Elevated rates of intracerebral hemorrhage

in individuals from a US clinical care

HIV cohort

Felicia C. Chow, ABSTRACT

MD, MAS Objective: To compare rates of intracerebral hemorrhage (ICH) in HIV-infected and uninfected in-
Wei He, MS dividuals in a large clinical care cohort and to assess risk factors associated with ICH.
Peter Bacchetti, PhD
Methods: We identified incident ICH in HIV-infected and uninfected control cohorts from the Part-
Susan Regan, PhD
ners Health Care system using ICD-9-CM codes. We constructed Cox proportional hazards mod-
Steven K. Feske, MD
els to estimate adjusted hazard ratios for HIV infection and other predictors of ICH.
James B. Meigs, MD,
MPH Results: The incidence rate of ICH was 2.29 per 1,000 person-years in HIV-infected individuals
Steven K. Grinspoon, compared with 1.23 per 1,000 person-years in uninfected individuals, with an unadjusted inci-
MD dence rate ratio of 1.85 (95% confidence interval 1.37–2.47, p , 0.001). In a multivariable
Virginia A. Triant, MD, model, HIV infection was independently associated with a higher hazard of ICH, although its
MPH effect diminished with increasing age. Female sex was associated with a lower hazard of ICH
in the uninfected cohort but not in the HIV cohort. CD4 count ,200 3 106 cells/L and antico-
agulant use were predictive of ICH.
Correspondence to Conclusions: HIV infection conferred an increased adjusted hazard of ICH, which was more pro-
Dr. Chow:
chowf@sfgh.ucsf.edu nounced in young patients and in women. Neurology® 2014;83:1705–1711

GLOSSARY
ART 5 antiretroviral therapy; CKD 5 chronic kidney disease; HR 5 hazard ratio; ICD-9-CM 5 International Classification of
Diseases, Ninth Revision, Clinical Modification; ICH 5 intracerebral hemorrhage; IRR 5 incidence rate ratio; PY 5 person-
years; RPDR 5 Research Patient Data Registry; SAH 5 subarachnoid hemorrhage.

Cardiovascular and cerebrovascular event rates are increased in HIV infection.1–7 Several studies
pointing to an elevated risk of cerebrovascular disease in HIV-infected individuals have focused
on ischemic stroke and excluded intracranial hemorrhage.5,8,9 Whereas ischemic stroke and
cardiovascular disease often coexist and share a similar risk profile and underlying mechanism,
the association between HIV and intracerebral hemorrhage (ICH), which has mechanisms
distinct from ischemic stroke, is less clear. We examined the association of HIV with the risk
of ICH by comparing the incidence of ICH in HIV-infected and uninfected individuals in a
large US clinical care cohort. Our primary hypothesis was that the rate of ICH in HIV-infected
individuals would be higher than in uninfected individuals. We further controlled for potential
confounders and identified predictors of ICH in HIV-infected individuals.

METHODS Study design and patient population. We conducted an observational study of an HIV cohort derived from the
Partners HealthCare System Research Patient Data Registry (RPDR), a clinical care database of all inpatient and outpatient encounters
from Massachusetts General Hospital and Brigham and Women’s Hospital. Other vascular outcomes have been studied in this
cohort.3,5 HIV-infected individuals were identified from the RPDR using ICD-9-CM codes 042 or V08, previously validated in the
cohort.5 The HIV cohort was matched by age, race, and sex in a 1:10 ratio to HIV-uninfected individuals, who constituted the control
cohort. The control cohort was intentionally overpopulated to allow for the possibility of missing data and application of the exclusion
criteria. Within the cohorts, individuals who were at least 18 years of age at the beginning of the observation period and who had at least
1 inpatient or 2 outpatient clinical encounters were eligible for the study. The observation period began on the latest of January 1, 1996,
Editorial, page 1690
date of first ICD-9-CM code for HIV, or the first encounter for controls. Individuals were censored at the earliest of date of ICH, date of
last encounter, or December 31, 2009.
Supplemental data
at Neurology.org
From the Departments of Neurology (F.C.C.) and Epidemiology and Biostatistics (P.B.), University of California San Francisco; the Department of
Neurology (S.K.F.), Brigham and Women’s Hospital, Boston; and the Division of General Medicine (W.H., S.R., J.B.M., V.A.T.), Program in
Nutritional Metabolism (S.K.G., V.A.T.), and Division of Infectious Diseases (V.A.T.), Massachusetts General Hospital, Boston.
Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.

