Abstract
Background and objectives: Proteinuria is a major determinant of chronic kidney disease. We aimed to characterize the prevalence and correlates of proteinuria in a cohort of HIV-infected and uninfected injection drug users.
Design, setting, participants, & measurements: A cross-sectional analysis was performed among 902 injection drug users (273 HIV-infected) in the AIDS Linked to the Intravenous Experience cohort. The primary outcome was proteinuria defined as having a urine protein/creatinine concentration ratio >200 mg/g. Poisson regression with robust variance was used to determine prevalence ratios.
Results: Overall, 24.8% of participants had proteinuria; the prevalence was 2.9 times higher among HIV-infected participants (45%) compared with HIV-uninfected participants (16%). In addition, age, health insurance, employment status, hepatitis B and C serostatus, diabetes, and high BP were associated with proteinuria. Neither antiretroviral therapy nor features of illicit drug use history were associated with proteinuria. In multivariate analysis, HIV infection, unemployment, increased age, diabetes, hepatitis C infection, and high BP were significantly associated with a higher prevalence of proteinuria.
Conclusions: In an aging, predominantly African-American cohort of injection drug users, we found a striking burden of proteinuria that was strongly associated with HIV status. In addition to being a pathway to ESRD, proteinuria is a potent risk factor for cardiovascular morbidity and mortality. Evaluation of aggressive screening and disease-modification strategies in this high-risk population is warranted.
With the advent of highly active antiretroviral therapy (HAART), HIV-infected people are living longer and increasingly facing the complications of aging including chronic kidney disease (CKD). Incidence rates of kidney disease appear higher among HIV-infected individuals compared with the general population (1,2). This higher incidence is likely explained by a direct effect of HIV infection as well as indirect effects and confounding related to known risk factors for renal disease commonly present among HIV-infected populations including low socioeconomic status, African-American race, and comorbid infectious disease. Beyond HIV itself, it has been hypothesized that certain antiretroviral medications may be nephrotoxic (3,4). It is also possible that cumulative exposure to antiretroviral therapy may have chronic harmful effects on the kidneys.
Injection drug users (IDUs) have long been identified as a risk population for HIV infection and may be at higher risk for kidney disease compared with the general U.S. population. In addition to HIV, IDUs also have high prevalence of other risk factors for kidney damage including illicit drug use (e.g., vascular damage from injected adulterants), viral hepatitis, low socioeconomic status, and poor access to medical care (5,6). Few data are available on the burden of CKD in IDU populations. The objective of this analysis was to investigate the prevalence and correlates of proteinuria in a well-characterized community-based cohort of HIV-infected and uninfected IDUs. Of particular interest were associations with HIV infection, hepatitis C infection, and drug use.
Materials and Methods
Study Population
The study population derives from the AIDS Linked to the IntraVenous Experience (ALIVE) cohort in Baltimore, Maryland, previously described in detail (7). Briefly, this community-based cohort began in 1988 with additional enrollments in 1994 through 1995, 1998, 2000, and 2005 through 2008 and was originally aimed at studying the incidence and natural history of HIV among IDUs. During all recruitment periods, 4376 total IDUs were recruited, of whom 1065 (24.3%) were HIV-infected at enrollment and 340 (7.7%) seroconverted during follow-up. The study sample for this analysis included 902 IDUs who were in active follow-up when collection of urine samples and testing of serum for creatinine was initiated in 2007 through 2008. The study was approved by the Johns Hopkins School of Public Health Institutional Review Board and all participants provided written informed consent.
Measurements and Definitions
At semiannual visits, participants answered questionnaires administered by trained interviewers about their medical history, risk behaviors, and current medications including antiretroviral regimens if HIV-infected. Participants completed questionnaires concerning sensitive risk behaviors, including drug use, on a computer in a private room. Participants also underwent BP measurement and biospecimen collection. High BP was defined as having either a systolic measurement ≥140 mmHg or a diastolic measurement ≥90 mmHg. History of hypertension and diabetes were defined as self-reported physician diagnosis of these conditions. After each visit, serum was tested for antibodies against HIV among uninfected participants and for HIV RNA and CD4 cell count among HIV-infected participants. HIV RNA levels <50 copies per milliliter were defined as undetectable on the basis of the assay used (Roche Amplicore version 1.5).
All ALIVE participants were tested for hepatitis B virus (HBV) and hepatitis C virus (HCV) infection during their study visits between 2005 and 2008. Previous HBV infection was defined as having detectable antibodies to the core antigen (anti-HBc) and being hepatitis B surface antigen (HBsAg)–negative whereas chronic infection was defined as being anti-HBc–positive and HBsAg-positive. HCV infection was defined as a positive HCV antibody test.
