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta (Dr Ding); Department of Epidemiology, Medicine, and Pediatrics, University of Alabama at Birmingham (Drs Ding, Wilson, McGwin, and Tang); and Departments of Medicine and Pediatrics and the Institute for Global Health, Vanderbilt University School of Medicine, Nashville, Tennessee (Drs Modjarrad and Vermund).

Arch Pediatr Adolesc Med. 2009;163(12):1100-1105. doi:10.1001/archpediatrics.2009.204

Objective  To examine the prevalence and biopsychosocial predictors of suboptimal virologic response to highly active antiretroviral therapy (HAART) among human immunodeficiency virus–infected adolescents.

Design  Population-based cohort study.

Setting  Sixteen academic medical centers across 13 cities in the United States.

Participants  One hundred fifty-four human immunodeficiency virus–infected adolescents who presented for at least 2 consecutive visits after initiation of HAART.

Main Outcome Measures  Viral load (plasma concentration of human immunodeficiency virus RNA) and CD4+ lymphocyte count.

Results  Of the 154 adolescents enrolled in the study, 50 (32.5%) demonstrated early and sustained virologic suppression while receiving HAART. The remaining 104 adolescents (67.5%) had a poor virologic response. Adequate adherence (>50%)—reported by 70.8% of respondents—was associated with 60% reduced odds of suboptimal virologic suppression in a multivariable logistic regression model (adjusted odds ratio = 0.4; 95% confidence interval, 0.2-1.0). Exposure to suboptimal antiretroviral therapy prior to HAART, on the other hand, was associated with more than 2-fold increased odds of suboptimal virologic response (adjusted odds ratio = 2.6; 95% confidence interval, 1.1-5.7).

Conclusions  Fully two-thirds of human immunodeficiency virus–infected adolescents in the current study demonstrated a suboptimal virologic response to HAART. Nonadherence and prior single or dual antiretroviral therapy were associated with subsequent poor virologic responses to HAART. These predictors of HAART failure echo findings in pediatric and adult populations. Given the unique developmental stage of adolescence, age-specific interventions are indicated to address high rates of nonadherence and therapeutic failure.

In 2006, adolescents aged 13 to 19 years accounted for 4% of newly diagnosed cases of human immunodeficiency virus (HIV) infection in the United States,1 half the percentage reported in 2003.2 Despite a decrease in the relative proportion of adolescents diagnosed as having HIV infection, incidence rates within this age group have been steadily increasing since the late 1990s, nearly doubling between 1998 and 2006.1 However, these data are likely to underestimate the true incidence and proportion of adolescent HIV cases, as overall incidence rate estimates were revised upward in 2008 and HIV testing coverage is much lower among adolescent populations compared with adult populations.3,4 Furthermore, the prolonged latent phase of HIV's natural history suggests that a large proportion of individuals diagnosed as having HIV infection in early adulthood were likely infected during adolescence.

Adolescents are less likely to be tested for HIV as they access health care services less frequently and with greater difficulty than their adult counterparts.5-7 Behavioral studies have demonstrated that high-risk behaviors, lack of knowledge about HIV, and logistic impediments to accessing health care systems may contribute to delays in testing adolescents for HIV infection.8,9 Delays in HIV diagnosis lead to delays in treatment initiation, which can result in poorer clinical outcomes. The delays to initiating care are compounded by the fact that adolescents lack adequate health care insurance and do not often have access to youth-friendly clinical settings.10,11 These disadvantages along with complicated therapeutic regimens may predispose adolescents to poor antiretroviral therapy (ART) adherence rates, which can translate into viral resistance and inadequate viral suppression.

The relationship between viral suppression and immune reconstitution in the context of highly active ART (HAART) has been well documented in HIV-infected adult and pediatric populations.12-16 Predictors of poor virologic, immunologic, and clinical outcomes have also been validated in these cohorts.17-20 Adolescents differ in their biological and behavioral characteristics from these age groups but are still often grouped with adults in epidemiologic studies. Given the paucity of data on HIV-infected youth populations, we sought to identify predictors of elevated viral load and decreased CD4+ lymphocyte counts among adolescents receiving HAART.