© 2014 American Academy of Neurology 1705

 Standard protocol approvals, registrations, and patient confounders. Statistical analyses were performed using Stata
consents. The Partners Human Research Committee and Uni- (StataCorp 2012; Stata Statistical Software: release 12; Stata
versity of California, San Francisco Committee on Human Corporation, College Station, TX).
Research approved the study.
RESULTS Demographic and clinical characteristics. A
Outcomes. We identified the primary endpoint of first-ever
ICH using ICD-9-CM code 431. This ICD-9-CM code has
total of 39,519 individuals were included in the study,
been previously demonstrated in a large administrative database of whom 4,251 were HIV-infected and 35,268 HIV-
to be 82% sensitive and 93% specific with positive predictive uninfected. These 2 cohorts contributed 237,760 PY
value of 80%.10 Individuals with concomitant ICD-9-CM codes of observation, with a median duration of follow up of
for skull fracture or intracranial injury (800.xx–801.xx, 803.xx– 5.47 years (interquartile range 1.48, 10.25). The
804.xx, 851.xx–854.xx) from the same encounter were excluded.
mean age was 41.6 years in the HIV cohort and
Patients were excluded from the study if an ICD-9-CM code for
ICH occurred prior to the first ICD-9-CM code for HIV or
40.6 years in the control cohort. Women
before the start of the observation period. constituted 31% of the HIV cohort compared with
Comorbid conditions that might impact the risk of ICH were 34% of controls. A total of 52% of the HIV and
also identified by ICD-9-CM codes documented at any time dur- control cohorts were white. The prevalence of most
ing the observation period, including hypertension (401.xx), dia- comorbidities, including hypertension, diabetes
betes (250.xx), chronic kidney disease (CKD) (585.1–6, 9; 403. mellitus, and smoking, was higher in the HIV
xx, 404.xx), and endocarditis (421.xx, 424.9x). Ascertainment of
cohort. Anticoagulation and statin use were also
hypertension and diabetes using ICD-9-CM codes has been pre-
viously validated in this cohort.11 Smoking status was ascertained more common in the HIV cohort (table 1).
by natural language processing, as described previously.12 Patients
Rates of ICH. We identified 320 cases of ICH, 58 in
with no smoking-related documentation in the medical record
were considered nonsmokers. We also obtained medication data the HIV and 262 in the control cohort. The inci-
on statins, antiplatelet agents, and anticoagulants. HIV parame- dence rate of ICH was 2.29 per 1,000 PY in HIV-
ters, including a history of any antiretroviral therapy (ART) use infected compared with 1.23 per 1,000 PY in
(ever vs never use by the end of the observation period), most uninfected individuals, with an unadjusted
recent CD4 count before the end of the observation period, CD4 incidence rate ratio (IRR) of 1.85 (table 2). After
count ,200 3 106 cells/L, and detectable HIV viral load
stratification by sex, a relative increase in ICH rates
(HIV RNA $ 400 copies/mL) at any time during the observation
period were determined.
was also seen overall in HIV-infected men and
women, although the IRRs comparing ICH rates in
Statistical analysis. We calculated age- and sex-stratified the HIV to control cohort were greater for women
incidence rates of ICH per 1,000 person-years (PY) of follow- than for men overall and in most age groups (table 2).
up time in the HIV and control cohorts, using incidence rate
In a multivariable model, HIV was independently
ratios to compare rates between cohorts. We constructed a
multivariable Cox proportional hazards regression analysis to
associated with ICH, with a 1.87 greater hazard of
model time to incident ICH, measured in years of age with the ICH at age 40 in men, after adjusting for age, race,
age at the start of the study being a “late entry” time, for the hypertension, diabetes, CKD, endocarditis, CNS
entire cohort of HIV-infected and uninfected individuals, infections or malignancy, smoking, antiplatelet
adjusting for demographics and potential ICH-related factors. therapy, anticoagulation, and statin use (table 3).
For the HIV cohort with available CD4 and viral load data
The effect of HIV on the hazard of ICH appeared
(n 5 2,278), we constructed a second multivariable Cox
to violate the proportional hazards assumption,
regression model of time to incident ICH, again measured in
years of age. Given the smaller number of events (n 5 26), we decreasing by an estimated factor of 0.96 for every
included only clinically relevant variables in a forward stepwise year that age increased. For example, at age 30, an
selection process. In each model, we tested the proportional HIV-infected man had 2.71 times the hazard of ICH
hazards assumption by plotting scaled Schoenfeld residuals for compared with an uninfected man, whereas at age 50,
each predictor against time, in addition to performing a test of the hazard of ICH for an HIV-infected man
nonzero slopes. We modeled variables with strong evidence for
compared to an uninfected man was 1.29 and by
nonproportional effects on the hazard of ICH by allowing their
effects to change with age. Because Cox regression does not age 60, the increased hazard was no longer present
explicitly estimate the effect of time (in our case age) on the risk (table 4).
of the outcome, we performed an auxiliary analysis fitting a
parametric Weibull model to obtain a simple quantitative Predictors of ICH. In addition to HIV status, hyper-
estimate of the effect of age among HIV-infected individuals. We tension predicted ICH (table 3). Diabetes mellitus
used the STATA streg command and extracted the hazard ratio was associated with a lower hazard of ICH, as were
(HR) for age 60 vs age 40 to illustrate the estimated magnitude of statin and antiplatelet use. We observed an interac-
the influence of age on hazard of ICH. For sensitivity analyses tion between female sex and HIV status on the hazard
where we evaluated the effect of the addition of a variable to a
of ICH, with a protective effect of female sex on ICH
model, a predetermined change-in-coefficient threshold of 10%
was considered significant. All p values were 2-sided with ,0.05 in HIV-uninfected individuals, but an estimate of no
considered statistically significant for testing the primary hypothesis protective effect among those with HIV (table 3). In a
that HIV is associated with ICH after controlling for potential sensitivity analysis, we excluded individuals with a

