To estimate the association of antiretroviral therapy initiation with incident acquired immunodeficiency syndrome (AIDS) or death while accounting for time-varying confounding in a cost-efficient manner, the authors combined a case-cohort study design with inverse probability-weighted estimation of a marginal structural Cox proportional hazards model. A total of 950 adults who were positive for human immunodeficiency virus type 1 were followed in 2 US cohort studies between 1995 and 2007. In the full cohort, 211 AIDS cases or deaths occurred during 4,456 person-years. In an illustrative 20% random subcohort of 190 participants, 41 AIDS cases or deaths occurred during 861 person-years. Accounting for measured confounders and determinants of dropout by inverse probability weighting, the full cohort hazard ratio was 0.41 (95% confidence interval: 0.26, 0.65) and the case-cohort hazard ratio was 0.47 (95% confidence interval: 0.26, 0.83). Standard multivariable-adjusted hazard ratios were closer to the null, regardless of study design. The precision lost with the case-cohort design was modest given the cost savings. Results from Monte Carlo simulations demonstrated that the proposed approach yields approximately unbiased estimates of the hazard ratio with appropriate confidence interval coverage. Marginal structural model analysis of case-cohort study designs provides a cost-efficient design coupled with an accurate analytic method for research settings in which there is time-varying confounding.

Background

We previously reported an increased risk of all-cause and AIDS mortality among HIV-infected women with albuminuria (proteinuria or microalbuminuria) enrolled in the Women’s Interagency HIV Study (WIHS) prior to the introduction of highly active antiretroviral therapy (HAART).

Methods

The current analysis includes 1,073 WIHS participants who subsequently initiated HAART. Urinalysis for proteinuria and semi-quantitative testing for microalbuminuria from two consecutive study visits prior to HAART initiation were categorized as follows: confirmed proteinuria (both specimens positive for protein), confirmed microalbuminuria (both specimens positive with at least one microalbuminuria), unconfirmed albuminuria (one specimen positive for proteinuria or microalbuminuria), or negative (both specimens negative). Time from HAART initiation to death was modeled using proportional hazards analysis.

Results

Compared to the reference group of women with two negative specimens, the hazard ratio (HR) for all-cause mortality was significantly elevated for women with confirmed microalbuminuria (HR 1.9; 95% CI 1.2–2.9). Confirmed microalbuminuria was also independently associated with AIDS death (HR 2.3; 95% CI 1.3–4.3), while women with confirmed proteinuria were at increased risk for non-AIDS death (HR 2.4; 95% CI 1.2–4.6).

Conclusions

In women initiating HAART, pre-existing microalbuminuria independently predicted increased AIDS mortality, while pre-existing proteinuria predicted increased risk of non-AIDS death. Urine testing may identify HIV-infected individuals at increased risk for mortality even after the initiation of HAART. Future studies should consider whether these widely available tests can identify individuals who would benefit from more aggressive management of HIV infection and comorbid conditions associated with mortality in this population.

Background

In the early highly active antiretroviral therapy (HAART) era, kidney dysfunction was strongly associated with death among HIV-infected individuals. We re-examined this association in the later HAART period to determine whether chronic kidney disease (CKD) remains a predictor of death after HAART-initiation.

Methods

To evaluate the effect of kidney function at the time of HAART initiation on time to all-cause mortality, we evaluated 1415 HIV-infected women initiating HAART in the Women’s Interagency HIV Study (WIHS). Multivariable proportional hazards models with survival times calculated from HAART initiation to death were constructed; participants were censored at the time of the last available visit or December 31, 2006.

Results

CKD (eGFR <60 ml/min/1.73 m2) at HAART initiation was associated with higher mortality risk adjusting for age, race, hepatitis C serostatus, AIDS history and CD4+ cell count (hazard ratio [HR]=2.23, 95% confidence interval [CI]: 1.45–3.43). Adjustment for hypertension and diabetes history attenuated this association (HR=1.89, CI: 0.94–3.80). Lower kidney function at HAART initiation was weakly associated with increased mortality risk in women with prior AIDS (HR=1.09, CI: 1.00–1.19, per 20% decrease in eGFR).

