Incidence of Statin Initiation Among People With and Without HIV in the United States: A Prospective Observational Study

Abstract

Background

To reduce atherosclerotic cardiovascular disease (ASCVD) among US people with human immunodeficiency virus (HIV, PWH), it is critical to assess shortfalls in statin initiation. We aimed to describe patterns in clinically indicated statin initiation among demographically similar people with or at increased vulnerability to HIV.

Methods

Using data from the Multicenter AIDS Cohort Study and Women's Interagency HIV Study, we followed 842 men and 852 women with an indication for statin use based on 2013 American College of Cardiology/American Heart Association guidelines for statin initiation between January 2014 and March 2020. We estimated 2-year incidence of statin initiation stratified by demographic, clinical, and behavioral characteristics and compared estimates using incidence differences.

Results

Within 2 years of statin indication, 20% of participants reported statin use. Initiation of statin therapy did not differ significantly by HIV status. However, initiation was lower among Black versus non-Black persons, especially among men (14.0% versus 22.3%; difference, −8.3%; 95% confidence interval, −13.8% to −2.8%). Compared with initiation among persons indicated based only on ≥7.5% 10-year predicted ASCVD risk (incidence: men, 14.2%; women, 15.2%), initiation was higher among those with existing ASCVD (incidence: men, 32.9%; women, 22.0%) or diabetes (incidence: men, 26.4%; women, 24.5%). Initiation was lower among current versus noncurrent smokers and higher among those with comorbidities.

Conclusions

Guideline-driven statin initiation was low, with large inequities by race. With indications for statin use expanding for PWH, redressing barriers to guideline implementation will be crucial to achieving improved outcomes through uptake of these evidence-based therapies.

Statin initiation was low among US people with or at increased vulnerability to human immunodeficiency virus. Within 2 years of eligibility, 80% failed to receive a statin. Incidence varied by clinical factors and was lower among Black versus non-Black participants.

Atherosclerotic cardiovascular disease (ASCVD) contributes to a disproportionate burden of premature morbidity and mortality among people with human immunodeficiency virus (HIV, PWH) [1–3] as well as people from historically marginalized racial and ethnic groups [4–6]. Dyslipidemia, hypertension, and diabetes are highly prevalent among both populations and have long been recognized as major risk factors for ASCVD [4, 7, 8]. Additionally, for PWH, HIV-associated inflammation and immune dysregulation persist after successful viral suppression [9] and may further increase risk of myocardial infarction, stroke, and heart failure [8].

Statin therapy is an effective intervention to reduce the risk of ASCVD in the general population [10–12], and recent evidence indicates benefits for PWH [13]. Historically, US guidelines emphasized statin use for cholesterol reduction [14]. However, guidelines updated by the American College of Cardiology/American Heart Association (ACC/AHA) at the end of 2013 shifted recommendations from using statins to achieve specific low-density lipoprotein (LDL) targets to using statins more broadly as preventive therapy based on 4 statin-benefit group profiles [15]. While this change increased the estimated number of US adults eligible for statin therapy by 1.3-fold [16] and US PWH by 2- to 3-fold [17, 18], subsequent incidence of statin initiation among demographically similar people with and without HIV is unknown. A small number of US studies have examined statin use among select populations of PWH after eligibility expanded but have mainly relied on cross-sectional data [19–22].

Longitudinal studies that characterize incidence of statin initiation after clinical indication for statin use are critical to assess guideline implementation and identify shortfalls in timely initiation of this evidence-based therapy. This information can provide important insight as recent policy changes have further expanded statin eligibility among PWH [23] in light of trial results that demonstrate substantial benefits of statin use at lower thresholds of ASCVD risk [13]. Using data from the Multicenter AIDS Cohort Study (MACS) and Women's Interagency HIV Study (WIHS) after 2013 ACC/AHA guideline publication, we estimated the incidence of statin initiation within 2 years of clinical indication for statin use. We quantified differences in incidence by demographic, clinical, and behavioral risk factors.

