Random weighted bootstrap method for recurrent events with informative censoring

Using the data from the AIDS Link to Intravenous Experiences cohort study as an example, an informative censoring model was used to characterize the repeated hospitalization process of a group of patients. Under the informative censoring assumption, the estimators of the baseline rate function and the regression parameters were shown to be related to a latent variable. Hence, it becomes impractical to directly estimate the unknown quantities in the moments of the estimators for the bandwidth selection of a smoothing estimator and the construction of confidence intervals, which are respectively based on the asymptotic mean squared errors and the asymptotic distributions of the estimators. To overcome these difficulties, we develop a random weighted bootstrap procedure to select appropriate bandwidths and to construct approximated confidence intervals. One can see that our method is simple and faster to implement from a practical point of view, and is at least as accurate as other bootstrap methods. In this article, it is shown that the proposed method is useful through the performance of a Monte Carlo simulation. An application of our procedure is also illustrated by a recurrent event sample of intravenous drug users for inpatient cares over time.

Mannose binding lectin genotypes influence recovery from hepatitis B virus infection

Mannose binding lectin (MBL) is a central component of the innate immune response and thus may be important for determining hepatitis B virus (HBV) persistence. Since single-nucleotide polymorphisms (SNPs) in the gene encoding MBL (mbl2) alter the level of functional MBL, we hypothesized that mbl2 genotypes are a determinant of HBV persistence or recovery from viral infection. We tested this hypothesis by using a nested case control design with 189 persons with HBV persistence matched to 338 individuals who had naturally recovered from HBV infection. We determined genotypes of two promoter and three exon 1 SNPs in mbl2 and grouped these genotypes according to the amount of functional MBL production. We found that the promoter SNP -221C, which leads to deficient MBL production, was more common in those subjects with viral persistence (odds ratio [OR], 1.38; 95% confidence interval [CI], 1.01 to 1.89; P = 0.04). Those subjects homozygous for the combination of promoter and exon 1 genotypes associated with the highest amount of functional MBL had significantly increased odds of recovery from infection (OR, 0.55; 95% CI, 0.37 to 0.84; P = 0.005). Conversely, those homozygous for the combination of promoter and exon 1 genotypes which produce the lowest amount of functional MBL were more likely to have viral persistence (OR, 1.76; 95% CI, 1.02 to 3.01; P = 0.04). These data are consistent with the hypothesis that functional MBL plays a central role in the pathogenesis of acute hepatitis B.

Kernel Estimation of Rate Function for Recurrent Event Data

Recurrent event data are largely characterized by the rate function but smoothing techniques for estimating the rate function have never been rigorously developed or studied in statistical literature. This paper considers the moment and least squares methods for estimating the rate function from recurrent event data. With an independent censoring assumption on the recurrent event process, we study statistical properties of the proposed estimators and propose bootstrap procedures for the bandwidth selection and for the approximation of confidence intervals in the estimation of the occurrence rate function. It is identified that the moment method without resmoothing via a smaller bandwidth will produce a curve with nicks occurring at the censoring times, whereas there is no such problem with the least squares method. Furthermore, the asymptotic variance of the least squares estimator is shown to be smaller under regularity conditions. However, in the implementation of the bootstrap procedures, the moment method is computationally more efficient than the least squares method because the former approach uses condensed bootstrap data. The performance of the proposed procedures is studied through Monte Carlo simulations and an epidemiological example on intravenous drug users.

MICA and recovery from hepatitis C virus and hepatitis B virus infections

The polymorphic MHC class I chain-related A (MICA) gene encodes a ligand that has different binding affinities for the NKG2D activating receptor of CD8+ T cells and natural killer (NK) cells. We hypothesized that MICA heterogeneity would affect recovery from hepatitis C virus (HCV) and hepatitis B virus (HBV) infections. To test the hypothesis, we initially typed known MICA polymorphisms for 228 persons who cleared HCV infection and 442 persons with persistent hepatitis C matched on other factors affecting viral persistence. Although MICA(*)015 was detected more than two-fold more often in persons with viral clearance (odds ratio 0.36, 95% confidence interval=0.19, 0.80), it occurred in fewer than 5% of the study population. In a similar analysis of 442 persons with chronic hepatitis B and 768 matched controls who recovered, MICA(*)015 was detected in 2.0% of persons with chronic hepatitis B and only 0.9% of controls. No significant associations were detected with other MICA polymorphisms. While further investigation may reveal a structural basis of the MICA(*)015 associations, these data provide little support for the hypothesis that differential distribution of MICA alleles substantially affects recovery from HCV and HBV infections.

Body habitus in a cohort of HIV-seropositive and HIV-seronegative injection drug users

We determined anthropometric measurements (including height, weight, circumferences, and skinfolds) and self-reported symptoms related to body habitus changes in 324 HIV-seropositive and HIV-seronegative inner city injection drug users (IDUs) who participated in a substudy from the ALIVE (AIDS Linked to Intravenous Experiences) cohort. Participants who reported lipoatrophy in body parts had consistently lower anthropometric measurements and those reporting adiposity had correspondingly higher anthropometric measurements than participants who did not report these changes. Peripheral lipoatrophy was more common among all HIV-seropositive than HIV-seronegative participants, however, it was not associated with highly active antiretroviral therapy (HAART) (39% HIV-seronegatives; 58% HIV-seropositive not receiving HIV treatment [No Tx]; 49% HAART, p = 0.04). Central adiposity was more common among HAART (52%) than No Tx (26.6%) and HIV-seronegative (42%) participants (p = 0.001). However, waist circumference, while somewhat higher among HAART than No Tx participants, did not differ significantly from HIV-seronegative participants (85.2 cm HIV-seronegatives; 83.3 cm No Tx; 85.8 cm HAART). A large proportion of those who reported peripheral lipoatrophy also reported central lipoatrophy (76.9% HIV-seronegatives; 69.6% No Tx; 66.2% HAART). A large proportion of those who reported central adiposity also reported adiposity of the peripheral sites (88.1% HIV seronegatives; 66.7% No Tx; 74.3% HAART). The combination of lipoatrophy and adiposity was associated with HAART treatment (6% HIV-seronegatives; 3% No Tx; 16% HAART, p = 0.002), but may be driven by the association with adiposity. These data suggest validity of self-reports for body habitus changes among injection drug users.

Cytotoxic T-lymphocyte antigen 4 gene and recovery from hepatitis B virus infection

Cytotoxic T-lymphocyte antigen 4 (CTLA-4) is an inhibitory T-cell receptor expressed by activated and regulatory T cells. We hypothesized that single-nucleotide polymorphisms (SNPs) in the gene encoding CTLA-4 may affect the vigor of the T-cell response to hepatitis B virus (HBV) infection, thus influencing viral persistence. To test this hypothesis, we genotyped six CTLA4 SNPs, from which all frequent haplotypes can be determined, using a large, matched panel of subjects with known HBV outcomes. Haplotypes with these SNPs were constructed for each subject using PHASE software. The haplotype distribution differed between those with viral persistence and those with clearance. Two haplotypes were associated with clearance of HBV infection, which was most likely due to associations with the SNPs -1722C (odds ratio [OR] = 0.60, P = 0.06) and +49G (OR = 0.73, P = 0.02). The wild-type haplotype, which contains an SNP leading to a decreased T-cell response (+6230A), was associated with viral persistence (OR = 1.32, P = 0.04). These data suggest that CTLA4 influences recovery from HBV infection, which is consistent with the emerging role of T regulatory cells in the pathogenesis of disease.