1706 Neurology 83 November 4, 2014

 Table 1 Clinical characteristics of HIV-infected and control cohorts

HIV-infected (n 5 4,251) Control (n 5 35,268) p Value

Hypertension, n (%) 1,472 (35) 9,644 (27) ,0.0001

Diabetes mellitus, n (%) 886 (21) 3,924 (11) ,0.0001

Chronic kidney disease, n (%) 254 (6) 888 (3) ,0.0001

Endocarditis, n (%) 159 (4) 277 (1) ,0.0001

a
CNS infections/malignancy, n (%) 157 (4) 35 (0.1) ,0.001

Smoking, n (%) 2,053 (48) 10,102 (29) ,0.0001

Antiplatelet use, n (%) 924 (22) 6,012 (17) ,0.0001

Anticoagulant use, n (%) 262 (6) 1,483 (4) ,0.0001

Statin use, n (%) 687 (16) 4,805 (14) ,0.0001

CD4 count (3106 cells/L), mean (SD) 474 (315) — —

Nadir CD4 count (310 cells/L), mean (SD)

6
286 (254) — —

HIV RNA (log, copies/mL), median (IQR) 0 (0, 6.7) — —

ART use, n (%) 2,113 (50) — —

NRTI use, n (% of ART users) 2,016 (95) — —

NNRTI use, n (% of ART users) 1,174 (56) — —

PI use, n (% of ART users) 1,403 (66) — —

Duration of ART use, y, mean (SD) 4 (4) — —

Abbreviations: ART 5 antiretroviral therapy; ICD-9-CM 5 International Classification of Diseases, Ninth Revision, Clinical
Modification; IQR 5 interquartile range; NNRTI 5 non-nucleoside reverse transcriptase inhibitors; NRTI 5 nucleoside
reverse transcriptase inhibitors; PI 5 protease inhibitors.
a
CNS infection/malignancy: ICD-9-CM codes 013 (tuberculosis), 321.0 (cryptococcus), 112.83 (candida), 054.3, 054.72
(herpes simplex virus), 130.0 (toxoplasma), 053.0, 053.10 (herpes zoster virus), 046.3 (progressive multifocal leukoen-
cephalopathy), 200.5 (CNS lymphoma), and 094.81, 094.9, 094.2, 091.81 (syphilis).