METHODS

Study Population

The MACS [24] and WIHS [25], now the MACS/WIHS Combined Cohort Study (MWCCS) [26], were interval cohort studies that followed PWH and demographically similar people with increased vulnerability to HIV across multiple US sites since 1984 and 1994, respectively. MACS enrolled 7358 men who had sex with men at 4 sites (Los Angeles, CA; Baltimore, MD; Chicago, IL; Pittsburgh, PA) over 4 recruitment periods (1984–1985, 1987–1990, 2001–2003, 2010–2017). WIHS enrolled 4982 women at 10 sites (Bronx/Manhattan, NY; Brooklyn, NY; Chicago, IL; Los Angeles, CA; San Francisco, CA; Washington, DC; Atlanta, GA; Birmingham, AL/Jackson, MS; Chapel Hill, NC; Miami, FL) over 4 recruitment periods (1994–1995, 2001–2002, 2011–2012, 2013–2015). Participants attended study visits approximately every 6 months, contributing sociodemographic, behavioral, clinical, and medication data through in-depth interviews, laboratory testing, and physical exams. Sites confirmed deaths through linkage to state death registries and the National Death Index. Institutional review boards at each site approved the MACS and WIHS, and all participants provided written informed consent to participate.

Eligibility for the present study was restricted to persons who participated in the cohorts after the 2013 ACC/AHA guidelines were released (ie, ≥1 follow-up visit after December 2013). Because we were interested in incidence of statin initiation following expanded criteria for statin use, participants who reported statin use at their study visit preceding the 2013 ACC/AHA guidelines were excluded. In our final sample, we included participants with an indication for statin use within the approximate 4 years following guideline updates.

Ascertainment of Clinical Indication for Statin Initiation

ACC/AHA guidelines outlined the following 4 statin-benefit groups: persons aged ≥21 years with existing ASCVD, persons with LDL cholesterol ≥190 mg/dL, persons aged 40–75 years with diabetes, and persons aged 40–75 years with ≥7.5% 10-year predicted risk of ASCVD estimated with the pooled cohort equations [15]. Guidelines depicted statin-benefit groups hierarchically; membership in group 2 was defined among persons who did not meet criteria for group 1, membership in group 3 was defined among persons who did not meet criteria for group 2 or group 1, and so on.

In implementing this framework, we assessed statin indications among participants at the visit interval that coincided with the period in which the 2013 ACC/AHA guidelines were released (October 2013–March 2014) through the interval ending in March 2018, allowing approximately 2 years of possible follow-up before April 2020. (Reenrollment in the MWCCS began in October 2020 [26].) We classified participants as having existing ASCVD based on self-report of previous myocardial infarction, stroke, or revascularization. Diabetes was based on 2 fasting glucose measures ≥126 mg/dL, fasting glucose ≥126 mg/dL and hemoglobin A1c ≥6.5%, or self-report of antidiabetic medication use. We used participant age, sex, race, serum cholesterol, systolic blood pressure, antihypertensive use, cigarette smoking, and diabetes to calculate 10-year ASCVD risk with the pooled cohort equations formula [15]. Because the formula requires nonmissing values, we addressed incomplete data as follows. In cases of missing cholesterol, values in WIHS were most often missing by design because lipid testing was only performed every other visit; we carried forward previous cholesterol values at visits when no testing was scheduled. In other instances of sporadic missing information for cholesterol, blood pressure, antihypertensive use, or smoking, we carried forward values from the study visit approximately 6 months prior.

Outcome Ascertainment

Our outcome of interest was statin initiation within approximately 2 years (ie, 4 study visits) of meeting 2013 ACC/AHA guideline criteria. Statin use was self-reported by participants at visits. We considered participants newly reporting use of a statin to have initiated statin therapy.

Statistical Analyses

We followed participants from the first visit at which they had an indication for statin use until reported statin use, death, administrative censoring after 4 follow-up visits, or the last follow-up visit before April 2020, whichever occurred first. Using the Aalen–Johansen estimator of the cumulative incidence function [27], we estimated 2-year incidence of statin initiation, accounting for the competing risk of death.