Epistatic effects of immunoglobulin GM and KM allotypes on outcome of infection with hepatitis C virus

Immunoglobulin GM and KM allotypes-genetic markers of gamma and kappa chains, respectively-are associated with immune responsiveness to several infectious pathogens and with survival in certain viral epidemics. We hypothesized that GM and KM allotypes affect the outcome of hepatitis C virus (HCV) infection. To test this hypothesis, we serologically allotyped 100 persons with well-documented clearance of HCV infection and 198 matched persistently infected persons. None of the GM or KM phenotypes by itself was associated with the clearance or persistence of HCV infection. Particular combinations of these phenotypes, however, were significantly associated with the outcome of HCV infection. Subjects with GM 1,17 5,13 and KM 1,3 phenotypes were over three times (odds ratio [OR], 3.57; 95% confidence interval [CI], 1.44 to 8.87) as likely to clear the infection as the subjects who lacked these phenotypes. This GM phenotype had a similar association with clearance in the absence of KM 3 (OR, 2.75; 95% CI, 1.21 to 6.23). The presence of GM 1,3,17 23 5,13 phenotype (in the absence of KM 3) was associated with persistence (OR, 0.21; 95% CI, 0.06 to 0.77), while its absence (in the presence of KM 1,3) was associated with the clearance of infection (OR, 2.03; 95% CI, 1.16 to 3.54). These results show epistatic interactions of genes on chromosomes 14 (GM) and 2 (KM) in influencing the outcome of an HCV infection. Further investigations involving candidate genes (GM, KM, HLA, and Fcgamma receptors) and cellular and humoral immune responses to HCV epitopes are needed to understand the mechanisms underlying these associations.

Prevalence and outcomes of anemia in individuals with human immunodeficiency virus: a systematic review of the literature

In patients with human immunodeficiency virus (HIV), anemia is a commonly encountered hematologic abnormality that has a significant impact on clinical outcomes and quality of life (QOL). This review describes the prevalence of anemia in several populations of patients with HIV and the effects of anemia on survival, morbidity, disease progression, transfusion requirements, and QOL. The prevalence of anemia in HIV disease varies considerably, ranging from 1.3% to 95%: it depends on several factors, including the stage of HIV disease, sex, age, pregnancy status, and injection-drug use as well as the definition of anemia used. In general, as HIV disease progresses, the prevalence and severity of anemia increase. Anemia is also more prevalent in HIV-positive women, children, and injection-drug users than in HIV-negative women, children, and injection-drug users. Anemia has been shown to be a statistically significant predictor of progression to the acquired immunodeficiency syndrome and is independently associated with an increased risk of death in patients with HIV. Treatment of anemia with epoetin-alpha has resulted in significantly fewer patients requiring transfusion as well as decreases in the mean number of units of blood transfused. Resolution of HIV-related anemia has been shown to improve QOL, physical functioning, energy, and fatigue in individuals with HIV. More recently, the use of highly active antiretroviral therapy has also been associated with a significant increase in hemoglobin concentrations and a decrease in the prevalence of anemia.

Semiparametric Analysis of Recurrent Events Data in the Presence of Dependent Censoring

Dependent censoring occurs in longitudinal studies of recurrent events when the censoring time depends on the potentially unobserved recurrent event times. To perform regression analysis in this setting, we propose a semiparametric joint model that formulates the marginal distributions of the recurrent event process and dependent censoring time through scale-change models, while leaving the distributional form and dependence structure unspecified. We derive consistent and asymptotically normal estimators for the regression parameters. We also develop graphical and numerical methods for assessing the adequacy of the proposed model. The finite-sample behavior of the new inference procedures is evaluated through simulation studies. An application to recurrent hospitalization data taken from a study of intravenous drug users is provided.

Clinical and immunologic progression in HIV-infected US women before and after the introduction of highly active antiretroviral therapy

OBJECTIVE

To examine factors associated with clinical and immunologic HIV disease progression in a cohort of US women.

DESIGN

Analysis of data from a prospective, longitudinal, case-control study of HIV-infected women followed every 6 months for 7 years.

SETTING

Four urban clinical centers in the United States.

PARTICIPANTS

648 HIV-infected women who did not have AIDS at time of entry into the study.

MEASUREMENTS

Structured clinical and behavioral interviews; protocol-directed physical examinations; CD4 lymphocyte counts; plasma HIV RNA; infectious pathogen serologies. RESULTS With 2304 women-years of follow-up, 46.1% of the women developed AIDS; however, 93.3% of the diagnoses were based on CD4 counts dropping to <200 cells/mm(3). Only 10.6% of the women with CD4 counts <200 cells/mm(3) developed an opportunistic infection. Baseline CD4 count was the strongest predictor of subsequent clinical progression. Illicit substance use, multiple pregnancies, demographic variables, and other infections were not associated with progression. Among women with CD4 counts >500 cells/mm(3) at baseline, those who were anemic or had hepatitis C were more likely to progress to AIDS. By the end of the study, only 52% of the participants were on highly active antiretroviral therapy (HAART).

CONCLUSIONS

Despite underutilization of HAART in this multicenter cohort of urban women, opportunistic infections were uncommon, despite CD4 declines.

Children of HIV-positive drug-using parents

OBJECTIVE

Associations between human immunodeficiency virus (HIV) status of injection drug users (IDUs) and their children's psychopathology and social functioning were examined.

METHOD

Parents (N = 61) were drawn from an ongoing longitudinal study of inner city, primarily African-American IDUs. Children (N = 79) were 6 to 11 years of age, currently living with the IDU parent. Parental variables included HIV status, apparent and inapparent HIV infection (with HIV-related medical symptoms and/or disclosure of parental HIV status to children; with neither medical symptoms nor disclosure, respectively), presence of HIV-related medical symptoms, HIV disclosure status to each child, and depression as ascertained by the Center for Epidemiologic Studies-Depression. Children's outcomes (competencies, psychiatric symptoms, and disorders) were assessed with the Child Behavior Checklist and the Schedule for Affective Disorders and Schizophrenia for School-Age Children.

RESULTS

Children of IDU parents exhibited high rates of psychopathology. Parental HIV infection per se had no discernible impact on children's outcomes. The apparent HIV infection of a parent was associated with an eightfold increase (odds ratio 7.80; 95% confidence interval 1.56-39.09) in the prevalence of disruptive behavior disorders (compared with children of HIV-negative parents). Parental depression was associated with a threefold increase in the prevalence of children's disruptive behavior disorders (odds ratio 3.49, 95% confidence interval 1.11-11.04).

CONCLUSIONS

Parental HIV status per se does not seem to have a differential impact on the affected children. The apparent HIV infection of a parent may be associated with children's externalizing symptoms and disorders.