In addition to the semiannual data collection, beginning in October 2007, spot urine samples were collected annually from all participants, whereas creatinine concentration was measured from serum samples. Urine samples were tested for protein and creatinine concentration levels. Serum creatinine measurements were used to calculate estimated GFR (eGFR) using the Modification of Diet in Renal Disease (MDRD) equation (8).
Outcome of Interest
The primary outcome of interest was proteinuria, which was defined as having a urine protein/creatinine concentration ratio (P/CrR) >200 mg/g, a cutoff recommended for usage in clinical practice by the Kidney Disease Outcomes Quality Initiative of the National Kidney Foundation (9). P/CrRs from spot urine samples have been found to correlate closely with 24-hour urine collections, and are simpler to obtain and may be less prone to collection errors (10).
Statistical Analyses
Correlates of proteinuria were determined using χ2 tests for categorical variables, Mann-Whitney tests for continuous variables, and univariate Poisson regression with robust variance. Poisson regression with robust variance was used instead of logistic regression because of the high prevalence of the outcome and because it allows for direct estimation of prevalence ratios (PRs), along with 95% confidence intervals [CIs] (11).
A multivariate model was constructed to identify independent correlates of proteinuria. Multivariate models included covariates determined to be statistically important on the basis of associations with proteinuria in the univariate analysis (P < 0.10) and a priori knowledge based on previous literature. The primary exposures of interest were HIV, HCV, and drug use. Other variables were selected for inclusion in models if they were hypothesized to be associated with any of these factors and causally associated with the outcome. HBV prevalence would be expected to be higher among people with HIV and HCV (shared common cause) and can also be hypothesized to be associated with renal disease. Employment status and health insurance were considered as markers of socioeconomic status, which is linked with all three exposures and the outcome. Hypertension and diabetes have also been linked to HIV and HCV, although the causal mechanism remains unknown, and are also known risk factors for renal outcomes. Also included were important demographic characteristics deemed to be possible confounders regardless of their associations in this population (i.e., race, gender, and age). In the final multivariate models, categorical variables were condensed when multiple categories were biologically similar and had statistically equivalent proteinuria prevalence.
A sensitivity analysis was conducted including those individuals who had two P/CrR measurements obtained 1 year apart (n = 753, 83.5%). We examined the correlation of the two measurements using 2 × 2 tables. In addition, regression models using two consecutive P/CrR measurements indicative of proteinuria as an outcome were compared with results of the initial analysis.
In addition to recent exposure, illicit drug use was assessed through summary variables designed to capture the cumulative exposure to drugs before the time of outcome assessment. Multiple measures were considered including the proportion of visits where use was reported, number of visits where use was reported, average number of injections per year, and time since initiation of injection. A prespecified subset analysis was also performed including only those who were HIV-infected at the time of P/CrR measurement. In this analysis, HIV RNA level, CD4 cell count, and use of antiretroviral therapy were considered in univariate and multivariate analyses. Variables included in the multivariate analysis of the HIV-infected population were selected in the same way as in the analysis of the full population. All statistical analyses were performed using SAS ver. 9.1 (SAS Institute, Cary, NC).
Results
Baseline Characteristics
The median age of the 902 participants was 49.5 years (range 21 to 80 years old). The population was predominately African American (91.6%) and 65.4% were male (Table 1). Thirty percent were HIV-infected at the time of measurement. Although all participants had a history of injection drug use, only 35.9% reported having injected drugs within the 6 months before the clinic visit when proteinuria assessment was performed.
Table 1.
