Presented During:

Monday, Mar 8, 2021: 11:30 AM  - 11:40 AM 
Virtual CROI 2021  

Abstract Number:


Abstract Type:

General Abstract  


Michael J. Silverberg1, Tory Levine-Hall1, Alexandra Anderson1, Stacey E. Alexeeff1, Jennifer O. Lam1, C. Bradley Hare2, Jason Flamm3, Andrew Williams4, Matthias Cavassini5, Kendall Bryant6, Michael A. Horberg7, Derek D. Satre8


1Kaiser Permanente Northern California, Oakland, CA, USA, 2Kaiser Permanente San Francisco Medical Center, San Francisco, CA, USA, 3Kaiser Permanente Sacramento Medical Center, Sacramento, CA, USA, 4Tufts University, Boston, MA, USA, 5Lausanne University Hospital, Lausanne, Switzerland, 6National Institute on Alcohol Abuse and Alcoholism, Bethesda, MD, USA, 7Kaiser Permanente Mid-Atlantic States, Rockville, MD, USA, 8University of California San Francisco, San Francisco, CA, USA

Presenting Author:

Dr Michael Silverberg, PhD MPH  


With higher risk of cardiovascular disease (CVD) in persons with HIV (PWH), management of hypertension, dyslipidemia, and diabetes mellitus is crucial. Here, we evaluate the extent to which PWH have successfully managed these conditions and their influence on CVD risk.


Cohort study of adult (≥18 years) PWH and 20:1 age-, sex-, race/ethnicity-matched persons without HIV (PWoH) who were members of an integrated healthcare system in Northern California during 2013-2017. We excluded subjects with prevalent CVD (coronary heart disease or ischemic stroke). In those with hypertension, dyslipidemia, and diabetes, we computed the disease management index (DMI), which accounts for both the amount and duration of person-time above treatment goal (vertical lines in Figure) over 6-month intervals. A DMI of 100% represents perfect control and DMI <100% is the amount in control relative to a reference treated population. Next, using Cox regression, we computed hazard ratios (HR) for incident CVD by HIV status overall, and in subgroups of those with successfully controlled risk factors (i.e., DMI 100%). Models were adjusted for other key modifiable risk factors (smoking and alcohol use), demographics, and clinical factors (Charlson comorbidity index, depression, body mass index, healthcare utilization).


The study included 8,285 PWH and 170,517 PWoH, with similar prevalences of hypertension (19% PWH; 22% PWoH), dyslipidemia (41% for both) and diabetes (8% PWH; 9% PWoH). PWH and PWoH had similar control of most conditions (with DMIs approaching 100%) except for triglycerides (worse control for PWH) and HbA1c (better control for PWH) (Figure). Among PWH, other factors, including smoking and unhealthy alcohol use, had only marginal associations with reduced DMIs. Overall, PWH had 450 CVD events (20.8 per 1,000 person-years) and PWoH had 7,648 events (17.0 per 1,000 person-years), with an adjusted HR of 1.18 (95% CI 1.07-1.30). The elevated risk of CVD for PWH was attenuated and not statistically significant when comparing PWH and PWoH with successfully controlled dyslipidemia (HR 1.10; 95% CI 0.91-1.34), and diabetes (HR 1.02; 0.72-1.42), but remained significant for those with successfully controlled hypertension (HR 1.35; 1.10-1.67).


Successful management of dyslipidemia and diabetes may help mitigate the CVD disparity in PWH. Further research is needed to evaluate whether more stringent hypertension treatment goals than <140/90 mm Hg are needed to prevent CVD in PWH.


(L) Cardiovascular Complications of HIV Infection and Antiretroviral Therapy


Cardiovascular disease

Supporting Image: Figure.jpg

Does this abstract include any aspects of research on SARS-CoV-2 or COVID-19?


Prior Presentation or Publication: In general, CROI does not accept work that has been previously published or publicly presented. Abstract text that is under copyright by a publication or another conference should not be submitted verbatim to CROI. Consideration may be given to a previously presented submission if meaningful newer data or different analyses are included or if the prior presentation was to a conference not focused on HIV- or SARS-CoV-2-related topics. Have your study data or abstract information been published, submitted for publication where publication is anticipated on or before the start of the CROI where you will present, or presented at any other major national or international scientific or medical HIV-related conferences (ie, generally 400 or more attendees)?