Psychological distress and progression to AIDS in a cohort of injection drug users

OBJECTIVE

We investigated whether distress was independently associated with more rapid progression to AIDS among HIV-infected injection drug users (IDUs).

DESIGN

A cohort study of IDUs in Baltimore followed from 1988 through 1999. A total of 451 IDUs met the eligibility criteria of being HIV-seropositive but AIDS-free at baseline.

METHODS

We used Cox proportional hazards models to investigate progression to AIDS within 2 years of baseline, controlling for CD4 lymphocyte count, HIV-1 viral load, and oral thrush.

RESULTS

Of the 451 participants, 76.3% were male and 95.8% were African-American; the median age at enrollment was 34 years. The cumulative incidence of AIDS within 2 years of baseline was 7.1%. In multiple regression analysis, distress was associated with more rapid time to AIDS (adjusted relative hazard [RH] = 2.39; 95% CI: 1.16-4.90). The strongest association was observed among IDUs with the lowest (<200 x 10 /L) CD4 counts (adjusted RH = 4.94; 95% CI: 1.30-18.77).

CONCLUSIONS

Psychological distress was independently associated with shorter time to AIDS among HIV-infected IDUs, especially among those with the lowest CD4 cell counts, but was not predictive of mortality in this cohort. Further study of the effects of psychological distress on AIDS progression within this population is warranted.

The role of Medicaid HMO enrollment in the longitudinal utilization of medical care services in a cohort of injecting drug users in Baltimore, Maryland

Although injecting drug users (IDUs) are highly dependent on Medicaid, the literature has not explored the effect of Medicaid HMO enrollment on medical care utilization. This longitudinal analysis compares 6-month utilization levels of IDUs in Medicaid HMOs with those of other Medicaid-enrolled IDUs and IDUs in other insurance arrangements. The analysis includes 1064 participants from the AIDS Link to Intravenous Experience (ALIVE) study, a natural history study of HIV infection in a community-based sample of IDUs in Baltimore, Maryland. A generalized linear modeling approach was used to examine self-reported inpatient admissions, ER visits, and ambulatory care visits from January 1995 through June 1997. Non-HMO Medicaid beneficiaries had more inpatient admissions and ambulatory care visits than Medicaid HMO enrollees. These findings suggest that HMO enrollment has been effective in lowering utilization among IDUs in Medicaid, but it is unclear whether utilization differences are the result of inappropriate utilization or access issues.

Readiness for cessation of drug use among recent attenders and nonattenders of a needle exchange program

Needle exchange programs (NEPs) represent a bridge to drug abuse treatment. NEP attenders tend to have more severe drug problems, however, and may be less ready to reduce their drug use than other drug users. This study investigated the relationship between NEP attendance and readiness for cessation of drug use. Since the period from 1988 through 1989, a community-based sample of injection drug users (IDUs) in Baltimore has undergone semiannual interview-administered questionnaires and HIV testing. A total of 288 IDUs completed a questionnaire on readiness for cessation of drug use. Readiness for drug use cessation was assessed from a 28-item validated scale of problem drug use and intention to quit, based on the "stages of change" model. Logistic regression was used to determine factors associated with readiness for cessation of drug use, including sociodemographics, drug use behaviors, and NEP attendance. Thirty percent of respondents attended the NEP in the past month. Stage of change in readiness for cessation of drug use did not differ between NEP attenders and nonattenders (OR = 0.9; 95% CI: 0.5-1.6). Similar proportions of persons recently attending and not attending the NEP were classified as ready to stop drug use (about 30%). In multivariate analysis, readiness for cessation of drug use was associated with speedball injection and previous enrollment in drug treatment but not with NEP attendance. NEP attenders, although exhibiting characteristics consistent with more severe drug dependence, were as motivated for cessation of drug use as were nonattenders. These findings suggest that formal collaboration between NEPs and drug treatment programs could increase the proportion of IDUs in treatment.

Severity and correlates of liver disease in hepatitis C virus-infected injection drug users

Between May 1996 and June 1998, 210 members of a cohort of 1,667 hepatitis C virus (HCV)-infected injection drug users (IDUs) were selected for liver biopsy procedure after stratification based on 2 consecutive serum alanine transaminase (ALT) levels. Liver histology, which could be fully evaluated for 207 subjects, was classified by using the modified Ishak scores. At the time of biopsy, the median age of subjects was 41.3 years and the median estimated duration of HCV infection was 20.7 years; 94% were African American; 78% men; 31% were human immunodeficiency virus (HIV) seropositive; and 76% had HCV genotype 1a or 1b. Total modified histologic activity index (MHAI) scores ranged from 0 to 9, and 26.6% had a total MHAI score of 5 or greater. Persons with a total MHAI score of 5 or greater were more likely to be HIV infected (P =.04). Higher fibrosis, indicated by Ishak modified fibrosis scores of 3 to 6, was present in 10.1% of subjects and was found more often in those older than 46 years of age (the highest quartile) (P <.01). Both fibrosis scores of 3 or greater and total scores of 5 or greater were associated with elevated ALT, aspartate transaminase (AST), and gamma-glutamyl transpeptidase (GGT) levels (P <.01). When serial values were considered, the results of liver enzyme testing could reduce the probability of an IDU having a fibrosis score of 3 or greater from 10% to 3%. In conclusion, these data indicate that severe liver disease is uncommon in this urban, HCV-infected IDU cohort, especially in younger persons and those with repeatedly normal liver enzymes.

Non-parametric methods for recurrent event data with informative and non-informative censorings

Recurrent event data are commonly encountered in health-related longitudinal studies. In this paper time-to-events models for recurrent event data are studied with non-informative and informative censorings. In statistical literature, the risk set methods have been confirmed to serve as an appropriate and efficient approach for analysing recurrent event data when censoring is non-informative. This approach produces biased results, however, when censoring is informative for the time-to-events outcome data. We compare the risk set methods with alternative non-parametric approaches which are robust subject to informative censoring. In particular, non-parametric procedures for the estimation of the cumulative occurrence rate function (CORF) and the occurrence rate function (ORF) are discussed in detail. Simulation and an analysis of data from the AIDS Link to Intravenous Experiences Cohort Study is presented.

The effects of sponsorship in 12-step treatment of injection drug users

What contributes to sustained abstinence from injection drug use by those who participate in community-based Narcotics Anonymous (NA) and Alcoholics Anonymous (AA) is not well understood. We know that sponsorship is a central element in these programs. To investigate the relationship between sponsorship and abstinence, we evaluated NA/AA sponsorship over a 1-year period in a longitudinal study of 500 former and current injection drug users in inner-city Baltimore recruited from the community-at-large, independent of treatment center affiliation. The findings indicated that having a sponsor in NA/AA for this population was not associated with any improvement in 1-year sustained abstinence rates than a non-sponsored group. However, being a sponsor over the same time period was strongly associated with substantial improvements in sustained abstinence rates for the sponsors, controlling for involvement with community organizations, NA/AA meeting attendance, marital status, employment, participation in drug and alcohol treatment centers and HIV status. Involvement in community organizations was also strongly associated with successful abstinence, controlling for the same variables. Of those participants involved with community organizations, more than half reported involvement in church activities. Our investigation suggests that, for NA/AA sponsors in this study population, providing direction and support to other addicts is associated with improved success in sustained abstinence for the sponsors but does little to improve the short-term success of the persons being sponsored.