To examine differences in statin initiation, we stratified analyses by statin-benefit group, race, and risk factors at time of statin indication. Namely, we stratified by the presence versus absence of each of the following ASCVD-associated comorbidities: hypertension [28], dyslipidemia (ie, total cholesterol ≥200 mg/dL, fasting LDL ≥130 mg/dL, high-density lipoprotein <40 mg/dL, fasting triglycerides ≥150 mg/dL, or use of nonstatin lipid-lowering medication with self-report of previous clinical diagnosis), metabolic syndrome [29], and obesity; substance use indicators including current cigarette smoking, self-reported use of crack or cocaine within the last 6 months, and binge drinking [30]; and age. We contrasted stratum-specific estimates using incidence differences and calculated 95% confidence intervals (CIs) using the delta method. Analyses were performed using SAS version 9.4 (SAS Institute, Inc, Cary, NC).

RESULTS

Of 2445 men and 2387 women who participated in MACS or WIHS after 2013 ACC/AHA guideline updates, 520 (21%) men and 237 (10%) women were excluded for prevalent statin use, 169 (7%) men and 194 (9%) women were excluded for missing clinical data needed to ascertain statin eligibility, and 842 (34%) men and 852 (36%) women met the indication for statin use at visits between October 2013 and March 2018 and were included in the analyses (Supplementary Figure 1). The median age of the analytic sample was 58 years (interquartile range [IQR], 51–63) among men and 52 years (IQR, 47–57) among women, 33% of men and 79% of women were Black, and 49% of men and 69% of women were diagnosed with HIV (Table 1). Most were hypertensive (men, 58%; women, 73%) and/or had dyslipidemia (men, 71%; women, 66%) at time of indication. Among men, 14% were indicated based on existing ASCVD, 4% based on LDL ≥190 mg/dL, 21% based on diabetes, and 61% based only on ≥7.5% 10-year predicted ASCVD risk of ASCVD. Among women, 26% were indicated based on existing ASCVD, 4% based on LDL ≥190 mg/dL, 33% based on diabetes, and 37% based only on ≥7.5% 10-year predicted ASCVD risk. The 2-year incidence of statin initiation was 19.6% (95% CI, 17.1% to 22.5%) for statin-eligible men and 20.6% (95% CI, 17.7% to 24.0%) for statin-eligible women (Figure 1).

Table 1.

Characteristics of 842 Multicenter AIDS Cohort Study Men and 852 Women's Interagency HIV Study Women Indicated for Statin Use Based on 2013 American College of Cardiology/American Heart Association Guidelines

Characteristic	Multicenter AIDS Cohort Study		Women's Interagency HIV Study
	N = 842		N = 852
Age, median [IQR], y	58 [51, 63]		52 [47, 57]
Black race	278	(33)	677	(79)
Hispanic ethnicity	73	(9)	130	(15)
Diagnosed with HIV	416	(49)	592	(69)
Hypertensiona	489	(58)	619	(73)
Metabolic syndromeb	290	(34)	191	(22)
Dyslipidemiac	617	(73)	565	(66)
Obesityd	203	(24)	454	(53)
Current smoking	273	(32)	447	(52)
Crack or cocaine usee	82	(9)	80	(9)
Binge drinkingf	51	(6)	100	(12)
Depressiong	251	(31)	271	(32)
Statin-benefit group	…	…	…	…
Existing ASCVD	117	(14)	219	(26)
Low-density lipoprotein ≥190 mg/dL	30	(4)	38	(4)
Diabetesh	178	(21)	284	(33)
≥7.5% 10-year predicted ASCVD risk	517	(61)	311	(37)
ASCVD risk, median [IQR], %i	10.5 [8.4, 15.1]		9.8 [8.3, 12.7]

Indication is based on the 2013 American College of Cardiology and American Heart Association Guideline on the Treatment of Blood Cholesterol to Reduce Atherosclerotic Cardiovascular Risk in Adults, which recommended statin use for the following groups: those with existing ASCVD, those with low-density lipoprotein (LDL) ≥190 mg/dL, those aged 40–75 years with diabetes, those aged 40–75 years with ≥7.5% 10-year predicted risk of ASCVD estimated with the pooled cohort equations formula. The study population comprises men and women who met 1 of the 4 criteria at study visits between October 2013 and March 2018 and were followed for approximately 2 years (ie, 4 visits) until the first of the following: reported statin use, death, administrative censoring after 4 visits, or last visit before April 2020.