Study design

The REACH Project (Reaching for Excellence in Adolescent Care and Health) of the Adolescent Medicine HIV/AIDS Research Network is a multicenter, prospective, observational study that longitudinally evaluated prevalent HIV infections and related behavioral and biological factors in adolescents (aged 12-19 years) for a minimum of 1 year (mean follow-up, 2.4 years) between 1996 and 2000. This time frame allowed the assessment of the transition from single and dual ART to HAART. The study cohort was restricted to participants who acquired HIV via sexual transmission or injection drug use and excluded youth infected through vertical transmission, contaminated blood products, or early childhood sexual abuse. Detailed descriptions of the REACH Project study design, objectives, and procedures have been published elsewhere.21,22

Participants included in the current analysis were required to meet 5 inclusion criteria: (1) HIV seropositive at baseline; (2) HAART naive at baseline; (3) receiving HAART for at least 2 consecutive visits following HAART initiation; (4) documented baseline viral load; and (5) viral load available before and after HAART initiation. Approvals for the original study and subsequent analyses were obtained from all participating universities' institutional review boards. Informed consent was obtained from all participants, parents, or guardians as deemed appropriate by local laws and/or university regulations.

Behavioral data were collected via face-to-face interviews, audio computer-assisted self-administered interviews, and medical record reviews. Study participants defined their race/ethnicity during direct interviews by self-identification. Interviewers measured and rated depressive signs and symptoms with the Center for Epidemiologic Study Depression Scale.23 Details of adherence assessments have been published elsewhere.24 For the purpose of this analysis, a participant who took all of his or her triple antiretroviral medications as prescribed more than 50% of the time prior to a particular visit was deemed to be adequately adherent at that specific visit. (Data on adherence to specific medications within a triple drug regimen were not collected.) If a participant then maintained adequate adherence at more than 50% of his or her visits, he or she was defined as having overall adequate adherence for the duration of the study. This definition of ART adherence deviates from standard criteria25 but was crafted to be more sensitive at the cost of being less specific to capture an adequate sample size in a population that has a high risk of nonadherence.

Host genetic factors that are collectively predictive of better immunologic control of HIV replication were captured by an algorithm that has been described previously.26 In brief, HLA class I antigens and chemokine receptor 2 and chemokine receptor 5 genotypes that have a favorable effect on controlling HIV replication and/or delaying disease progression were assigned a score of +1, while unfavorable genotypes received a score of −1. At the individual level, a genetic score was calculated by summing the assigned values for all contributing HLA, chemokine receptor 2, and chemokine receptor 5 variants within the individual. Within the study population, genetic scores ranged from −3 to +1, although individuals with extremely low scores were rare. Therefore, we collapsed scores of −3,−2, and −1 into a single category of −1 for the purpose of analysis. Virologic and immunologic measures were obtained according to previously defined protocols.21,22


Participants were categorized according to the magnitude and sustainability of viral load reductions while receiving HAART. We defined early viral responders as those who had a viral load reduction greater than 1.0 log10 copies/mL between the pre-HAART visit and the initial 2 post-HAART visits. We identified individuals as demonstrating a sustained viral response if their mean viral load from the remaining post-HAART visits was at least 0.5 log10 copies/mL lower than their pre-HAART value. Participants were ultimately classified into 2 categories: (1) those who had both early and sustained virologic suppression were termed optimal responders; and (2) those who either demonstrated early but unsustained virologic suppression or showed no early response at all were classified as suboptimal responders. The latter group is a composite of 2 types of responders because few participants (n = 15) showed an early but unsustained response. The 2 groups that composed the suboptimal response categories were immunologically equivalent as measured by CD4+ lymphocyte count over the study's time course (data not shown). Analyses were performed using both categorization schemes. Results differed marginally between the 2 analyses. Hence, we present the findings of the dichotomous categorization only.