Intimate relationship characteristics associated with condom use among drug users and their sex partners: a multilevel analysis

Few studies have examined sexual risk behaviors among drug users in the context of their primary sexual relationships. The aim of this study is to examine characteristics of sexual partners associated with inconsistent condom use among a sample of 703 cocaine and opiate users from Baltimore, MD who participated in a network oriented intervention study of human immunodeficiency syndrome (HIV) risk behaviors among drug users. The current study examines relationship characteristics between drug users and their sexual partners (n=1003). Some of the study participants had more than one sexual partner, accounting for the greater number of sexual pairs than study participants. Study participants reported using multiple drugs (56% injected, 48% sniffed heroin, and 48% smoked crack) and that 70% of their sexual partners also used heroin and/or cocaine. Forty percent reported consistent condom use in the past 3 months. In multiple logistic regression, characteristics associated with consistent condom use were, the drug user being HIV infected; not living with their sexual partners; and not being financially interdependent. Previous research has demonstrated lower rates of condom use with main compared with casual and exchange partners. Primary sexual partners provide a sense of stability and introducing condoms may not be desirable or realistic. HIV prevention programs that target drug users should focus on the benefits, such as trust and a sense of security, and risks of not using condoms in primary relationships. Programs should also focus on enhancing communication and negotiation skills through targeting individuals as well as couples.

Illicit drug use in one's social network and in one's neighborhood predicts individual heroin and cocaine use

PURPOSE

The nature of competing social environmental factors' influence on substance abuse is unclear. A longitudinal study was undertaken to determine the relative power of social network and neighborhood characteristics to predict continuing illicit drug use.

METHODS

Three hundred forty-two inner-city adults with a history of injection drug use were followed for 1 year; their heroin and cocaine use were assessed semiannually. Multiple logistic regression models were fit to determine the degree to which social network and neighborhood characteristics, assessed at baseline, predicted continuing heroin and/or cocaine use throughout the study period.

RESULTS

Two hundred thirty-six (69%) participants reported continuing heroin and/or cocaine use. Drug use by members of the social network was a stronger predictor of participants' continuing drug use (OR = 4.31, 95% CI 2.51 to 7.40) than was a high level of drug-related arrests in the participant's neighborhood (OR = 2.41, 95% CI 1.24 to 4.71), after adjusting for drug treatment and demographic variables. Both seemed to have independent effects on study participants' drug use.

CONCLUSIONS

These findings reiterate the importance of breaking ties with drug-using associates, even for those who reside in high-risk environments. Further work is needed to develop interventions that increase drug users' success in altering social network composition or also treat drug-using network members.

Improvement of anemia among HIV-infected injection drug users receiving highly active antiretroviral therapy

Although anemia is common during HIV infection, it is unclear whether potent antiretroviral therapy would improve or worsen anemia. We conducted a study to examine the impact of highly active antiretroviral therapy (HAART) on anemia in a cohort of HIV-positive injection drug users (IDUs) in Baltimore, Maryland. At baseline, the overall prevalence of anemia was 40%. During mean follow-up of one year, among 102 subjects who received HAART, there was a mean increase in hemoglobin of 3.6 +/- 1.7 g/L (p =.0003) [corrected] and a mean decrease in log(10) plasma HIV load of 0.25 +/- 0.06 copies/ml (p <.0002) [corrected]. Among 103 control subjects who were not receiving antiretroviral medications, there was a mean decrease in hemoglobin of 4.2 +/- 1.1 g/L (p <.04) [corrected] and mean increase in log(10) plasma HIV load of 0.78 +/- 0.17 copies/ml (p <.0001) [corrected]. Multivariate analysis using mixed linear models showed that HAART was associated with an increase of hemoglobin of 0.223 g/L per month (p <.0001) after adjusting for body mass index, opportunistic infections, and gender. HAART was associated with an improvement in anemia, and potential mechanisms that may be involved include a reduction in opportunistic infections and the anemia of chronic disease and an improvement in nutritional status.

Human immunodeficiency virus seroconversion and evolution of the hepatitis C virus quasispecies

When chronic hepatitis C virus (HCV) infections are complicated by acquisition of human immunodeficiency virus (HIV), liver disease appears to accelerate and serum levels of HCV RNA may rise. We hypothesized that HIV might affect the HCV quasispecies by decreasing both complexity (if HIV-induced immunosuppression lessens pressure for selecting HCV substitutions) and the ratio of nonsynonymous (d(N)) to synonymous (d(S)) substitutions, because d(N) may be lower (if there is less selective pressure). To test this hypothesis, we studied the evolution of HCV sequences in 10 persons with chronic HCV infection who seroconverted to HIV and, over the next 3 years, had slow or rapid progression of HIV-associated disease. From each subject, four serum specimens were selected with reference to HIV seroconversion: (i) more than 2 years prior, (ii) less than 2 years prior, (iii) less than 2 years after, and (iv) more than 2 years after. The HCV quasispecies in these specimens was characterized by generating clones containing 1 kb of cDNA that spanned the E1 gene and the E2 hypervariable region 1 (HVR1), followed by analysis of clonal frequencies (via electrophoretic migration) and nucleotide sequences. We examined 1,320 cDNA clones (33 per time point) and 287 sequences (median of 7 per time point). We observed a trend toward lower d(N)/d(S) after HIV seroconversion in 7 of 10 subjects and lower d(N)/d(S) in those with rapid HIV disease progression. However, the magnitude of these differences was small. These results are consistent with the hypothesis that HIV infection alters the HCV quasispecies, but the number of subjects and observation time may be too low to characterize the full effect.

Initial plasma HIV-1 RNA levels and progression to AIDS in women and men

BACKGROUND

It is unclear whether there are differences between men and women with human immunodeficiency virus type 1 (HIV-1) infection in the plasma level of viral RNA (the viral load). In men, the initial viral load after seroconversion predicts the likelihood of progression to the acquired immunodeficiency syndrome (AIDS), but the relation between the two has not been assessed in women. Currently, the guidelines for initiating antiretroviral therapy are applied uniformly to women and men.

METHODS

From 1988 through 1998, the viral load and the CD4+ lymphocyte count were measured approximately every six months in 156 male and 46 female injection-drug users who were followed prospectively after HIV-1 seroconversion.

RESULTS

The median initial viral load was 50,766 copies of HIV-1 RNA per milliliter in the men but only 15,103 copies per milliliter in the women (P<0.001). The median initial CD4+ count did not differ significantly according to sex (659 and 672 cells per cubic millimeter, respectively). HIV-1 infection progressed to AIDS in 29 men and 15 women, and the risk of progression did not differ significantly according to sex. For each increase of 1 log in the viral load (on a base 10 scale), the hazard ratio for progression to AIDS was 1.55 (95 percent confidence interval, 0.97 to 2.47) among the men and 1.43 (95 percent confidence interval, 0.76 to 2.69) among the women. The median initial viral load was 77,822 HIV-1 RNA copies per milliliter in the men in whom AIDS developed and 40,634 copies per milliliter in the men in whom it did not; the corresponding values in the women were 17,149 and 12,043 copies per milliliter. Given the recommendation that treatment should be initiated when the viral load reaches 20,000 copies per milliliter, 74 percent of the men but only 37 percent of the women in our study would have been eligible for therapy at the first visit after seroconversion (P<0.001).