Characteristics of study participants
|
n (%) |
|||
|---|---|---|---|
| P/CrR ≤200 | P/CrR >200 | Total | |
| Race | |||
| other | 66 (86.8%) | 10 (13.2%) | 76 (8.4%) |
| African American | 612 (74.1%) | 214 (25.9%) | 826 (91.6%) |
| Gender | |||
| male | 452 (76.6%) | 138 (23.4%) | 590 (65.4%) |
| female | 226 (72.4%) | 86 (27.6%) | 312 (34.6%) |
| Mean age (years) | 48.4 ± 8.0 | 50.4 ± 7.2 | 48.9 ± 7.9 |
| Work statusa | |||
| unemployed | 490 (72.1%) | 190 (27.9%) | 680 (75.6%) |
| employed | 186 (84.9%) | 33 (15.1%) | 219 (24.4%) |
| Health insurancea | |||
| uninsured | 187 (80.3%) | 46 (19.7%) | 233 (26.0%) |
| insured | 486 (73.4%) | 176 (26.6%) | 662 (74.0%) |
| Injection drug use in prior 6 months | |||
| none | 438 (75.8%) | 140 (24.2%) | 578 (64.1%) |
| cocaine only | 24 (68.6%) | 11 (31.4%) | 35 (3.9%) |
| heroin only | 77 (80.2%) | 19 (19.8%) | 96 (10.6%) |
| both | 139 (72.0%) | 54 (28.0%) | 193 (21.4%) |
| Crack use in prior 6 months | |||
| no | 469 (73.5%) | 169 (26.5%) | 638 (70.7%) |
| yes | 209 (79.2%) | 55 (20.8%) | 264 (29.3%) |
| HIV serostatus | |||
| negative | 529 (84.1%) | 100 (15.9%) | 629 (69.7%) |
| positive | 149 (54.6%) | 124 (45.4%) | 273 (30.3%) |
| Hepatitis B serostatusa | |||
| uninfected (anti-HBc−) | 158 (82.7%) | 33 (17.3%) | 191 (21.4%) |
| anti-HBc+/HBsAg− | 494 (72.9%) | 184 (27.1%) | 678 (76.1%) |
| anti-Hbc+/HBsAg+ | 21 (95.5%) | 1 (4.6%) | 22 (2.5%) |
| Hepatitis C antibodya | |||
| negative | 116 (89.9%) | 13 (10.1%) | 129 (14.3%) |
| positive | 562 (72.8%) | 210 (27.2%) | 772 (85.7%) |
| Diabetesa | |||
| never diagnosed | 616 (76.7%) | 187 (23.3%) | 803 (89.2%) |
| ever diagnosed | 61 (62.9%) | 36 (37.1%) | 97 (10.8%) |
| BP and hypertension diagnosisa | |||
| normal BP/no diagnosed hypertension | 343 (81.1%) | 80 (18.9%) | 423 (47.2%) |
| normal BP/diagnosed hypertension | 96 (76.2%) | 30 (23.8%) | 126 (14.1%) |
| high BP | 236 (68.0%) | 111 (32.0%) | 347 (38.7%) |
| eGFRb | |||
| ≥60 | 662 (78.3%) | 183 (21.7%) | 845 (93.7%) |
| <60 | 16 (28.1%) | 41 (71.9%) | 57 (6.3%) |
Some missing data due to participants choosing not to answer or test results being unavailable.
eGFR was calculated using serum creatinine measurements and the MDRD equation.
Prevalence and Correlates of Proteinuria
Among the 902 participants, 224 (24.8%) participants had proteinuria. Of these 224, 41 (18.3%) also had an eGFR <60, whereas of those without proteinuria, 16 (2.4%) had a eGFR <60. Compared with HIV-uninfected participants, the distribution of P/CrR values in HIV-infected participants was bimodal and shifted to the right (Figure 1). The prevalence of proteinuria was significantly higher in HIV-infected versus uninfected people (45.4 versus 15.9%, respectively; PR, 2.86; 95% CI 2.29 to 3.57) (Table 2).
Figure 1.
Distribution of urine protein/creatinine ratios by HIV status. Values represent log-transformed data; vertical reference line represents ln(200) urine protein/creatinine concentration ratio used as cutoff for clinical proteinuria.
Table 2.