Statistical analysis

All analyses were performed using SAS statistical software version 9.1 (SAS Institute, Inc, Cary, North Carolina). Differences in proportions of participant characteristics between the 2 viral response groups were assessed by χ2 test. Cochran-Mantel-Haenszel χ2 was used for ordinal categories. After determining normality assumptions to be robust, mean group differences of baseline CD4+ lymphocyte count and logarithmically transformed viral load were compared using Wilcoxon rank sum test and t test, respectively. Multivariable logistic regression models were used to examine the relationship between viral response and baseline age, race/ethnicity, and all dichotomous variables. Magnitudes of association are reported as adjusted odds ratios and 95% confidence intervals for each covariate in the model.

Mean viral loads and CD4+ lymphocyte counts were calculated for pre- and post-HAART visits across the 2 responder groups. The pre- to post-HAART differences between the 2 groups were compared using t test. Changes in viral load across all time points were examined using longitudinal data analytic methods. Specifically, we estimated changes in viral load within participants and between responder groups using a mixed linear model for repeated measures. Although we included multiple covariates in our model, we did not have adequate statistical power to analyze outcomes by type of first- or second-line antiretroviral regimens. Model-selection strategies were used and baseline age, sex, and race/ethnicity were adjusted for potential confounding. Variance-covariance structure and model fit diagnostics were assessed for each model.


Of the 154 adolescents included in the study, most were black (72.7%), female (73.4%), and between the ages of 17 and 19 years (70.8%) at enrollment (Table 1). Nearly half of the participants (49.4%) were depressed as assessed by the Center for Epidemiologic Study Depression Scale. More than two-thirds of the adolescents (68.4%) reported ever having used alcohol. Of those, 83 (79.8%) reported alcohol use in the 3 months prior to interview. Illicit drug use was even more prevalent (83.6%). Seventy-three adolescents (47.4%) had a history of using single or dual ART before commencing HAART. More than 100 participants (70.8%) met our definition of medication adherence during the course of follow-up.

Table 1. 
Baseline Characteristics of 154 Human Immunodeficiency Virus–Infected Adolescents According to Virologic Response to Highly Active Antiretroviral Therapy
Baseline Characteristics of 154 Human Immunodeficiency Virus–Infected Adolescents According to Virologic Response to Highly Active Antiretroviral Therapy

According to the Centers for Disease Control and Prevention HIV staging system,27 58 adolescents (37.7%) met clinical criteria for early HIV disease, 65 (42.2%) were at the intermediate stage of infection, and 31 (20.1%) were at the late stage of their disease course. The mean (SD) baseline viral load of the study population was 3.93 (0.88) log10 copies/mL, while the mean (SD) CD4+ lymphocyte count was 495 (248) cells/μL. Fifty adolescents (32.5%) demonstrated early and sustained viral suppression (optimal responders), while 15 (9.7%) had an early but unsustained viral load reduction and 89 (57.8%) showed little or no viral response at all while receiving HAART. The latter 2 groups composed the suboptimal responders.

When we compared individuals according to virologic response to HAART, we found no differences in age, sex, race, education, evidence of depression, or history of drug use. Clinical stage of disease progression, immunogenetic profile, baseline CD4+ lymphocyte count, and viral load values were also similar across groups. However, optimal responders had a higher proportion of individuals who reported ever using alcohol than suboptimal responders (80.0% vs 61.5%, respectively; P = .03). Adherence levels and prior use of single or dual ART also differed by virologic response. Optimal responders were less likely to have had a history of ART use than suboptimal responders (30.0% vs 55.8%, respectively; P = .003). Also, optimal responders had been significantly more adherent to their HAART regimen compared with suboptimal responders (82.0% vs 65.3%, respectively; P = .03).