CONCLUSIONS

Although the initial level of HIV-1 RNA was lower in women than in men, the rates of progression to AIDS were similar. Treatment guidelines that are based on the viral load, rather than the CD4+ lymphocyte count, will lead to differences in eligibility for antiretroviral treatment according to sex.

Antipneumolysin antibody titers in HIV-seropositive injection drug users before and after pneumococcal bacteremia

Lower baseline antipneumolysin antibody (alpha-PLY) levels have been found in populations with a higher incidence of pneumococcal infections. To determine whether predisease alpha-PLY titer is associated with invasive pneumococcal disease in HIV-seropositive injection drug users (IDU), we utilized a prospective cohort of IDU in Baltimore to compare alpha-PLY titers before bacteremia in 28 HIV- seropositive IDU cases with alpha-PLY titers in 56 matched (CD4 and seroconversion date) HIV-seropositive IDU control subjects and 28 matched (calendar time) HIV-seronegative IDU control subjects remaining free of pneumococcal disease. We also compared the postinfection fold-rise of alpha-PLY titers in cases relative to the change in alpha-PLY titers in control subjects during the same interval; alpha-PLY titers were measured using quantitative ELISA, and functional activity was assessed using antihemolysin assays. Predisease alpha-PLY titer did not differ between cases (66 units) and HIV-seropositive control subjects (70 units, p = 0.56) or HIV-seronegative control subjects (80 units, p = 0.10). There was a significant difference in fold-rise of alpha-PLY titers postdisease between cases (1.18) and HIV-seronegative control subjects (0.76), p = 0.03. Baseline alpha-PLY titers do not differ significantly between HIV-seropositive IDU who develop pneumococcal bacteremia from HIV-seropositive and HIV-seronegative IDU control subjects remaining free of severe pneumococcal disease.

Analyzing Recurrent Event Data With Informative Censoring

Recurrent event data are frequently encountered in longitudinal follow-up studies. In statistical literature, noninformative censoring is typically assumed when statistical methods and theory are developed for analyzing recurrent event data. In many applications, however, the observation of recurrent events could be terminated by informative dropouts or failure events, and it is unrealistic to assume that the censoring mechanism is independent of the recurrent event process. In this article we consider recurrent events of the same type and allow the censoring mechanism to be possibly informative. The occurrence of recurrent events is modeled by a subject-specific nonstationary Poisson process via a latent variable. A multiplicative intensity model is used as the underlying model for nonparametric estimation of the cumulative rate function. The multiplicative intensity model is also extended to a regression model by taking the covariate information into account. Statistical methods and theory are developed for estimation of the cumulative rate function and regression parameters. As a major feature of this article, we treat the distributions of both the censoring and latent variables as nuisance parameters. We avoid modeling and estimating the nuisance parameters by proper procedures. An analysis of the AIDS Link to Intravenous Experiences cohort data is presented to illustrate the proposed methods.

Characteristics of injection drug using parents who retain their children

While it is known that injection drug users (IDUs) often have their children removed or place them voluntarily, little is known about factors associated with whether IDU parents live with their children. We identified a community sample of 391 IDU parents with at least one child under age 14 (index IDU parents). For these IDU parents, 62% did not have any of their children under age 14 living with them. We assessed whether certain health factors, risk related behaviors, social indicators, and active drug use were related to whether children of IDUs were living with the index IDU parent. IDU parents who were living with their children were overwhelmingly more likely to be female, more likely to have health insurance, and engage in no-risk or low-risk drug practices, as compared to moderate/high-risk practices. Additionally, HIV negative and HIV positive asymptomatic parents were about three times more likely to be living with their children than HIV positive parents with clinical symptoms commonly seen among those suffering from HIV-related illnesses. HIV-related clinical symptoms, rather than HIV status per se, seem to be associated with retention of children.

The prevalence of homelessness among injection drug users with and without HIV infection

Cross-sectional investigations of homelessness have many potential biases. Data from 2,452 individuals enrolled in a longitudinal cohort study of Baltimore, Maryland, residents recruited in 1988-1989 with a history of injection drug use were analyzed to identify the extent and determinants of homelessness. Proportions having ever experienced homelessness were compared across subgroups of injection drug users (IDUs) who were human immunodeficiency virus (HIV) negative, HIV positive, and HIV seroconverting. Logistic regression identified independent predictors of homelessness. In the cohort, 1,144 (46.7%) participants experienced homelessness during the course of the study. There were differences in prevalence of homelessness by serostatus: 42.4% (n = 621) of participants who remained HIV negative were ever homeless, while 50.6% (n = 346) of HIV-infected individuals and 58.9% (n = 178) of those who seroconverted during the study were ever homeless (P < .001). Participants who consistently denied active injection drug use during follow-up were unlikely to experience homelessness (19%). Independent predictors of homelessness were male sex, HIV seroprevalence, and HIV seroconversion. Following participants over time captures more experiences of homelessness than cross-sectional studies and more accurately identifies risk characteristics. Our data suggest that homelessness is a significant problem among IDUs, especially those with HIV/AIDS.

Gonadal hormone levels in injection drug users

Levels of serum sex hormones, particularly testosterone, luteinizing hormone, and follicle-stimulating hormone, have been shown to change in various disease states. Hypogonadism has been well-documented in patients with HIV infection. As HIV infection and injection drug use are closely linked, we examined whether injection drug users (IDU's) exhibit hormone abnormalities. We evaluated men participating in the ALIVE study (AIDS Linked to Intravenous Experiences), a large cohort study conducted in Baltimore, MD. We found that 20% of 40 IDU subjects (20 HIV+, 20 HIV-) with a mean age 41.5+/-0.9, had low serum total testosterone concentration. We were unable to detect a direct correlation between drug use and hormone levels. Further study is needed on the hormonal milieu of the IDU patient.

Oxidative stress in HIV-1-infected injection drug users

Low serum antioxidant levels observed in many HIV-infected populations could be largely due to an increase in oxidative stress (defined as a disturbance in the equilibrium status of prooxidant/antioxidant systems of intact cells). In HIV infection, oxidative stress may be caused by both overproduction of reactive oxygen intermediates (ROIs) and a simultaneous deficiency of antioxidant defenses. Furthermore, injection drug use has been associated with increased levels of oxidative stress in animal models. Currently, there is widespread use of self-prescribed antioxidant supplementation among the HIV-infected population and a prevailing belief that high-dose supplementation is beneficial, or at the very least, not harmful. Data from our studies show that HIV-positive injection drug users (IDUs) who are on antiretroviral combination therapies including a protease inhibitor have significantly higher mean serum levels of several antioxidants, independent of dietary and supplemental intake, compared with both HIV-negative IDUs and HIV-positive IDUs not taking protease inhibitors. This suggests that oxidative stress may be reduced in patients taking protease inhibitors. Preliminary data suggest that the future of antioxidant supplementation therapy, if any, will be one in which different doses of supplements are recommended for HIV-infected patients on the various antiretroviral treatment regimens. More research is needed to determine the interactions among injection drug use, oxidative stress, antiretroviral therapy, and the use of antioxidant supplements in HIV infection. Until more is known, caution should be exercised when using or recommending high-dose antioxidant supplementation in HIV-infected individuals, particularly in those on protease inhibitors, since moderate levels of oxidative stress are involved in a number of useful physiologic processes.