Abbreviations: ASCVD, atherosclerotic cardiovascular disease; HDL, high density lipoprotein; HIV, human immunodeficiency virus; IQR, interquartile range.

^aSystolic blood pressure ≥140 mmHg, diastolic blood pressure ≥90 mmHg, self-reported hypertension, or use of antihypertensive medications.

^bEvidence of meeting at least 3 of the following diagnostic criteria: waist circumference >88 cm for women, >102 cm for men; fasting triglycerides ≥150 mg/dL or fibrates or niacin medications with diagnosis; fasting glucose ≥110 mg/dL or antidiabetics with diagnosis; high-density lipoprotein (HDL) <50 mg/dL for women, <40 mg/dL for men, or niacin medications with diagnosis; systolic blood pressure ≥130 mmHg, diastolic blood pressure ≥85 mmHg, or antihypertensives with diagnosis.

^cTotal cholesterol ≥200 mg/dL, fasting LDL ≥130 mg/dL, HDL <40 mg/dL, fasting triglycerides ≥150 mg/dL, or use of lipid-lowering medication with diagnosis.

^dBody mass index ≥30 kg/m²; missing for 20 men and 7 women.

^eSelf-reported use of crack or cocaine in last 6 months.

^fSelf-reported consumption of ≥5 drinks for men or ≥4 drinks for women at 1 sitting in last 6 months.

^gScore ≥16 on the Center for Epidemiologic Studies Depression Scale; missing for 27 men and 8 women.

^hTwo fasting glucose measurements ≥126 mg/dL, fasting glucose ≥126 mg/dL and hemoglobin A1c ≥6.5%, or ever self-reported antidiabetic medication use.

ⁱMedian ASCVD risk among those whose statin-benefit group was ≥7.5% 10-year predicted ASCVD risk.

Figure 1.

Cumulative incidence (95% confidence interval) of statin initiation within 2 years of clinical indication for statin use among Multicenter AIDS Cohort Study statin-eligible men (A) and Women's Interagency HIV Study statin-eligible women (B).

Figure 2 and Supplementary Table 1 present cumulative incidence of statin initiation by statin-benefit group, excluding persons indicated based on LDL ≥190 mg/dL due to small sample size (men, n = 30; women, n = 38). Two-year incidence was highest among men indicated based on existing ASCVD (incidence, 32.9%; 95% CI, 24.5% to 41.4%) and lowest among men indicated based only on ≥7.5% 10-year predicted ASCVD risk (incidence, 14.2%; 95% CI, 11.6% to 16.9%). Among women, incidence was highest among those indicated based on diabetes (incidence, 24.5%; 95% CI, 20.1% to 28.8%) and lowest among those indicated based only on ≥7.5% 10-year predicted ASCVD risk (incidence, 15.2%; 95% CI, 11.1% to 19.4%).

Figure 2.

Cumulative incidence (95% confidence interval) of statin initiation within 2 years of clinical indication for statin use among Multicenter AIDS Cohort Study statin-eligible men (A) and Women's Interagency HIV Study statin-eligible women (B) by criterion for meeting indication for statin use.

Incidence of statin initiation was lower among Black versus non-Black participants, particularly for men (Figure 3). Notably, the racial disparity among men increased across the study period. At 6 months following an indication for statin use, incidence of statin initiation was 4.8% points lower (95% CI, −8.0% to −1.6%) among Black versus non-Black men, whereas at 2 years, that disparity had increased in magnitude to 8.3% points lower (95% CI, −13.2% to −2.8%).

Figure 3.

Cumulative incidence (top panels) and incidence difference (bottom panels) of statin initiation within 2 years of clinical indication for statin use among Black versus non-Black Multicenter AIDS Cohort Study statin-eligible men (A) and Women's Interagency HIV Study statin-eligible women (B).