Despite statistically significant differences between groups on 3 separate parameters, only prior ART use was still significantly associated with suboptimal virologic response on multivariable logistic regression modeling (adjusted odds ratio = 2.6; 95% confidence interval, 1.1-5.7) (Table 2). There was also a strong trend toward protection from suboptimal virologic response with adequate HAART adherence (adjusted odds ratio = 0.4; 95% confidence interval, 0.2-1.0). Comparisons between HAART response groups revealed that the mean pre-HAART plasma concentration of HIV RNA was 0.36 log10 copies/mL higher in the optimal responders (mean [SD], 4.13 [0.76] log10 copies/mL) compared with suboptimal responders (mean [SD], 3.77 [0.90] log10 copies/mL) (P < .001) (Table 3). At post-HAART visits, optimal responders had a 7-fold viral load decrement compared with suboptimal responders (mean [SD], −2.11 [0.68] vs −0.29 [0.93] log10 copies/mL; P < .001) (Figure). Also, the optimal responders maintained a plasma HIV RNA concentration at least 1.0 log10 copies/mL lower than suboptimal responders on all subsequent post-HAART initiation visits except the last (P < .001). After the baseline visit at which the optimal responders had a lower CD4+ lymphocyte count than the suboptimal responders (458 cells/μL vs 503 cells/μL, respectively; P = .03), the optimal responders consistently exhibited significantly higher CD4+ lymphocyte counts across all remaining visits (P < .001).

Pre– and post–highly active antiretroviral therapy (HAART) viral load (A) and CD4+ lymphocyte count (B) according to virologic response.

Pre– and post–highly active antiretroviral therapy (HAART) viral load (A) and CD4+ lymphocyte count (B) according to virologic response.

Table 2. 
Multivariable-Adjusted Likelihood of Suboptimal Virologic Response to Highly Active Antiretroviral Therapy in Association With Sociodemographic Characteristics Among 154 Human Immunodeficiency Virus–Infected Adolescentsa
Multivariable-Adjusted Likelihood of Suboptimal Virologic Response to Highly Active Antiretroviral Therapy in Association With Sociodemographic Characteristics Among 154 Human Immunodeficiency Virus–Infected Adolescentsa
Table 3. 
Mixed Model of Plasma Concentrations of Human Immunodeficiency Virus RNA and CD4+ Lymphocyte Counts Among Adolescent Optimal Responders and Suboptimal Responders to Highly Active Antiretroviral Therapy Before and After Highly Active Antiretroviral Therapy Initiationa
Mixed Model of Plasma Concentrations of Human Immunodeficiency Virus RNA and CD4+ Lymphocyte Counts Among Adolescent Optimal Responders and Suboptimal Responders to Highly Active Antiretroviral Therapy Before and After Highly Active Antiretroviral Therapy Initiationa


Data on the predictors of clinical outcomes in HIV-infected adolescents are sparse and primarily limited to long-term survivors who had perinatal transmission.28,29 Because adolescents with HIV infection may not know their status and/or may not be in clinical care, it is difficult for any single site outside of developing nations30-32 to have enough HIV-infected adolescents to tease out the factors salient to clinical failure. The REACH Project made it possible for us to overcome these limitations by combining 16 sites from 13 diverse locales in US cities for a prospective care-based epidemiologic study.

We found that adolescents with a history of ART use were more than twice as likely to have a suboptimal virologic response to HAART than if they had been completely ART naive. These outcomes are consistent with previous studies in both adults and children. Use of single or dual therapy is inherently suboptimal and precipitates the emergence of resistant HIV strains. We acknowledge the possibility that adolescents who had received prior ART may have been nonadherent with their medication regimen as well. We found, however, that adolescents with at least 50% adherence to their HAART regimen through at least 50% of their study visits were 60% less likely to exhibit a poor virologic response. That even modest ART adherence should have had such a large protective effect is encouraging. Still, moderate adherence poses a high risk for emergence of viral resistance as it is well established that the lower the proportion of adherent participants is, the higher the mean viral load is.33,34 Most studies have indicated that exceptionally high adherence (>95%) is needed to maintain adequate virologic suppression.35-37 These results are therefore consistent with findings from studies in both adults and children showing that HAART adherence is essential to achieve sustained suppression of HIV replication and immunologic reconstitution, although an elevated risk of developing resistant viral strains remains.38-40