Nutritional assessment in intravenous drug users with HIV/AIDS

Studying metabolic, endocrine, and gastrointestinal (MEG) disorders in drug abuse and HIV infection is important. Equally important, however, are the tools we use to assess these disorders. Assessment of nutritional status may include any combination of biochemical and body composition measurements, dietary intake assessment, and metabolic studies. Each method has its strengths and weaknesses and there is no perfect tool. When assessing nutritional status in injection drug users (IDU) and in HIV-infected people, the decision on which method or methods to use becomes even more complex. A review of studies reported during the XII World Conference on AIDS reveals that of 64 abstracts on the topic of nutrition in HIV-infected adults, only 11 assessed diet, 41 assessed anthropometry, and 24 assessed some form of biochemical measure. The most commonly reported methods for dietary intake included 24-hour recalls, food records, and food frequencies. The commonest methods used for measuring body composition included height, weight, bioimpedance, and dual-energy x-ray absorptiometry (DEXA). Biochemical measurements included various blood nutrients, lipids, and albumin. Methods varied greatly between studies, and caution should be taken when trying to compare results across studies, especially among those using different methods. Currently, few studies deal with the development of methods that can be used for research in HIV-infected and IDU populations. We need to work toward better tools in dietary intake assessment, body composition, and biochemical measurements, especially methods that will allow us to track changes in nutritional status over time.

Design, implementation, and evaluation at entry of a prospective cohort study of homosexual and bisexual HIV-1-negative men in Belo Horizonte, Brazil: Project Horizonte

BACKGROUND AND OBJECTIVES

Project Horizonte, an open cohort of homosexual and bisexual HIV-1-negative men, is a component of the Minas Gerais AIDS Vaccine Program of the Federal University of Minas Gerais, Belo Horizonte, Brazil. Its objectives included the evaluation of seroincidence of HIV, to ascertain the role of counseling on behavior modification and to assess their willingness to participate in future HIV vaccine trials.

METHODS

Various means of recruitment were used, including pamphlets, notices in community newspapers, radio, and television, at anonymous testing centers, and by word of mouth.

RESULTS

From October 1994 to May 1999, 470 volunteers were enrolled. Their mean age was 26 years and over 70% of them had high school or college education. During the follow-up, they were seen every 6 months, when they received counseling and condoms, and when HIV testing was done. Eighteen seroconversions were observed, and the incidence rates estimates were 1.75 per 100 and 1.99 per 100 person-years, for 36 and 48 months of follow-up, respectively. During the entire period, 139 volunteers were lost to follow-up. Among them, 59 (42.4%) never returned after the initial visit and 51 (36.7) came only once after their initial visit. No losses were observed for those observed during follow-up for more than 3 years. At enrollment, 50% of participants said they would participate in a vaccine trial, and 30% said they might participate.

CONCLUSIONS

The results obtained up to this moment confirm the feasibility of following this type of cohort for an extended period, estimating HIV incidence rate, and evaluating counseling for safe sexual practices in preparation for clinical trials with candidate HIV vaccines in Brazil.

Correlates of enrollment in methadone maintenance treatment programs differ by HIV-serostatus

OBJECTIVES

To identify correlates of enrollment in methadone maintenance treatment programs (MMTP) among a prospective cohort of injection drug users (IDUs) in Baltimore, Maryland.

METHODS

A total of 1480 IDUs undergoing semi-annual HIV tests and interviews were studied between 1994 and 1998, during which time a needle exchange program was introduced. Longitudinal analysis using generalized estimating equations was used to identify correlates of MMTP participation over time.

RESULTS

Although similar proportions of HIV-seropositive and -seronegative IDUs enrolled in MMTP during follow-up (26 versus 22%, respectively), correlates of enrollment differed by HIV-serostatus. Among HIV-seropositive participants, older age [adjusted odds ratio (AOR, 1.37)] was associated with enrollment in MMTP. Among HIV-seronegative IDUs, factors associated with not enrolling in MMTP were being African American (AOR, 0.22) and having been recently incarcerated (AOR, 0.62) or homeless (AOR, 0.72). In both groups, females were twice as likely to be enrolled in MMTP, and those with Medicaid were 1.5 times more likely to be enrolled. When behavioral factors were lagged one visit, needle exchange program attendance was positively associated with MMTP enrollment among HIV-negative IDUs (AOR, 2.10); however, this association diminished significantly over time as dedicated treatment slots for needle exchange program participants became saturated.

CONCLUSIONS

These findings underscore the need to improve access to MMTP, especially to certain subgroups such as African-Americans, the homeless, incarcerated and uninsured. Our data suggest that health care providers and needle exchange programs can facilitate enrollment into MMTP provided that adequate treatment slots are consistently available to this particularly vulnerable population.

Consideration of a North American heroin-assisted clinical trial for the treatment of opiate-dependent individuals

Untreated opiate addiction remains a major public health problem in North America (US and Canada). Increased morbidity and mortality as well as criminal behavior related to untreated opiate dependence constitute significant social and economic burdens. While the principal treatment modality to opiate addiction in North America has been methadone treatment since the 1960s, its reach and effectiveness has been limited; at any given time, only about 25% (US) and 15-20% (Canada) of all opiate addicts are in methadone treatment. Reasons for low levels of treatment participation among this subset of users include perceptions among users that treatment programs are punitive and that the medication is fraught with side effects. In the meantime, alternatives to methadone have been recently approved or are in development, including levo-alpha acetylmethadol and buprenorphine. However, the extent to which they will solve the current problem is still unknown, and therefore development of additional treatment strategies needs to continue. Recent studies of heroin-assisted treatment in Europe (Switzerland, the Netherlands and Great Britain) produced preliminary yet encouraging results in attracting and retaining long-term, treatment-resistant addicts in treatment, as well as improving treatment outcomes. However encouraging, the North American context differs from Europe. A study performed in North America would provide critical information on whether utilizing injectable opiates enhances the overall therapeutic attractiveness and effectiveness of substance abuse treatment to a subset of recalcitrant users. Implications of positive results would expand the continuum of effective interventions in the US and Canada, and increase the number of long-term, treatment-resistant opiate addicts in treatment.

Prevention of HIV infection in street-recruited injection drug users. The Collaborative Injection Drug User Study (CIDUS)

BACKGROUND

Injection drug users (IDUs) and their sex partners account for an increasing proportion of new AIDS and HIV cases in the United States, but public debate and policy regarding the effectiveness of various HIV prevention programs for them must cite data from other countries, from non-street-recruited IDUs already in treatment, or other programs, and from infection rates for pathogens other than HIV.