Results that contrasted statin initiation among those with versus without specific risk factors revealed no significant difference in 2-year incidence of statin initiation by HIV (Figure 4, Supplementary Table 1). However, 2-year incidence was substantially higher among men and women with versus without hypertension, with versus without dyslipidemia, with versus without metabolic syndrome, and with versus without obesity. Two-year incidence of statin initiation was lower among men and women who reported smoking versus no current smoking, recent crack/cocaine use, and binge drinking. The difference was more pronounced among men for smoking (difference, −5.9%; 95% CI, −11.3% to −.4%) and crack/cocaine use (difference, −8.1%; 95% CI, −15.6% to −.6%) and among women for binge drinking (difference, −7.8%; 95% CI, −15.4% to −.1%). While incidence of statin initiation was higher with increasing age among men, incidence was highest among women aged 50–60 years versus women aged <50 or >60 years (Supplementary Figure 2).

Figure 4.

Difference in the 2-year incidence of statin initiation among Multicenter AIDS Cohort Study statin-eligible men (A) and Women's Interagency HIV Study statin-eligible women (B) with versus without clinical and behavioral risk factors at time of clinical indication for statin use.

DISCUSSION

In longitudinal analyses that characterized statin initiation among participants in the MACS and WIHS following 2013 ACC/AHA guideline expansion of eligibility criteria, we found extremely low incidence of statin initiation; 80% of men and women who were eligible under 2013 guidelines failed to receive guideline-recommended statin therapy within 2 years of clinical indication. Moreover, the 2-year incidence of statin initiation was especially low among persons for whom the statin use indication was having ≥7.5% 10-year predicted ASCVD risk versus existing ASCVD or diabetes, those without versus with traditionally recognized ASCVD-associated comorbidities, those with versus without recent substance use, and Black versus non-Black persons.

While prior studies suggest suboptimal utilization of statins, frequency of guideline-driven new statin initiation among PWH or at increased vulnerability to HIV following the 2013 ACC/AHA guideline changes has not been extensively examined. Generally, studies have quantified proportions of statin-eligible people on therapy and used previous National Cholesterol Education Program Adult Treatment Panel (ATP) III guidelines to determine eligibility [31], data collected prior to 2013 ACC/AHA updates [32–34], or cross-sectional designs that included data from arbitrary time points after guideline updates [19–21, 35–37]. One previous study estimated statin incidence in WIHS but was restricted to the period before 2013 ACC/AHA guideline updates [17]. More recently, Coburn et al estimated annual proportions of statin initiation among statin-indicated PWH enrolled in clinical cohorts that contributed data to the North American AIDS Cohort Collaboration on Research and Design but cautioned that estimates should be considered cross-sectional [22].

To achieve maximum reductions in ASCVD morbidity and mortality, early initiation of preventive statin therapy may be necessary, especially among PWH for whom ASCVD risk is often underestimated [38]. In the Randomized Trial to Prevent Vascular Events in HIV (REPREIVE), PWH treated with pitavastatin at lower thresholds of predicted ASCVD risk than recommended for statin use by ACC/AHA guidelines experienced substantial therapeutic benefits and decreased risk of major adverse ASCVD events and death [13]. Consequently, guidelines published in 2024 recommend statins for PWH at lower risk thresholds [23], greatly expanding the number of PWH eligible for statin initiation once again.

Thus, our work makes an important contribution to existing literature by offering evidence of shortfalls in timely provision of clinically indicated preventive therapy that may be critical to optimally improve health outcomes among persons with or at increased vulnerability to HIV. Using longitudinal data from 2 of the largest cohort studies of people with or at increased vulnerability to HIV in the United States, we were able to follow individuals from new indication for statin use to estimate 2-year incidence of statin initiation and quantify differences in incidence by numerous characteristics. The Aalen–Johansen estimator allowed us to appropriately account for the competing risk of death prior to statin initiation [27, 39], which was nonnegligible in WIHS and could result in overestimating incidence if deaths were instead censored. Furthermore, the MACS and WIHS include participants from multiple regions in the Northeast, Southeast, Midwest, and West [26], improving generalizability of findings to US PWH compared with previous studies that have predominately included patients from limited regions or healthcare systems [19, 20, 31, 35]. Additionally, the cohorts have successfully enrolled Black and Hispanic participants, providing a study population whose racial/ethnic distribution is similar to that of the US population of PWH [26].