Correlates of proteinuriaa among all study participants
| Variable | Prevalence of Proteinuriaa |
Univariate |
Multivariate Modelb |
|||
|---|---|---|---|---|---|---|
| % (n) | P | PR | 95% CI | PR | 95% CI | |
| Race | ||||||
| other | 13.2 (10) | 1.00 | 1.00 | |||
| African American | 25.9 (214) | 0.014 | 1.97 | 1.09 to 3.55 | 1.31 | 0.74 to 2.32 |
| Gender | ||||||
| male | 23.4 (138) | 1.00 | 1.00 | |||
| female | 27.6 (86) | 0.168 | 1.18 | 0.93 to 1.49 | 1.15 | 0.92 to 1.44 |
| Age | ||||||
| by 10 years (continuous) | 1.29 | 1.12 to 1.48 | 1.24 | 1.05 to 1.47 | ||
| Work status | ||||||
| unemployed | 27.9 (190) | 1.00 | 1.00 | |||
| employed | 15.1 (33) | <0.001 | 0.54 | 0.39 to 0.76 | 0.67 | 0.48 to 0.94 |
| Health insurance | ||||||
| uninsured | 19.7 (46) | 1.00 | 1.00 | |||
| insured | 26.6 (176) | 0.038 | 1.35 | 1.01 to 1.80 | 0.90 | 0.67 to 1.20 |
| Injection drug use in prior 6 months | ||||||
| none | 24.2 (140) | 1.00 | 1.00 | |||
| cocaine only | 31.4 (11) | 1.30 | 0.78 to 2.16 | 1.39 | 0.82 to 2.34 | |
| heroin only | 19.8 (19) | 0.82 | 0.53 to 1.25 | 1.16 | 0.74 to 1.82 | |
| both | 28.0 (54) | 0.353 | 1.16 | 0.88 to 1.51 | 1.27 | 0.97 to 1.65 |
| Crack use in prior 6 months | ||||||
| no | 26.5 (169) | 1.00 | ||||
| yes | 20.8 (55) | 0.074 | 0.79 | 0.60 to 1.03 | 0.85 | 0.65 to 1.11 |
| HIV status | ||||||
| negative | 15.9 (100) | 1.00 | 1.00 | |||
| positive | 45.4 (124) | <0.001 | 2.86 | 2.29 to 3.57 | 2.92 | 2.29 to 3.72 |
| Hepatitis B serostatus | ||||||
| uninfected (anti-HBc−) | 17.3 (33) | 1.00 | 1.00 | |||
| anti-HBc+/HBsAg− | 27.1 (184) | 1.41 | 1.03 to 1.91 | 0.81 | 0.59 to 1.10 | |
| anti-Hbc+/HbsAg+ | 4.6 (1) | 0.002 | 0.24 | 0.03 to 1.63 | 0.17 | 0.02 to 1.13 |
| Hepatitis C antibody | ||||||
| negative | 10.1 (13) | 1.00 | 1.00 | |||
| positive | 27.2 (210) | <0.001 | 2.70 | 1.59 to 4.58 | 1.84 | 1.03 to 3.27 |
| Diabetes | ||||||
| never diagnosed | 23.3 (187) | 1.00 | 1.00 | |||
| ever diagnosed | 37.1 (36) | 0.003 | 1.59 | 1.20 to 2.13 | 1.52 | 1.12 to 2.05 |
| BP | ||||||
| normal | 20.0 (110) | 1.00 | 1.00 | |||
| high | 32.0 (111) | <0.001 | 1.60 | 1.28 to 2.01 | 1.46 | 1.17 to 1.82 |
95% CI, for an alpha of 0.05.
Proteinuria defined as a urine protein-creatinine ratio >200.
Model adjusted for all other covariates in column.
Variables associated in univariate analysis with higher prevalence of proteinuria included African-American race, older age, possession of health insurance, unemployment, prior HBV infection, prior HCV infection, diabetes diagnosis, and elevated BP. Other variables that were considered, but were not associated with proteinuria in univariate analysis, include cigarette smoking and incarceration in the previous 6 months. In multivariate analysis, HIV remained significantly associated with proteinuria (PR = 2.92, 95% CI = 2.29 to 3.72). Other factors that remained significantly associated with proteinuria after adjustment for other possible explanatory variables included older age, unemployment, prior HCV infection, diabetes diagnosis, and elevated BP (Table 2).
In the sensitivity analysis, 18% of the 753 participants with two consecutive measurements had a P/CrR >200 at both measurements. Sixty-one percent of HIV-uninfected participants with a first measurement of proteinuria had proteinuria at second measurement, and 81% of HIV-infected participants with a first measurement of proteinuria had proteinuria at second measurement. The associations of the multivariate model were not diminished when we considered only those who had evidence of proteinuria at both visits to have the outcome of interest (data not shown). For example, the association with HIV was strengthened (PR = 3.47, 95% CI = 2.50 to 4.81).
Measures of recent or cumulative illicit drug use were not significantly associated with proteinuria. No association was observed with self-reported injection drug use (cocaine and/or heroin) or crack cocaine use in the 6 months before the assessment visit (Table 2). Moreover, we found no significant association between various cumulative measures of drug use including proportion of study visits in which injection use was reported since cohort enrollment (PR per 10% increase in proportion = 0.99, 95% CI = 0.68 to 1.42), the number of injections per year of reported injection drug use (PR per 100 injections per year = 0.97, 95% CI = 0.91 to 1.04), and duration of injection drug use and proteinuria (PR per 5 years of injection drug use: 1.06; 95% CI = 0.97 to 1.16). The results of these analyses were unchanged after adjustment for age, race, sex, and employment status and in a sensitivity analysis including only those with two consecutive abnormal measurements (data not shown).