METHODS

Participants were recruited from the street at six sites (Baltimore [Maryland], New York [two sites], Chicago [Illinois], San Jose [California], Los Angeles [California], and at a state women's correctional facility [Connecticut]), interviewed with a standard questionnaire, and located and reinterviewed at one or more follow-up visits (mean, 7.8 months later). HIV serostatus and participation in various programs and behaviors that could reduce HIV infection risk were determined at each visit.

RESULTS

In all, 3773 participants were recruited from the street, and 2306 (61%) were located and interviewed subsequently. Of 3562 initial serum specimens, 520 (14.6%) were HIV-seropositive; at subsequent assessment, 19 people, all from the East Coast and Chicago, had acquired HIV. Not using previously used needles was substantially protective against HIV acquisition (relative risk [RR], 0.29; 95% confidence interval [CI], 0.11-0.80 ) and, in a multivariate model, was significantly associated with use of needle and syringe exchange programs (adjusted odds ratio [ORadj], 2.08; 95% CI, 1.15-3.85). Similarly, reduction of injection frequency was very protective against seroconversion (RR, 0.33; 95% CI, 0.14-0.80), and this behavior was strongly associated with participation in drug treatment programs (ORadj, 3.54; 95% CI, 2.50-5.00). In a separate analysis, only 37.5% of study-participants had sufficient new needles to meet their monthly demand.

CONCLUSIONS

In this large multicity study of IDUs in the United States, several HIV prevention strategies appeared to be individually and partially effective; these results indicate the continued need for, and substantial gaps in, effective approaches to preventing HIV infection in drug users.

Correlates of HIV infection among young adult short-term injection drug users

OBJECTIVES

To identify risks associated with HIV infection among young adult short-term injection drug users.

METHODS

Current injection drug users, between 18 and 29 years of age, were recruited through street outreach to participate in a cross-sectional survey of HIV prevalence by circumstances of drug injection initiation, HIV-related risk behaviors, and a follow-up to estimate HIV incidence.

RESULTS

At enrollment, 33 (14.4%) of 229 participants were HIV-seropositive. Significant bivariate associations with HIV at the time injection drug use was initiated included age less than or equal to 18 years, having receptive anal sex with the person who assisted with initiation, and having two or more 'trainers' before being able to self-inject. Injecting risks positively associated with HIV included cocaine or speedball (heroin and cocaine together) injection versus heroin or amphetamine injection, injecting five or more times per day, daily crack smoking, backloading, sharing needles at peak drug use, and not using a new needle for every injection. Sexual practices associated with HIV included reporting > 100 lifetime sex partners, a history of sexual assault, being gay or bisexual, and trading sex for money or drugs after starting to inject. In a multivariate model, trading anal sex for money or drugs after initiating injection drug use [odds ratio (OR), 14.2; 95% confidence interval (CI) 3.2-62.3], cocaine/speedball injection (OR, 10.3; 95% CI, 2.2-47.9), daily crack smoking (OR, 4.2; 95% CI, 1.7-10.5), and having two or more trainers (OR, 2.6; 95% CI, 1.1 - 5.9) were independently associated with HIV. During 12 months of follow-up, four persons seroconverted for HIV (annual incidence: 2.6%; 95% CI, 1.1 -5.9%)

CONCLUSIONS

Among short-term injectors, both sexual and injecting practices were important predictors of HIV infection, indicating that a proportion of HIV infections among young injection drug users can be attributed to sexual transmission. The incidence rate for HIV infection suggests that immediate steps should be taken to prevent new infections among young injection drug users.

Improved antioxidant status among HIV-infected injecting drug users on potent antiretroviral therapy

Low serum antioxidant levels in HIV-infected people have been attributed to altered metabolism associated with excess oxidative stress. We conducted a study to examine serum antioxidant levels in 175 HIV-positive and 210 HIV-negative injecting drug users (IDUs) in Baltimore, Maryland. At the time of data collection, 30 of the HIV-positive IDUs were receiving antiretroviral therapies (ART) including a protease inhibitor (PI), 43 ART without a PI, 22 monotherapies, and 80 not on any ART. Serum antioxidants examined included retinol, alpha-tocopherol and gamma-tocopherol, alpha-carotene and beta-carotene, lycopene, lutein/zeaxanthin, and beta-cryptoxanthin. Mean serum levels of lycopene and lutein/zeaxanthin were significantly lower in HIV-positive IDUs than HIV-negative IDUs. Contrary to the findings in other studies, however, levels of the remaining antioxidants in HIV-positive study subjects were not lower than in HIV-negative study subjects. In fact, serum alpha-tocopherol levels were significantly higher in HIV-positive IDUs than HIV-negative IDUs (medians = 744 microg/dl and 718 microg/dl, respectively; p = .04). Among HIV-positive study subjects, there were significant differences in antioxidant levels by ART regimen. In multivariate models adjusting for injecting drug use, dietary intake, supplement intake, gender, and alcohol intake, significant overall differences by ART regimen were observed for alpha-tocopherol, beta-carotene, and beta-cryptoxanthin. Serum levels of these three antioxidants were significantly higher in the PI group than in the other three ART groups combined (p = .0008, 0.02, and 0.02, respectively). These data provide indirect evidence of the effectiveness of PIs in lowering oxidative stress levels in HIV-positive IDUs.

Hypervariable region 1 sequence stability during hepatitis C virus replication in chimpanzees

The putative envelope 2 (E2) gene of hepatitis C virus (HCV) contains a highly variable region referred to as hypervariable region 1 (HVR1). We hypothesized that this genetic variability is driven by immune selection pressure, rather than representing the accumulation of random mutations in a region with relatively little functional constraint. To test this hypothesis, we examined the E2 sequence of a human inoculum that was passaged through eight chimpanzees, which appear to have a replicative rate (opportunity for chance mutation) similar to that of humans. Acute-phase plasma samples from a human (the inoculum) and six of eight serially infected chimpanzees were studied. For each, 33 cloned cDNAs were examined by a combined heteroduplex-single-stranded conformational polymorphism assay to assess quasispecies complexity and optimize selection of clones with unique gel shift patterns (clonotypes) for sequencing. The sequence diversity of HCV was significantly lower in the chimpanzees than in the humans, and during eight serial passages there was no change in the sequence of the majority clonotype from each animal examined. Similarly, the rates of protein sequence altering (nonsynonymous) substitution were lower in the chimpanzees than in the humans. These findings demonstrate that nonsynonymous mutations indicate selection pressure rather than being an incidental result of HCV replication.

Screening for hepatitis C virus in human immunodeficiency virus-infected individuals

Immunosuppression from human immunodeficiency virus (HIV) may impair antibody formation, and false-negative hepatitis C virus antibody (anti-HCV) tests have been reported in individuals coinfected with HIV and HCV. It is unknown if the frequency of false-negative tests is sufficiently high to change screening recommendations in this setting. Thus, the prevalence of false-negative results for anti-HCV by third-generation tests was determined with samples from HIV-infected individuals. Sera from 559 HIV-infected and 944 HIV-negative prospectively followed injection drug users were tested for anti-HCV by a third-generation enzyme immunoassay and for HCV RNA by using a branched DNA assay and the HCV COBAS AMPLICOR system. Of 559 HIV-infected participants, 547 (97.8%) were anti-HCV positive. One of the remaining 12 anti-HCV-negative participants was HCV RNA positive, and she later developed detectable anti-HCV. Of the 944 HIV-negative participants, 825 (87.4%) were anti-HCV positive. One of the remaining 119 anti-HCV-negative participants was HCV RNA positive, and she also developed detectable anti-HCV at a later visit. These data indicate that HIV infection does not alter the approach to hepatitis C virus screening, which should be performed with third-generation assays for anti-HCV unless acute infection is suspected.