history of endocarditis or CNS infection or had little effect on the hazard of ICH and did not alter
malignancy from the multivariate model. The effect the HR for any other predictors by more than 10%,
of HIV on the hazard of ICH did not change other than CD4 count ,200 3 106 cells/L, for which
significantly (HR 1.89, 95% CI 1.22–2.91, p 5 the HR changed from 4.61 (95% CI 2.09–10.17,
0.004 for a man at age 40), nor did the HR for p , 0.001) to 3.88 (95% CI 1.68–8.96, p 5 0.002).
other covariates in the model.
DISCUSSION In a large US clinical care cohort, we
Predictors of ICH among HIV-infected individuals. In found an increased rate of ICH among HIV-
the HIV cohort, 2,278 HIV-infected individuals infected compared with uninfected individuals. The
had available data on CD4 count and viral load. hazard of ICH associated with HIV was increased
Using a forward stepwise selection process, we even after adjustment for recognized ICH risk
included sex, CD4 count ,200 3 106 cells/L, factors, suggesting that a component of the risk
anticoagulant use, ART use, and statin use in the may be mediated by factors unique to HIV. Indeed,
multivariable model. CD4 count ,200 3 106 cells/L among the HIV cohort, a CD4 count ,200 3 106
at any point during the observation period was cells/L was associated with a markedly higher risk of
associated with a higher rate of ICH, as was ICH, raising the possibility that more advanced HIV
anticoagulant use (table 5). We found strong disease stage may increase ICH risk.
evidence from fitting a parametric Weibull model to Several risk factors for ICH are more prevalent in
the HIV cohort, with the same covariates as those HIV-infected individuals than in the general population,
chosen by forward stepwise selection, that age including hypertension,13–16 a strong independent pre-
influenced hazard of ICH in the HIV cohort (p 5 dictor of ICH.17–21 Moreover, at younger ages, African
0.038 vs constant hazard), as in the general Americans, who have been disproportionately affected
population. For example, the ratio of hazards in by the HIV epidemic,22 have a greater risk of ICH com-
HIV-infected individuals at age 60 compared with pared with whites.17,20 The fact that ICH is associated
age 40 was 1.74 (95% CI 1.02–5.80). with worse functional outcomes and higher case fatality
In a sensitivity analysis, we included having a rates than ischemic stroke23,24 could lead to consider-
detectable viral load in the HIV-only model. This able loss of quality-adjusted life-years, particularly for

Neurology 83 November 4, 2014 1707

 was not included. Conversely, in the United States,
Table 2 Incidence rates for intracerebral hemorrhage stratified by age group
and sex
data from the Nationwide Inpatient Sample indicate
that the proportion of all ischemic strokes attributable
HIV-infected Control to HIV-infected individuals rose in the United States
from 1997 to 2006, whereas the proportion of ICH
Rate per Rate per Unadjusted IRR
Events 1,000 PY Events 1,000 PY (95% CI) hospitalizations in HIV-infected individuals was stable
Age group, y to decreased.28 Our results support a study that observed
18–35 14 1.55 38 0.41 3.80 (1.90–7.17)
a higher rate of a composite outcome of ICH and sub-
arachnoid hemorrhage (SAH) in a Canadian HIV
36–45 22 2.23 74 1.04 2.15 (1.27–3.50)
cohort compared with population-based controls.29
46–55 15 3.26 91 2.78 1.18 (0.63–2.04)
Our study advances the existing literature by focusing
56–65 3 2.29 37 3.52 0.65 (0.13–2.06) specifically on ICH as opposed to a composite outcome
66–75 3 7.55 12 3.41 2.21 (0.40–8.20) of ICH and SAH, an important distinction given differ-
>75 1 5.04 10 7.67 0.66 (0.02–4.62) ences in underlying pathophysiology and etiology, and
Total 58 2.29 262 1.23 1.85 (1.37–2.47)
by including HIV-related clinical factors.
Mounting evidence points to a link among inade-
Women, y
quately controlled HIV, chronic inflammation,
18–35 4 1.13 8 0.20 5.76 (1.27–21.50)
immune dysregulation, and accelerated atherosclero-
36–45 11 4.06 16 0.69 5.85 (2.45–13.42)
sis in HIV infection.30–37 Existing data conflict
46–55 2 1.65 17 1.75 0.94 (0.11–3.96) regarding the specific association between CD4 count
56–65 0 0 6 1.91 0 and cerebrovascular and cardiovascular risk.5,6,33,38,39
66–75 1 7.75 5 3.48 2.23 (0.05–19.92) Similar to our finding that a CD4 count ,200 3 106
>75 0 0 5 6.97 0
cells/L was associated with higher ICH risk, we pre-
viously found a higher CD4 count decreased the haz-
Total 18 2.24 57 0.72 3.09 (1.71–5.34)
ard of ischemic stroke, although this did not reach
Men, y
statistical significance.5 Furthermore, ART use, for
18–35 10 1.82 30 0.57 3.18 (1.39–6.68) which there is mixed evidence in relation to stroke
36–45 11 1.54 58 1.21 1.28 (0.60–2.46) risk,5,6,9,38 was associated with a lower, but not statis-
46–55 13 3.84 74 3.21 1.20 (0.61–2.18) tically significant, hazard of ICH (table 5). These
56–65 3 3.12 31 4.20 0.74 (0.15–2.38)
findings suggest that more advanced or less well-
controlled HIV infection may contribute to increased
66–75 2 7.46 7 3.37 2.21 (0.22–11.62)
cerebrovascular risk. While inflammation may play a
>75 1 10.43 5 8.53 1.22 (0.03–10.92)
role in the development of ischemic stroke in HIV-
Total 40 2.31 205 1.53 1.50 (1.04–2.12) infected individuals, an analogous association with
Abbreviations: CI 5 confidence interval; IRR 5 incidence rate ratio; PY 5 person-years. ICH has not been explored. Inflammatory markers
that are elevated in HIV and strongly predictive of
mortality,40 however, have been shown to predict
younger HIV-infected individuals. Few studies have ICH in the general population.e1–e3 Lipohyalinosis,
specifically evaluated ICH risk in HIV-infected com- a destructive process affecting intracerebral small ves-
pared with uninfected individuals. Data from studies sels marked by “a loss of normal arterial architecture,
performed before combination ART are confounded mural foam cells and, in acute cases, evidence of
by advanced disease and associated opportunistic in- fibrinoid vessel wall necrosis,” may underlie both
fections that may increase ICH risk.25 Furthermore, small vessel lacunar infarcts and a subset of primary
ICH was less common than ischemic stroke in many ICH.e4–e6 Endothelial dysfunction in the setting of
of these studies,26,27 reflecting the typical breakdown of chronic inflammation and immune dysfunction is
stroke by subtype and rendering it difficult to draw postulated to play a role in the pathogenesis of cere-
conclusions based on small numbers of events. Studies bral small vessel disease and lipohyalinosise7,e8 and
evaluating ICH from the current era of combination may be a contributing factor in the association
ART conflict. A Danish observational cohort between HIV and increased ICH risk.
of HIV-infected individuals compared with a We found that HIV conferred a relatively greater
population-based control group evaluated cerebral risk of ICH in women compared with men. While
infarction separately from ICH.6 ICH incidence was absolute ICH rates were higher in men, the relative
increased in HIV-infected individuals with a much risk of ICH comparing HIV-infected to uninfected
more pronounced effect among HIV-infected IV drug individuals was greater for women at any age. Exam-
users, although information on other potential con- ined another way, the protective effect of female sex
founders including smoking and anticoagulant use on the hazard of ICH, which is in line with the