Despite the consistency of our results with prior studies in adult and pediatric populations, our study had several limitations. Adherence data that are based on self-report are subject to imprecision and recall bias. To reduce such bias, REACH Project study interviews were conducted in a uniform fashion, making a differential bias between the 2 outcome groups less likely. We also used less rigorous criteria for the definition of adherence because a stricter definition of adherence (>90% or >95%) was rare in this population and would have reduced our statistical power considerably. In addition, we assessed adherence to HAART only; adherence to any prior antiretroviral use was not considered. Compared with other adult cohorts, our study's sample size was much smaller. However, the current study is one of the largest analyses of an HIV-infected adolescent population, drawing on a geographically diverse set of sites across the United States. Finally, owing to cost restraints, we were not able to perform resistance genotyping or phenotyping on study participants.

Youth with HIV infection have many special needs in coping with their disease.41 While we found that predictors of virologic failure in adolescents are similar to those in children and adults, adolescent psychosocial issues must be addressed to ensure clinical success. Many adolescents live at home and may not reveal their status to parents and guardians, who are then unable to provide assistance in adhering to therapy.42,43 Other youth not living with parents or guardians may have transient housing arrangements or may be living on the street, which complicates adequate adherence as housing, food, and shelter often take precedence over medication adherence.44,45 Furthermore, even when treatment is sought, additional barriers related to availability, acceptability, and equity of services make adherence difficult.

Adherence to HAART will require youth-friendly clinical settings and social support through community outreach efforts.9,46 The adolescent population is in a unique transition state from childhood to adulthood; this requires tailored services to ensure proper adherence. Our data add support for increased focus on adherence support services for HIV-infected adolescents, particularly as fully two-thirds of youth failed to have an adequate virologic response to HAART. The strong association observed in this study between better adherence and both early and sustained virologic responses among adolescents suggests an urgent need to maximize supportive adherence programs in the context of the broader social needs of HIV-infected youth.

Correspondence: Kayvon Modjarrad, MD, PhD, Institute for Global Health, Vanderbilt University School of Medicine, 2215 Garland Dr (319 Light Hall), Nashville, TN 37232-0242 (

Accepted for Publication: May 19, 2009.

Author Contributions: Dr Ding had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Ding and Vermund. Acquisition of data: Ding, Wilson, Tang, and Vermund. Analysis and interpretation of data: Ding, Wilson, Modjarrad, McGwin, Tang and Vermund. Drafting of the manuscript: Ding, Wilson, Modjarrad, McGwin, Tang, and Vermund. Critical revision of the manuscript for important intellectual content: Ding, Modjarrad, McGwin, Tang, and Vermund. Statistical analysis: Ding, Modjarrad, and McGwin. Administrative, technical, and material support: Modjarrad and Tang. Study supervision: Modjarrad, Tang, and Vermund.

Financial Disclosure: None reported.

Funding/Support: The REACH Project study was funded by grant U01-HD32842 from the National Institute of Child Health and Human Development and by the National Institute on Drug Abuse, the National Institute of Allergy and Infectious Diseases, and the National Institute of Mental Health. The analytic work was supported by the Vanderbilt-Meharry Center for AIDS Research and by grant P30-AI054999 from the National Institute of Allergy and Infectious Diseases.

Additional Information: Participating institutions of the REACH Project of the Adolescent Medicine HIV/AIDS Research Network are as follows: University of Miami, Miami, Florida; Children's Hospital of Philadelphia, Philadelphia, Pennsylvania; Tulane Medical Center, New Orleans, Louisiana; Children's National Medical Center, Washington, DC; Montefiore Medical Center, Bronx, New York; University of Maryland, Baltimore; Children's Hospital of Los Angeles, Los Angeles, California; Cook County Hospital and University of Chicago, Chicago, Illinois; Mt Sinai Medical Center, New York, New York; Alabama Children's Hospital, Birmingham; Emory University, Atlanta, Georgia; St Jude Children's Research Hospital, Memphis, Tennessee; SUNY Health Science Center at Brooklyn, Brooklyn, New York; Children's Diagnostic and Treatment Center, University of Puerto Rico, San Juan, Puerto Rico; and University of Medicine and Dentistry of New Jersey, Newark.