The mere approximately 15% incidence of statin initiation among persons indicated based on 10-year predicted ASCVD risk may be influenced by lag time between guideline publication and implementation. This component of determining statin eligibility represented a major paradigm shift from previous clinical practices, and it is possible that providers were not using the pooled cohort equations calculator to guide treatment decisions. Surveys conducted among US providers found that approximately 50% were aware of the ≥7.5% 10-year risk threshold for statin use [40] 1 year following guideline publication and approximately 70% after 2 years [41]. However, in WIHS, we did not find notable differences in 2-year incidence of statin initiation between women indicated for statin use within the first 2 years of guideline updates versus women indicated afterward (Supplementary Figure 3). Additionally, among 20% of MACS participants who would have also met previous ATP III guideline eligibility criteria [14], incidence was only 2% higher compared with those solely indicated based on 2013 ACC/AHA criteria (Supplementary Figure 4).

Participants might have initiated statins based predominately on presentation of traditionally recognized risk factors for ASCVD (ie, dyslipidemia, hypertension, metabolic syndrome, obesity) given significantly higher incidence of initiation among those with versus without each comorbidity. Additionally, incidence was higher with increasing number of comorbidities (Supplementary Table 2). Nonetheless, two-thirds of men and more than three-quarters of women who reported a previous myocardial infarction, stroke, or revascularization event failed to receive therapy.

The racial disparities we observed are consistent with previous findings [20, 22, 31, 35, 36]. Supplementary analyses revealed that racial disparities were even more pronounced among men with HIV, men with diabetes, women with hypertension, and women with a statin use indication based on 10-year predicted ASCVD risk (Supplementary Table 3). This is concerning given that ASCVD disproportionately affects historically marginalized racial and ethnic groups, with Black women experiencing the highest death rates from ASCVD [4]. Reducing ASCVD-related morbidity and mortality will require special attention to inequitable receipt of preventive therapy in addition to increasing timely statin initiation among all people indicated for statin use.

Though lack of health insurance and inaccessible healthcare can be substantial barriers to uptake of clinical treatments, this did not appear to be a major driver of the low incidence of statin initiation observed in our study population. Approximately 95% of men and women in our study reported health insurance or use of other prescription medications (eg, antihypertensives, antidiabetics; Supplementary Table 4). Additionally, 82% of men and 89% of women reported having seen a medical provider in an outpatient setting in the past year. Thus, low statin initiation may have been driven by challenges previously highlighted by primary care providers [42], including uncertainty regarding frequency of CVD risk assessment and how to communicate risk to patients, patient preferences, and competing priorities when other comorbidities, psychosocial factors, or substance use require attention. It is possible that substance use was targeted for intervention prior to (or instead of) statin initiation, which is consistent with the lower statin initiation we observed among men who reported current smoking, men who reported recent crack or cocaine use, and women who reported binge drinking.

Several limitations should be considered in the interpretation of our findings. First, we only had data on self-reported medication use, which may not capture all medications prescribed and may be subject to outcome misclassification if, for example, participants were not aware they were taking a statin. However, in addition to self-reporting medications, participants were explicitly asked if they were taking a cholesterol-lowering medication, which provided another opportunity to determine statin use. Second, we did not have data on statin dosage to determine whether those who initiated therapy were receiving the correct statin intensity. Studies in select populations of PWH reported that only 30%–57% of those prescribed statins received appropriate dose intensities [35, 37]. Consequently, it is likely that incidence of initiating appropriate statin regimens was substantially lower than our observed 20% incidence of initiating any statin regimen. Additionally, updated ACC/AHA guidelines released at the end of 2018 highlighted HIV as a “risk enhancer” and further emphasized moderate-intensity statin use among PWH with a 7.5%–20% 10-year predicted ASCVD risk [43]. Incidence of statin initiation may have improved after 2018, but we did not have a large enough sample or follow-up time to assess this in the MACS and WIHS. However, while Pan et al noted a potential increase in statin prescriptions among PWH after 2018, they found that most statin-eligible PWH still did not receive statins [44]. Finally, “existing ASCVD” was based on self-report and may be subject to measurement error as we did not have adjudicated event data.