Conclusions

Kidney function at HAART initiation remains an independent predictor of death in HIV-infected individuals, especially in those with a history of AIDS. Our study emphasizes the necessity of monitoring kidney function in this population. Additional studies are needed to determine mechanisms underlying the increased mortality risk associated with CKD in HIV-infected persons.

Background

The clinical importance of the association of HIV infection and antiretroviral therapy (ART) with low bone mineral density (BMD) in premenopausal women is uncertain because BMD stabilizes on established ART and fracture data are limited.

Methods

We measured time to first new fracture at any site with median follow-up of 5.4 years in 2391 (1728 HIV-infected, 663 HIV-uninfected) participants in the Women’s Interagency HIV Study (WIHS). Self-report of fracture was recorded at semiannual visits. Proportional hazard models assessed predictors of incident fracture.

Results

At baseline, HIV-infected women were older (40 ± 9 vs. 36 ± 10 years, P <0.0001), more likely to report postmenopausal status and be hepatitis C virus-infected, and weighed less than HIV-uninfected women. Among HIV-infected women, mean CD4+ cell count was 482 cells/μl; 66% were taking ART. Unadjusted incidence of fracture did not differ between HIV-infected and uninfected women (1.8 vs. 1.4/100 person-years, respectively, P = 0.18). In multivariate models, white (vs. African-American) race, hepatitis C virus infection, and higher serum creatinine, but not HIV serostatus, were statistically significant predictors of incident fracture. Among HIV-infected women, older age, white race, current cigarette use, and history of AIDS-defining illness were associated with incidence of new fracture.

Conclusion

Among predominantly premenopausal women, there was little difference in fracture incidence rates by HIV status, rather traditional risk factors were important predictors. Further research is necessary to characterize fracture risk in HIV-infected women during and after the menopausal transition.

Prevalence of microalbuminuria is increased in patients with HIV. Microalbuminuria is associated with increased mortality in other populations, including diabetics, for whom microalbuminuria testing is standard of care. We investigated whether microalbuminuria is associated with mortality in HIV-infected women not receiving antiretroviral therapy.

Methods

Urinalysis for proteinuria and semi-quantitative testing for microalbuminuria were performed in specimens from two consecutive visits in 1,547 HIV-infected women enrolled in the Women’s Interagency HIV Study in 1994–1995. Time to death was modeled using proportional hazards analysis.

Results

Compared to women without albuminuria, the hazard ratio (HR) for all-cause mortality was increased in women with one (HR 3.4; 95% CI 2.2–5.2) or two specimens positive for either proteinuria or microalbuminuria (HR 3.9; 95% CI 2.1–7.0). The highest risk was observed in women with both specimens positive for proteinuria (HR 5.8; 95% CI 3.4–9.8). The association between albuminuria and all-cause mortality risk remained significant after adjustment for demographics, HIV disease severity, and related comorbidities. Similar results were obtained for AIDS death.

Conclusions

We identified a graded relationship between albuminuria and the risk of all-cause and AIDS mortality.

We assessed the prevalence and predictors of latent Toxoplasma infection in a large group of human immunodeficiency virus (HIV)–infected and HIV-uninfected at-risk US women. The prevalence of latent Toxoplasma infection was 15% (380 of 2525 persons) and did not differ by HIV infection status. HIV-infected women aged ≥50 years and those born outside of the United States were more likely to have latent Toxoplasma infection, with prevalences of 32% and 41%, respectively.

Background

Opiate use is common in HIV- and hepatitis C virus (HCV)-infected individuals, however its contribution to the risk of diabetes mellitus is not well understood.

Methods

Prospective study of 1,713 HIV-infected and 652 uninfected participants from the Women’s Interagency HIV Study between October 2000 and March 2006. Diabetes defined as fasting glucose ≥126 mg/dl, or self-report of diabetes medication use or confirmed diabetes diagnosis. Opiate use determined using an interviewer-administered questionnaire. Detectable plasma HCV RNA confirmed HCV infection.

Results

Current opiate users had a higher prevalence of diabetes (15%) than non-users (10%, p=.03), as well as a higher risk of incident diabetes (adjusted relative hazard [RHadj] 1.58, 95% CI 1.01, 2.46), after controlling for HCV infection, HIV/antiretroviral therapy status and diabetes risk factors including age, race/ethnicity, family history of diabetes and body mass index. HCV infection was also an independent risk factor for diabetes (RHadj 1.61, 95% CI 1.02, 2.52). HCV-infected women reporting current opiate use had the highest diabetes incidence (4.83 cases/100 person-years).