Additional Contributions: Katherine L. Allen, MSc, provided assistance with manuscript development. We acknowledge the contributions of investigators and staff (listed in J Adolesc Health. 2001;29[suppl]:5-6) of the Adolescent Medicine HIV/AIDS Research Network (1994-2001, then superseded by the Adolescent Trials Network) and the youth who participated in the research.

Centers for Disease Control and Prevention, HIV/AIDS Surveillance Report, 2006.  Atlanta, GA Centers for Disease Control and Prevention2008;
Rosenberg  PSBiggar  RJ Trends in HIV incidence among young adults in the United States.  JAMA 1998;279 (23) 1894- 1899PubMedGoogle Scholar
Hall  HISong  RRhodes  P  et al. HIV Incidence Surveillance Group, Estimation of HIV incidence in the United States.  JAMA 2008;300 (5) 520- 529PubMedGoogle Scholar
Rangel  MCGavin  LReed  CFowler  MGLee  LM Epidemiology of HIV and AIDS among adolescents and young adults in the United States.  J Adolesc Health 2006;39 (2) 156- 163PubMedGoogle Scholar
Murphy  DAMitchell  RVermund  SHFutterman  DAdolescent Medicine HIV/AIDS Research Network, Factors associated with HIV testing among HIV-positive and HIV-negative high-risk adolescents: the REACH Study. Reaching for Excellence in Adolescent Care and Health.  Pediatrics 2002;110 (3) e36Http://Pediatrics.Aappublications.Org/Cgi/Content/Full/110/3/E36. Accessed December 15, 2008PubMedGoogle Scholar
Obermeyer  CMOsborn  M The utilization of testing and counseling for HIV: a review of the social and behavioral evidence.  Am J Public Health 2007;97 (10) 1762- 1774PubMedGoogle Scholar
Vermund  SHWilson  CM Barriers to HIV testing: where next?  Lancet 2002;360 (9341) 1186- 1187PubMedGoogle Scholar
Johnson  RLMartinez  JBotwinick  G  et al.  Introduction: what youth need: adapting HIV care models to meet the lifestyles and special needs of adolescents and young adults.  J Adolesc Health 2003;33 (2) ((suppl)) 4- 9PubMedGoogle Scholar
Tylee  AHaller  DMGraham  TChurchill  RSanci  LA Youth-friendly primary-care services: how are we doing and what more needs to be done?  Lancet 2007;369 (9572) 1565- 1573PubMedGoogle Scholar
Johnson  RLBotwinick  GSell  RL  et al.  The utilization of treatment and case management services by HIV-infected youth.  J Adolesc Health 2003;33 (2) ((suppl)) 31- 38PubMedGoogle Scholar
Murphy  DASarr  MDurako  SJMoscicki  ABWilson  CMMuenz  LRAdolescent Medicine HIV/AIDS Research Network, Barriers to HAART adherence among human immunodeficiency virus–infected adolescents.  Arch Pediatr Adolesc Med 2003;157 (3) 249- 255PubMedGoogle Scholar
De Rossi  A  Primary HIV infection in infants: impact of highly active antiretroviral therapy on the natural course.  J Biol Regul Homeost Agents 2002;16 (1) 53- 57PubMedGoogle Scholar
Detels  RMuñoz  A McFarlane  G  et al. Multicenter AIDS Cohort Study Investigators, Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration.  JAMA 1998;280 (17) 1497- 1503PubMedGoogle Scholar
Erb  PBattegay  MZimmerli  WRickenbach  MEgger  M Effect of antiretroviral therapy on viral load, CD4 cell count, and progression to acquired immunodeficiency syndrome in a community human immunodeficiency virus-infected cohort: Swiss HIV Cohort Study.  Arch Intern Med 2000;160 (8) 1134- 1140PubMedGoogle Scholar
Hunt  PWDeeks  SGRodriguez  B  et al.  Continued CD4 cell count increases in HIV-infected adults experiencing 4 years of viral suppression on antiretroviral therapy.  AIDS 2003;17 (13) 1907- 1915PubMedGoogle Scholar