Needle-exchange attendance and health care utilization promote entry into detoxification

This study was undertaken to identify factors associated with entry into detoxification among injection drug users (IDUs), and to assess the role of needle-exchange programs (NEPs) as a bridge to treatment. IDUs undergoing semiannual human immunodeficiency virus (HIV) tests and interviews were studied prospectively between 1994 and 1998, during which time an NEP was introduced in Baltimore. Logistic regression was used to identify independent predictors of entry into detoxification, stratifying by HIV serostatus. Of 1,490 IDUs, similar proportions of HIV-infected and uninfected IDUs entered detoxification (25% vs. 23%, respectively). After accounting for recent drug use, hospital admission was associated with four-fold increased odds of entering detoxification for HIV-seronegative subjects. Among HIV-infected subjects, hospital admission, outpatient medical care, and having health insurance independently increased the odds of entering detoxification. After accounting for these and other variables, needle-exchange attendance also was associated independently with entering detoxification for both HIV-infected (adjusted odds ratio [AOR] = 3.2) and uninfected IDUs (AOR = 1.4). However, among HIV-infected subjects, the increased odds of detoxification associated with needle exchange diminished significantly over time, concomitant with statewide reductions in detoxification admissions. These findings indicate that health care providers and NEPs represent an important bridge to drug abuse treatment for HIV-infected and uninfected IDUs. Creating and sustaining these linkages may facilitate entry into drug abuse treatment and serve the important public health goal of increasing the number of drug users in treatment.

Acute hepatitis C virus structural gene sequences as predictors of persistent viremia: hypervariable region 1 as a decoy

We hypothesized that hepatitis C virus (HCV) persistence is related to the sequence variability of putative envelope genes. This hypothesis was tested by characterizing quasispecies in specimens collected every six months from a cohort of acutely HCV-infected subjects (mean duration of specimen collection, 72 months after seroconversion). We evaluated 5 individuals who spontaneously cleared viremia and 10 individuals with persistent viremia by cloning 33 1-kb amplicons that spanned E1 and the 5' half of E2, including hypervariable region 1 (HVR1). To assess the quasispecies complexity and to detect variants for sequencing, the first PCR-positive sample was examined by using a previously described method that combines heteroduplex analysis and analysis of single-stranded conformational polymorphisms. The ratio of nonsynonymous to synonymous substitutions (dN/dS) within each sample was evaluated as an indicator of relative selective pressure. Amino acid sequences were analyzed for signature patterns, glycosylation signals, and charge. Quasispecies complexity was higher and E1 dN/dS ratios (selective pressure) were lower in those with persistent viremia; the association with persistence was strengthened by the presence of a combination of both characteristics. In contrast, a trend toward higher HVR1 dN/dS ratios was detected among those with persistent viremia. We did not detect any such association for factors that may affect complexity such as serum HCV RNA concentration. HVR1 had a lower positive charge in subjects with persistent viremia, although no consistent motifs were detected. Our data suggest that HCV persistence is associated with a complex quasispecies and immune response to HVR1.

HIV prevalence and risk behaviors among new initiates into injection drug use over the age of 40 years old

Injection drug users (IDUs) who have recently initiated an injecting career have high risk behaviors for HIV infection. The average age of an IDUs first injection is typically reported as 19-20 years, and some literature has reported 'maturing out' of drug use typically around 40 years old. The purpose of this study was to discriminate risk behaviors among newer injectors by age of initiation. This cross-sectional study includes volunteers enrolled using extensive community recruitment techniques in 1988 and 1989 in Baltimore, MD. Of the 722 injection drug users who had initiated injection within the prior six years, 124 were over 35 years old of whom 53 were 40 years and older. Rates of HIV were lower among those over 40 years (13.2%) than those who were 35-39 years (22.5%) or under 35 years old (20.9%). The behavioral characteristics of those aged 35-39 were similar to those under 35 years old, but those over 40 years old when compared with less than 39 years reported injecting at least daily less frequently (88.7% vs 76.5%, p = 0.042), using a needle from a sterile wrapper (19.0% vs 36.6%, p = 0.006), and injecting with their own works (52.6% vs 28.8%, p = 0.021). This study shows that people initiate injection drug use across a wide age range and that needle hygiene practices early after initiation tend to be safer in older compared to younger initiates.

Direct comparison of time to AIDS and infectious disease death between HIV seroconverter injection drug users in Italy and the United States: results from the ALIVE and ISS studies. AIDS Link to Intravenous Experiences. Italian Seroconversion Study

OBJECTIVE

To compare the rate of HIV disease progression in a sample of polydrug injectors (AIDS Link to Intravenous Experiences [ALIVE] study) with that in a sample of predominantly opiate injectors (Italian Seroconversion Study [ISS]).

DESIGN

Prospective cohort studies of HIV-positive individuals whose date of seroconversion (SC) is known with a good degree of precision. The ALIVE study involves a community-based cohort of injection drug users (IDU) in the United States and the ISS reports on a clinic-based cohort of seroconverters in Italy with different exposure modalities to HIV.

METHODS

Data from the two cohorts were combined. The date of SC was estimated as the midpoint in time between the last negative and the first positive HIV test. Time-to-event (i.e., AIDS or death from an infectious disease) statistical methods were used. Relative hazards (RH) of progression to event were adjusted by age at SC, gender, and year of SC.

RESULTS

Of the 1003 IDUs (251 from ALIVE and 752 from ISS), 226 progressed to AIDS, and 146 died after AIDS or from an infectious disease; of these, 10 were without an AIDS diagnosis. The two groups of IDUs differed in terms of age at SC (median, 35 years for ALIVE and 25 years for ISS), proportion of women (24% versus 31%), race (7.6% versus 100% white), and year of seroconversion (i.e., ISS participants seroconverted, on average, earlier than ALIVE participants). Although the univariate analysis suggested possible differences for progression to AIDS, or to death from infectious disease between cohorts, multivariate analyses that adjusted for age showed no significant differences by cohort, gender, race, or time of seroconversion. The median time to AIDS for 25-year-old persons was 12.3 years for ALIVE and 11.8 years for ISS; for 35-year-old persons, it was 8.5 and 8.2 years, respectively. These estimates were similar to those for non-IDUs observed in the ISS and to those from large cohort of homosexual men.

CONCLUSION

Our results confirm the importance of accounting for age when considering the incubation period for HIV infection. Despite differences in drug use characteristics, the similar median times to AIDS, for each age, between the two cohorts of IDUs and between the IDUs and the non-IDUs suggest a negligible effect of injection drug use on HIV progression.