Proteinuria among HIV-Infected IDUs
In analysis restricted to HIV-infected participants, an increase of 50 cells in participants' CD4+ count was associated with a 9% lower likelihood of prevalence of proteinuria (95% CI = 5 to 12%). Furthermore, there was an association between higher viral load and proteinuria prevalence (PR per log10 increase: 1.10; 95% CI = 1.06 to 1.14; Table 3). However, self-reported HAART use in the past 6 months and time since HAART initiation were not significantly associated with proteinuria. In multivariate analysis, CD4+ count remained significantly associated with proteinuria, whereas having detectable HIV RNA did not.
Table 3.
Correlates of proteinuriaa among HIV-infected participants
| Variable | Proteinuriaa |
Univariate |
Multivariateb |
||||
|---|---|---|---|---|---|---|---|
| Noc | Yesc | P | PR | 95% CI | PR | 95% CI | |
| CD4+ cells | |||||||
| cell count | 348 (214 to 559) | 202 (87 to 371) | |||||
| [per 50 cell increase] | <0.001 | 0.91 | 0.88 to 0.95 | 0.92 | 0.88 to 0.97 | ||
| HIV RNA detection | |||||||
| undetectable | 69 (64.49) | 38 (35.51) | |||||
| detectable | 84 (48.00) | 91 (52.00) | 0.003 | 1.59 | 1.17 to 2.16 | 1.44 | 0.91 to 2.28 |
| HIV viral load | |||||||
| viral load | 84 (40 to 8800) | 6370 (40 to 55450) | |||||
| [log-transformed viral load increase] | <0.001 | 1.10 | 1.06 to 1.14 | ||||
| HAART use in last 6 months | |||||||
| no | 60 (49.18) | 62 (50.82) | |||||
| yes | 81 (55.48) | 65 (44.52) | 0.217 | 0.85 | 0.66 to 1.10 | ||
| Time since first HAART used | 1.49 (0 to 6.85) | 1.95 (0 to 7.11) | 0.331 | 1.02 | 0.98 to 1.05 | 1.01 | 0.96 to 1.07 |
| years | |||||||
95% CI, for an alpha of 0.05.
Proteinuria defined as a urine protein/creatinine ratio >200.
Adjusted for all variables in addition to age, African-American race, and gender, and except for log (HIV viral load) and HAART use in last 6 months so as to avoid collinearity.
For categorical variables cells show n (percentage across rows); for continuous variables cells show median (interquartile range).
Refers to first reported use of HAART at a study visit.
Discussion
In this large, community-based cohort of predominantly African-American IDUs in Baltimore, Maryland, we observed a heavy burden of proteinuria with nearly a one quarter of the population having a P/CrR >200 at their first visit and almost one fifth having a P/CrR >200 at two consecutive visits. The burden was even more striking among HIV-infected participants with nearly one half having proteinuria. HIV infection clearly represents a strong risk factor for proteinuria; however, classic risk factors such as age, diabetes, and hypertension also substantially contributed to proteinuria risk. Because proteinuria is a strong risk factor for both progression to ESRD and for cardiovascular disease events and mortality (12,13), the high proteinuria prevalence will likely translate into severe consequences. Our P/CrR measurements followed a bimodal distribution that may be indicative of a susceptible subset within HIV-infected individuals. Importantly, however, most of these proteinuria risk factors are potentially modifiable, highlighting the substantial need for appropriate screening and management approaches to prevent development and progression of kidney disease among IDUs.
Our estimates of the burden of proteinuria among IDUs and among HIV-infected people exceed those reported from prior studies among comparable populations. In a study of 202 Hispanic HIV-infected and -uninfected drug users, the prevalence of proteinuria was only 10.9% (14). A meta-analysis of HIV/HCV co-infected populations, to which our HIV-infected participants should be similar, suggested a pooled estimate of CKD prevalence of 28% (15), which is 21% lower than the estimate among our HIV/HCV co-infected participants. The heavier burden of proteinuria in our population may collectively reflect the predominance of African-American participants in our study and the high prevalence of HIV, HCV, and HIV/HCV co-infection. Proteinuria as a measure of CKD was more prevalent than low eGFR (25.2 versus 6.3%) in our population. By comparison, in the 1999 through 2004 National Health and Nutrition Examination Surveys (NHANES), the estimated population prevalence of albuminuria and eGFR <60 ml/min per 1.73 m2 were similar (8.2 and 5.6%, respectively) (16). This was also true in the 40 to 60 years of age category of NHANES (albuminuria prevalence of 7.9% and eGFR <60 ml/min per 1.73 m2 prevalence of 5.7%) in which the majority of our population would be categorized (17). However, it was also demonstrated that these measures may be identifying different subpopulations (17). The implications of this different pattern of CKD phenotype are not known, but proteinuria may serve as a predictor of kidney failure independent of eGFR (18).