Press  NTyndall  MWWood  EHogg  RSMontaner  JS Virologic and immunologic response, clinical progression, and highly active antiretroviral therapy adherence.  J Acquir Immune Defic Syndr 2002;31 ((suppl 3)) S112- S117PubMedGoogle Scholar
Barreiro  PMDona  MCCastilla  JSoriano  V Patterns of response (CD4 cell count and viral load) at 6 months in HIV-infected patients on highly active antiretroviral therapy.  AIDS 1999;13 (4) 525- 526PubMedGoogle Scholar
Florence  ELundgren  JDreezen  C  et al. EuroSIDA Study Group, Factors associated with a reduced CD4 lymphocyte count response to HAART despite full viral suppression in the EuroSIDA study.  HIV Med 2003;4 (3) 255- 262PubMedGoogle Scholar
Jevtović  DSalemović  DRanin  JPesić  IZerjav  SDjurković-Djaković  O The dissociation between virological and immunological responses to HAART.  Biomed Pharmacother 2005;59 (8) 446- 451PubMedGoogle Scholar
Le Moing  VThiebaut  RChene  G  et al. ANRS CO8 (APROCO/COPILOTE) Study Group, Long-term evolution of CD4 count in patients with a plasma HIV RNA persistently <500 copies/mL during treatment with antiretroviral drugs.  HIV Med 2007;8 (3) 156- 163PubMedGoogle Scholar
Rogers  ASFutterman  DKMoscicki  ABWilson  CMEllenberg  JVermund  SH The REACH Project of the Adolescent Medicine HIV/AIDS Research Network: design, methods, and selected characteristics of participants.  J Adolesc Health 1998;22 (4) 300- 311PubMedGoogle Scholar
Wilson  CMHouser  JPartlow  CRudy  BJFutterman  DCFriedman  LBAdolescent Medicine HIV/AIDS Research Network, The REACH (Reaching for Excellence in Adolescent Care and Health) project: study design, methods, and population profile.  J Adolesc Health 2001;29 (3) ((suppl)) 8- 18PubMedGoogle Scholar
Radloff  LS The CES-D Scale: a self-report depression scale for research in the general population.  Appl Psychol Meas 1977;1 (3) 385- 40110.1177/014662167700100306Google Scholar
Murphy  DAWilson  CMDurako  SJMuenz  LRBelzer  MAdolescent Medicine HIV/AIDS Research Network, Antiretroviral medication adherence among the REACH HIV-infected adolescent cohort in the USA.  AIDS Care 2001;13 (1) 27- 40PubMedGoogle Scholar
Reisner  SLMimiaga  MJSkeer  MPerkovich  BJohnson  CVSafren  SA A review of HIV antiretroviral adherence and intervention studies among HIV-infected youth.  Top HIV Med 2009;17 (1) 14- 25PubMedGoogle Scholar
Tang  JWilson  CMMeleth  S  et al. REACH Study Group, Host genetic profiles predict virological and immunological control of HIV-1 infection in adolescents.  AIDS 2002;16 (17) 2275- 2284PubMedGoogle Scholar
 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults.  MMWR Recomm Rep 1992;41 (RR-17) 1- 19PubMedGoogle Scholar
Flynn  PMRudy  BJLindsey  JC  et al. PACTG 381 Study Team, Long-term observation of adolescents initiating HAART therapy: three-year follow-up.  AIDS Res Hum Retroviruses 2007;23 (10) 1208- 1214PubMedGoogle Scholar
Rudy  BJLindsey  JCFlynn  PM  et al. Pediatric Aids Clinical Trials Group 381 Study Team, Immune reconstitution and predictors of virologic failure in adolescents infected through risk behaviors and initiating HAART: week 60 results from the PACTG 381 cohort.  AIDS Res Hum Retroviruses 2006;22 (3) 213- 221PubMedGoogle Scholar