Overall, our findings illustrate substantial room for improvement in ASCVD prevention among PWH and demographically similar people at increased vulnerability to HIV, especially among Black people. Further research to elucidate points at which barriers to statin initiation are occurring and actively challenge racial inequities will be crucial to designing effective interventions to improve guideline implementation and reduce ASCVD-related morbidity and mortality. With statin eligibility expanding among PWH [23], it is imperative to ensure initiation of evidence-based preventive therapy without delay.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. The authors thank the study participants for their contributions and the dedication of the staff at the Multicenter AIDS Cohort Study (MACS)/ Women's Interagency HIV Study (WIHS) Combined Cohort Study (MWCCS) sites. The authors also thank Dr Adaora A. Adimora, who provided substantial guidance, mentorship, and contributions to this work including feedback on earlier abstracts and results.

Data for this article were collected by the MACS and WIHS, now the MWCCS. MWCCS (principal investigators): Atlanta Clinical Research Center (CRS) (Ighovwerha Ofotokun, Anandi Sheth, and Gina Wingood), U01-HL146241; Baltimore CRS (Todd Brown and Joseph Margolick), U01-HL146201; Bronx CRS (Kathryn Anastos, David Hanna, and Anjali Sharma), U01-HL146204; Brooklyn CRS (Deborah Gustafson and Tracey Wilson), U01-HL146202; Data Analysis and Coordination Center (Gypsyamber D'Souza, Stephen Gange, and Elizabeth Topper), U01-HL146193; Chicago–Cook County CRS (Mardge Cohen, Audrey French, and Ryan Ross), U01-HL146245; Chicago–Northwestern CRS (Steven Wolinsky, Frank Palella, and Valentina Stosor), U01-HL146240; Northern California CRS (Bradley Aouizerat, Jennifer Price, and Phyllis Tien), U01-HL146242; Los Angeles CRS (Roger Detels and Matthew Mimiaga), U01-HL146333; Metropolitan Washington CRS (Seble Kassaye and Daniel Merenstein), U01-HL146205; Miami CRS (Maria Alcaide, Margaret Fischl, and Deborah Jones), U01-HL146203; Pittsburgh CRS (Jeremy Martinson and Charles Rinaldo), U01-HL146208; UAB-MS CRS (Mirjam-Colette Kempf, James B. Brock, Emily Levitan, and Deborah Konkle-Parker), U01-HL146192; UNC CRS (M. Bradley Drummond and Michelle Floris-Moore), U01-HL146194.

Disclaimer. The contents of this publication are solely the responsibility of the authors and do not represent the official views of the National Institutes of Health.

Data sharing. Access to individual-level data from the MWCCS may be obtained upon review and approval of a MWCCS concept sheet. Links and instructions for online concept sheet submission are on the study website.

Financial support. The MWCCS is funded primarily by the National Heart, Lung, and Blood Institute, with additional cofunding from the Eunice Kennedy Shriver National Institute of Child Health & Human Development, National Institute on Aging, National Institute of Dental and Craniofacial Research, National Institute of Allergy And Infectious Diseases, National Institute of Neurological Disorders And Stroke, National Institute of Mental Health, National Institute on Drug Abuse, National Institute of Nursing Research, National Cancer Institute, National Institute on Alcohol Abuse and Alcoholism, National Institute on Deafness and Other Communication Disorders, National Institute of Diabetes and Digestive and Kidney Diseases, National Institute on Minority Health and Health Disparities, and in coordination and alignment with the research priorities of the Office of AIDS Research. MWCCS data collection is also supported by the National Center for Advancing Translational Sciences awards: UL1-TR000004 (University of California, San Francisco), UL1-TR003098 (Johns Hopkins University), UL1-TR001881 (University of California, Los Angeles), UL1-TR001409 (Georgetown University), KL2-TR001432 (Georgetown University), TL1-TR001431 (Georgetown University); and by the National Institute of Allergy and Infectious Diseases awards: P30-AI-050409 (Center for AIDS Research at Emory University), P30-AI-073961 (Miami Center for AIDS Research), P30-AI-050410 (University of North Carolina at Chapel Hill Center for AIDS Research), P30-AI-027767 (University of Alabama at Birmingham Center for AIDS Research), and P30-MH-116867 (Miami Center for HIV and Research in Mental Health). This work was also supported in part by R01AI157758.