By far the strongest predictor of proteinuria in this population was HIV infection. In adjusted analysis, HIV infection was associated with a nearly threefold higher prevalence of proteinuria. The magnitude of effect observed in our study is similar to associations reported from previous studies, both of those that explored measurements of late-stage and early kidney disease outcomes. An analysis of ESRD in African Americans in Baltimore (including ALIVE participants) found a risk ratio of ESRD of 3.9 associated with HIV infection (Gregory M. Lucas, Johns Hopkins School of Medicine, personal communication, 2009; see reference 19). Other studies done in more racially diverse populations have found similar associations between HIV and other indicators of kidney disease including elevated cystatin C and eGFR <60 ml/min per 1.73 m2 (20,21). Although studies examining proteinuria have been sparse, a study including HIV-infected and HIV-uninfected Hispanic drug users found that prevalence of proteinuria was 15% in the HIV-infected group and only 1% in the HIV-uninfected group (14). In another study HIV-infected participants had 5 times the odds of microalbuminuria as compared with uninfected controls (22), and microalbuminaria has been shown to predict development of proteinuria in patients with HIV infection (23).
Among HIV-infected participants, the prevalence of proteinuria was greater among people with higher viral replication and immunosuppression. Although HAART use did not appear protective against proteinuria, it is possible that some positive effects of HAART are reflected in the effects of CD4 and viral load. It is possible that failure to detect an association with HAART reflects unmeasured confounding or selection factors related to who received therapy in this cohort, but these data do suggest that the elevated prevalence of proteinuria in HIV-infected individuals is not solely due to HAART exposure.
Interestingly, neither recent nor history of illicit drug use was strongly associated with proteinuria. Cocaine and heroin use have previously been linked to kidney disease (24,25), but there has not been detailed longitudinal characterization of the effects of illicit drugs on CKD. Although we were limited by not having in our sample people that had never used drugs, we did have people with varying trajectories and types of substance use (26).
It is currently recommended by the Infectious Diseases Society of America that, upon diagnosis with HIV, individuals should have a urinalysis to screen for proteinuria, and HIV-infected individuals should undergo annual urinalysis if they are African American, have a low CD4+ count, have a high viral load, or have diabetes mellitus, hypertension, or HCV co-infection (3). Our findings support these characteristics as important in targeting higher risk populations, as diabetes, elevated BP, and HCV infection were significantly associated with proteinuria in this population. HCV infection is of particular importance in this population given the high prevalence in this cohort and IDUs in general (85.6% of our study population had been HCV-infected).
Although we did not observe an association between African-American race and proteinuria, this may reflect the small sample size of Caucasians, and the overall high prevalence of proteinuria in this cohort may at least in part reflect the predominance of African Americans. In previous studies, which drew from this and similar populations, strong associations were observed between African-American race and development of ESRD, and survival rates after dialysis initiation were poor (15,27). Early diagnosis and care for CKD may help to prevent more serious kidney damage, eventual kidney failure, and the need for dialysis.
This study has several limitations. First, the consensus definition of CKD requires that kidney damage or decreased kidney function be present for ≥3 months (28). Our study was based on single measurements of urine protein and, therefore, may overestimate the true prevalence of proteinuria. However, similar methodology has been used by NHANES to assess the prevalence of CKD in the United States (29). In addition, our prevalence estimate may be more generalizable than results reported from clinic-based populations because our study is community-based, with study visits performed independently of medical evaluation. Our diagnosis, treatment, and behavioral variables were limited in that they were collected through self-report. However, all treatment and clinical information was collected by trained interviewers whereas sensitive behavioral information was collected by audio computer-assisted self-interview in a safe, private setting, previously shown to increase reporting of stigmatized behaviors (30).
Our assessment of illicit drug and antiretroviral history may not fully account for exposure before study entry. Additionally, the lack of observation of an association between illicit drug use and proteinuria may at least in part reflect a healthy drug user bias. Specifically, declining health may be a reason for reduced use of illicit drugs, making drug use appear to be associated with better overall health. Adherence to antiretroviral medications also could not be assessed. Although our study did not show a significant association between illicit or antiretroviral drug use and proteinuria, this may be due to limited study power. Associations have been found in other studies and this relationship should be studied further to determine whether these are important components of risk.