Ferris  MBurau  KConstantin  AM  et al.  Influence of institutionalization on time to HIV disease progression in a cohort of Romanian children and teens.  Pediatrics 2007;120 (6) e1476- e1480PubMedGoogle Scholar
Gouws  EStanecki  KALyerla  RGhys  PD The epidemiology of HIV infection among young people aged 15-24 years in southern Africa.  AIDS 2008;22 ( Suppl 4) S5- 16PubMedGoogle Scholar
Kabue  MMKekitiinwa  AMaganda  ARisser  JMChan  WKline  MW Growth in HIV-infected children receiving antiretroviral therapy at a pediatric infectious diseases clinic in Uganda.  AIDS Patient Care STDS 2008;22 (3) 245- 251PubMedGoogle Scholar
Braithwaite  RSKozal  MJChang  CC  et al.  Adherence, virological and immunological outcomes for HIV-infected veterans starting combination antiretroviral therapies.  AIDS 2007;21 (12) 1579- 1589PubMedGoogle Scholar
Gandhi  MAmeli  NBacchetti  P  et al. Women's Interagency HIV Study (WIHS), Protease inhibitor levels in hair strongly predict virologic response to treatment.  AIDS 2009;23 (4) 471- 478PubMedGoogle Scholar
Boden  DHurley  AZhang  L  et al.  HIV-1 drug resistance in newly infected individuals.  JAMA 1999;282 (12) 1135- 1141PubMedGoogle Scholar
Chesney  M Adherence to HAART regimens.  AIDS Patient Care STDS 2003;17 (4) 169- 177PubMedGoogle Scholar
García de Olalla  PKnobel  HCarmona  AGuelar  ALópez-Colomés  JLCaylà  JA Impact of adherence and highly active antiretroviral therapy on survival in HIV-infected patients.  J Acquir Immune Defic Syndr 2002;30 (1) 105- 110PubMedGoogle Scholar
Gifford  ALBormann  JEShively  MJWright  BCRichman  DDBozzette  SA Predictors of self-reported adherence and plasma HIV concentrations in patients on multidrug antiretroviral regimens.  J Acquir Immune Defic Syndr 2000;23 (5) 386- 395PubMedGoogle Scholar
Gross  RBilker  WBFriedman  HMStrom  BL Effect of adherence to newly initiated antiretroviral therapy on plasma viral load.  AIDS 2001;15 (16) 2109- 2117PubMedGoogle Scholar
Nachega  JBHislop  MDowdy  DWChaisson  RERegensberg  LMaartens  G Adherence to nonnucleoside reverse transcriptase inhibitor–based HIV therapy and virologic outcomes.  Ann Intern Med 2007;146 (8) 564- 573PubMedGoogle Scholar
Hein  KFutterman  D Guidelines for the care of children and adolescents with HIV infection: medical management in HIV-infected adolescents.  J Pediatr 1991;119 (1, pt 2) S18- S20PubMedGoogle Scholar
D'Angelo  LJAbdalian  SESarr  MHoffman  NBelzer  MAdolescent Medicine HIV/AIDS Research Network, Disclosure of serostatus by HIV infected youth: the experience of the REACH study: Reaching for Excellence in Adolescent Care and Health.  J Adolesc Health 2001;29 (3) ((suppl)) 72- 79PubMedGoogle Scholar
Tornero  CSantamaria  AGil  E The presence of a caregiving relative implicated in the administration of antiretroviral drug administration improves virologic response.  J Acquir Immune Defic Syndr 2007;46 (5) 658PubMedGoogle Scholar
Aidala  AALee  GAbramson  DMMesseri  PSiegler  A Housing need, housing assistance, and connection to HIV medical care.  AIDS Behav 2007;11 (6) ((suppl)) 101- 115PubMedGoogle Scholar
Martinez  JBell  DDodds  S  et al.  Transitioning youths into care: linking identified HIV-infected youth at outreach sites in the community to hospital-based clinics and or community-based health centers.  J Adolesc Health 2003;33 (2) ((suppl)) 23- 30PubMedGoogle Scholar
Futterman  DC HIV and AIDS in adolescents.  Adolesc Med Clin 2004;15 (2) 369- 391PubMedGoogle Scholar