Conclusions

Among women with or at-risk for HIV, opiate use is associated with increased diabetes risk independently of HCV infection. Diabetic screening should be part of care for opiate users, and those infected with HCV.

Typical applications of marginal structural time-to-event (e.g., Cox) models have used time on study as the time scale. Here, the authors illustrate use of time on treatment as an alternative time scale. In addition, a method is provided for estimating Kaplan-Meier–type survival curves for marginal structural models. For illustration, the authors estimate the total effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome (AIDS) or death in 1,498 US men and women infected with human immunodeficiency virus and followed for 6,556 person-years between 1995 and 2002; 323 incident cases of clinical AIDS and 59 deaths occurred. Of the remaining 1,116 participants, 77% were still under observation at the end of follow-up. By using time on study, the hazard ratio for AIDS or death comparing always with never using highly active antiretroviral therapy from the marginal structural model was 0.52 (95% confidence interval: 0.35, 0.76). By using time on treatment, the analogous hazard ratio was 0.44 (95% confidence interval: 0.32, 0.60). In time-to-event analyses, the choice of time scale may have a meaningful impact on estimates of association and precision. In the present example, use of time on treatment yielded a hazard ratio further from the null and more precise than use of time on study as the time scale.

Background

Low bone mineral density (BMD) has been reported in HIV + women, but less is known about the longitudinal evolution of BMD and fracture incidence.

Methods

In 100 HIV+ and 68 HIV− premenopausal women in the Women’s Interagency HIV Study (WIHS), BMD was measured by dual energy x-ray absorptiometry at the femoral neck (FN) and lumbar spine (LS) at index visit and after a median of 2.5 years.

Results

In HIV+ women, BMD at index visit was normal but 5% lower at the LS and FN than in HIV− women. Annual percent decrease in BMD did not differ between HIV+ and HIV− women at the LS (−0.8±0.2% vs −0.4±0.2%, p=0.20) or FN (−0.8±0.3% vs −0.6±0.3%, p=0.56), and remained similar after adjustment for age, weight, and BMD at index visit. Among HIV+ women, bone loss was associated with vitamin D deficiency and opiate use but not with use or class of antiretrovirals. Incidence of self-reported fracture was 0.74/100 person-years in HIV+ women, and similar in HIV− women.

Conclusions

In premenopausal HIV+ women, index BMD was lower than comparable HIV− women; however, rates of bone loss at the LS and FN were similar over 2.5 years of observation, irrespective of ART.

To estimate the net effect of imperfectly measured highly active antiretroviral therapy on incident acquired immunodeficiency syndrome or death, the authors combined inverse probability-of-treatment-and-censoring weighted estimation of a marginal structural Cox model with regression-calibration methods. Between 1995 and 2007, 950 human immunodeficiency virus–positive men and women were followed in 2 US cohort studies. During 4,054 person-years, 374 initiated highly active antiretroviral therapy, 211 developed acquired immunodeficiency syndrome or died, and 173 dropped out. Accounting for measured confounders and determinants of dropout, the weighted hazard ratio for acquired immunodeficiency syndrome or death comparing use of highly active antiretroviral therapy in the prior 2 years with no therapy was 0.36 (95% confidence limits: 0.21, 0.61). This association was relatively constant over follow-up (P = 0.19) and stronger than crude or adjusted hazard ratios of 0.75 and 0.95, respectively. Accounting for measurement error in reported exposure using external validation data on 331 men and women provided a hazard ratio of 0.17, with bias shifted from the hazard ratio to the estimate of precision as seen by the 2.5-fold wider confidence limits (95% confidence limits: 0.06, 0.43). Marginal structural measurement-error models can simultaneously account for 3 major sources of bias in epidemiologic research: validated exposure measurement error, measured selection bias, and measured time-fixed and time-varying confounding.