We observed an alarmingly high prevalence of proteinuria in this cohort of IDUs, particularly among those who were HIV-infected. Our findings suggest that in addition to the unique implications of HIV infection, CKD risk factors among IDU populations may parallel those in the general population. Longitudinal data will allow us to confirm these risk factor relationships and also to evaluate the influence of appropriate access to care and to define rates of progression. Although our data clearly support the current Infectious Diseases Society of America guidelines for identifying kidney disease among HIV-infected peoples, these findings also warrant further investigation of expanded screening and more aggressive approaches to ensuring appropriate risk factor modification with the aim of slowing CKD progression in this high-risk population.
Disclosures
None.
Acknowledgments
E.L.Y. performed the data analysis, interpreted results, and drafted the initial manuscript. G.M.L. and S.H.M. additionally participated in the interpretation of results. S.H.M., D.V., and G.D.K. conceived study design and coordinated the study. All authors participated in discussion and review of the manuscript.
We acknowledge the contributions of the study participants in the ALIVE cohort and also all of the study staff.
This research was supported by the National Institutes of Health (R01DA12568, R01DA04434, K23 DA015616, and R01DA018577).
These data were presented at the Society of Epidemiologic Research Annual Meeting; June 23 through 26, 2009; Anaheim, CA.
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
References
- 1.Winston J, Deray G, Hawkins T, Szczech L, Wyatt C, Young B: Kidney disease in patients with HIV infection and AIDS. Clin Infect Dis 47: 1449–1457, 2008 [DOI] [PubMed] [Google Scholar]
- 2.Deeks SG, Phillips AN: HIV Infection, antiretroviral treatment, aging, and non-AIDS related morbidity. BMJ 338: 288–292, 2009 [DOI] [PubMed] [Google Scholar]
- 3.Gupta SK, Eustace JA, Winston JA, Boydstun II, Ahuja TS, Rodriquez RA, Tashima KT, Roland M, Franceschini N, Palella FJ, Lennox JL, Klotman PE, Nachman SA, Hall SD, Szczech LA: Guidelines for the management of chronic kidney disease in HIV-infected patients: Recommendations of the HIV Medicine Association of the Infectious Diseases Society of America. Clin Infect Dis 40: 1559–1585, 2005 [DOI] [PubMed] [Google Scholar]
- 4.Campbell LJ, Ibrahim F, Fisher M, Holt SG, Hendry BM, Post FA: Spectrum of chronic kidney disease in HIV-infected patients. HIV Med 10: 329–336, 2009 [DOI] [PubMed] [Google Scholar]
- 5.Fine DM, Garg N, Haas M, Rahman MH, Lucas GM, Scheel PJ, Atta MG: Cocaine use and hypertensive renal changes in HIV-infected individuals. Clin J Am Soc Nephrol 2: 1125–1130, 2007 [DOI] [PubMed] [Google Scholar]
- 6.Jaffe JA, Kimmel PL: Chronic nephropathies of cocaine and heroin abuse: A critical review. Clin J Am Soc Nephrol 1: 655–667, 2006 [DOI] [PubMed] [Google Scholar]
- 7.Vlahov D, Anthony JC, Munoz A, Margolick J, Nelson Ke, Celantano DD, Solomon L, Polk BF: The ALIVE Study, a longitudinal study of HIV-1 infection in intravenous drug users: Description of methods and characteristics of participants. NIDA Res Monogr 109: 75–100, 1991 [PubMed] [Google Scholar]
- 8.Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S, Kusek JW, Van Lente F: Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med 145: 247–254, 2006 [DOI] [PubMed] [Google Scholar]
- 9.National Kidney Foundation: KDOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification and stratification. Am J Kidney Dis 39 [suppl 1]: S1– S000, 2002 [PubMed] [Google Scholar]
- 10.Eknoyan G, Hostetter T, Bakris GL, Hebert L, Levey AS, Parving H-H, Steffes MW, Toto R: Proteinuria and other markers of chronic kidney disease: A position statement of the National Kidney Foundation (NKF) and the National Institute of Diabetes and Digestive Kidney Diseases (NIDDK). Am J Kidney Dis 42: 617–622, 2003 [DOI] [PubMed] [Google Scholar]
- 11.Barros AJD, Hirakata VN: Alternatives to logistic regression in cross-sectional studies: An empirical comparison of model that directly estimate the prevalence ratio. BMC Med Res Methodol 3: 21, 2003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ritz E: Proteinuria and vascular damage—The vicious twins. N Engl J Med 348: 2349–2352, 2003 [DOI] [PubMed] [Google Scholar]