Abstract

Since the introduction of highly active antiretroviral therapy (HAART) and the subsequent increased life expectancy in HIV-infected persons, non-HIV–related diseases have become an important cause of morbidity and mortality. This cross-sectional study reports the prevalence of overweight and obesity, and sociodemographic, psychological, and substance use-related risk factors for elevated body mass index (BMI) among 2157 HIV-seropositive (HIV+) in comparison to 730 HIV-seronegative (HIV−) participants in the Women's Interagency HIV Study (WIHS). Separate univariable and multivariate linear regression analyses were completed for HIV+ and HIV− women. Our study revealed a similar proportion of obesity (body mass index [BMI] ≥30) among HIV+ (33%) and HIV− women (29%) (p = 0.12), as well as comparable median BMI (HIV+: 26.1 versus HIV−: 26.7, p = 0.16). HIV+ compared to HIV− women, respectively, were significantly (p < 0.01) older (median = 35.6 versus. 32.5), but similar (p = 0.97) by race/ethnicity (57% African American, 28% Hispanic, and 15% white for both). In multivariate models for both HIV+ and HIV− women, African American race/ethnicity was significantly (p < 0.05) associated with higher BMI, while higher quality of life score and illicit hard drug use were associated with lower BMI. Additionally, smoking, alcohol use, markers of advanced HIV infection (AIDS diagnosis, elevated HIV viral load, low CD4 count), and a history of antiretroviral therapy use (ART) were also associated with lower BMI among HIV+ women. In conclusion, risk factors for elevated BMI were similar for HIV+ and HIV− women in the WIHS. For HIV+ women, all markers of advanced HIV infection and ART use were additionally associated with lower BMI.

Introduction

The use of hormonal contraception (HC) is increasing in HIV-infected women. Both HC and HIV infection have been associated with adverse metabolic outcomes. We investigated the association of progestin-only and combined (estrogen/progestin) HC with disorders of glucose and lipid metabolism in HIV-infected and uninfected women.

Methods

Linear mixed models evaluated the association of HC type with fasting HDL, LDL, triglycerides, the homeostasis model assessment estimate of insulin resistance (HOMA-IR), and glucose in 885 HIV-infected and 408 HIV-uninfected women from the Women's Interagency HIV Study seen between October 2000 and September 2005.

Results

Compared to non-HC users, progestin-only HC was independently associated with lower HDL (-3mg/dL;95% confidence interval[CI]:-5,-1 in HIV-infected and -6mg/dL;95% CI:-9,-3 in HIV-uninfected women), greater HOMA (+0.86;95% CI:0.51,1.22 and +0.56;95% CI:0.12,1.01). Combined HC was associated with higher HDL(+5mg/dL;95% CI:2,7 and +5mg/dL;95% CI:3,7).

Conclusion

Progestin–only HC is associated with lower HDL and greater HOMA-IR than non-HC users. Combined HC may be preferred in HIV-infected women of reproductive age at risk for cardiovascular disease, but interactions with antiretroviral therapy that may impair contraceptive efficacy have been reported. Alternative HC methods that minimize adverse outcomes but maintain efficacy require further study.

Objective

Obesity is common in women and associated with a number of adverse health outcomes including cardiovascular disease, infectious diseases, and cancer. We explore the relationship between obesity and immune cell counts in women.

Design

Longitudinal study of 322 women from 1999 through 2003 enrolled as HIV-negative comparators in the Women’s Interagency HIV Study.

Methods

Body mass index (BMI) was categorized as normal weight (BMI 18.5 - 24.9), overweight (BMI 25 - 29.9), obese (BMI 30 - 34.9), and morbid obesity (BMI ≥35). CD4 and CD8 counts and percents, total lymphocyte and white blood cell (WBC) counts were measured annually using standardized techniques. A mixed model repeated measures analysis was performed using an autoregressive correlation matrix.

Results

At the index visit, 61% of women were African-American; mean age was 35 years, and median BMI was 29 kg/m2. Immunologic parameters were in the normal range (median CD4 count: 995 cells/mm3; CD8 count: 488 cells/mm3; total lymphocyte count: 206 cells/mm3; median WBC: 6 × 103 cells/mm3). In multivariate analyses, being overweight, obese or morbidly obese were independently associated with higher CD4, total lymphocyte, and WBC counts than being normal weight; morbid obesity was associated with a higher CD8 count. The strongest associations between body weight and immune cell counts were demonstrated in the morbidly obese.

Conclusion

Increasing body weight is associated with higher CD4, CD8, total lymphocyte, and WBC counts in women. Investigation into the impact of obesity on immune function and long term adverse outcomes is needed.