- 13.De Lorenzo F, Collot-Teixeira S, Boffito M, Feher M, Gazzard B, McGregor JL: Metabolic-inflammatory changes, and accelerated atherosclerosis in HIV patients: Rationale for preventative measures. Curr Med Chem 15: 2991–2999, 2008 [DOI] [PubMed] [Google Scholar]
- 14.Rhee MS, Schmid CH, Stevens LA, Forrester JE: Risk factors for proteinuria in HIV-infected and -uninfected Hispanic drug users. Am J Kidney Dis 52: 683–690, 2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wyatt CM, Malvestutto C, Coca SG, Klotman PE, Parikh CR: The impact of hepatitis C virus co-infection on HIV-related kidney disease: A systematic review and meta-analysis. AIDS 22: 1799–1807, 2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, Van Lente F, Levey AS: Prevalence of chronic kidney disease in the United States. JAMA 298: 2038–2047, 2007 [DOI] [PubMed] [Google Scholar]
- 17.Garg AX, Kiberd BA, Clark WF, Haynes RB, Clase CM: Albuminuria and renal insufficiency prevalence guides population screening: Results from the NHANES III. Kidney Int 61: 2165–2175, 2002 [DOI] [PubMed] [Google Scholar]
- 18.Hemmelgarn BR, Manns BJ, Lloyd A, James MT, Klarenbach S, Quinn RR, Wiebe N, Tonelli M: Relation between kidney function, proteinuria, and adverse outcomes. JAMA 303: 423–429, 2010 [DOI] [PubMed] [Google Scholar]
- 19.Lucas GM, Mehta SH, Atta MG, Kirk GD, Galai N, Vlahov D, Moore RD: End-stage renal disease and chronic kidney disease in a cohort of African-American HIV-infected and at-risk HIV-seronegative participants followed between 1988 and 2004. AIDS 21: 2435–2443, 2007 [DOI] [PubMed] [Google Scholar]
- 20.Overton ET, Nurutidinova D, Freeman J, Seyfried W, Mondy KE: Factors associated with renal dysfunction within an urban HIV-infected cohort in the era of highly active antiretroviral therapy. HIV Med 10: 343–350, 2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Odden MC, Scherzer R, Bacchetti P, Szczech LA, Sidney S, Grunfeld C, Shlipak MG: Cystatin C Level as a Marker of Kidney Function in Human Immunodefiency Virus Infection. Arch Intern Med 167: 2213–2219, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Szczech LA, Grunfeld C, Scherzer R, Canchola JA, van der Horst C, Sidney S, Wohl D, Shlipak MG: Microalbuminuria in HIV infection. AIDS 21: 1003–1009, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Szczech LA, Menezes P, Quinlivan EB, van der Horst C, Bartlett JA, Svetkey LP: Microalbuminaria predicts overt proteinuria among patients with HIV infection. HIV Med January4, 2010. [epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bakir AA, Dunea G: Drugs of abuse and renal disease. Curr Opin Nephrol Hypertens 5: 122–126, 1996 [DOI] [PubMed] [Google Scholar]
- 25.Bemanian S, Motallebi M, Nosrati SM: Cocaine-induced renal infarction: Report of a case and review of the literature. BMC Nephrol 6: 10, 2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Galai N, Safaeian M, Vlahov D, Bolotin A, Celentano DD: Longitudinal patterns of drug injection behavior in the ALIVE Study cohort, 1988–2000: Description and determinants. Am J Epidemiol 158: 695–704, 2003 [DOI] [PubMed] [Google Scholar]
- 27.Atta MG, Fine DM, Kirk GD, Mehta SH, Moore RD, Lucas GM: Survival during renal replacement therapy among African Americans infected with HIV type 1 in urban Baltimore, Maryland. Clin Infect Dis 45: 1625–1632, 2007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Stevens LA, Levey AS: Measurement of kidney function. Med Clin North Am 89: 457–473, 2005 [DOI] [PubMed] [Google Scholar]
- 29.Coresh J, Astor BC, Greene T, Eknoyan G, Levey AS: Prevalence of chronic kidney disease and decreased kidney function in the adult US population: Third National Health and Nutrition Examination Survey. Am J Kidney Dis 41: 1–12, 2003 [DOI] [PubMed] [Google Scholar]
- 30.Macalino GE, Celentano DD, Latkin C, Strathdee SA, Vlahov D: Risk behaviors by audio computer-assisted self-interviews among HIV-seropositive and HIV-seronegative injection drug users. AIDS Educ Prev 14: 367–378, 2002 [DOI] [PubMed] [Google Scholar]