1708 Neurology 83 November 4, 2014

 rebound and treatment interruption, which could
Table 3 Hazard ratios for predictors of incident intracerebral hemorrhage
among HIV-infected individuals and uninfected controls
potentially contribute to the decreased protective
effect from being a woman on ICH risk among
HIV-infected 1 uninfected controls HIV-infected individuals.e10–e15 Furthermore, sex-
(n 5 39,519)
specific differences in the immune response to HIV
HR (95% CI) p Value have been elucidated, with higher levels of immune
Main effects activation in HIV-infected women compared with
HIV infection (in men at age 40 years) 1.87 (1.23–2.84) 0.003
men at a given level of viral suppressione16 and com-
a
pared with uninfected women.e17 Additional investi-
Female sex among HIV-uninfected 0.49 (0.36–0.66) ,0.001
gation into differential effects of sex on ICH hazard
Female sex among HIV-infected 1.01 (0.57–1.77) 0.98
among HIV-infected individuals is warranted.
White 1.01 (0.80–1.27) 0.94 As with myocardial infarction and ischemic
Hypertension 1.29 (1.01–1.66) 0.045 stroke,1,3,5,38 the absolute rates of ICH among HIV-
Diabetes mellitus 0.66 (0.47–0.94) 0.020 infected individuals increased with older age. In the
Chronic kidney disease 1.52 (0.97–2.38) 0.067
overall cohort, however, we demonstrated that the
effect of HIV status on the hazard of ICH decreased
Endocarditis 1.76 (0.96–3.21) 0.066
as age increased. A relatively greater effect of HIV in
CNS infections/malignancy 2.73 (1.31–5.71) 0.008
younger age groups has been demonstrated for other
Smoking (ever vs never) 0.99 (0.78–1.25) 0.91
vascular outcomes, including myocardial infarction
Anticoagulation use 1.21 (0.81–1.81) 0.36 and ischemic stroke.1,5 The decrease in the effect of
Antiplatelet use 0.68 (0.50–0.93) 0.016 HIV on the hazard of ICH with older age may reflect
Statin use 0.53 (0.38–0.75) ,0.001 a relatively greater contribution of traditional risk
Time-varying effects
factors to ICH risk as individuals age, consonant with
the rising incidence and prevalence of various comor-
HIVb 0.96 (0.94–0.99) 0.005
bidities in aging HIV-infected individuals.e18,e19
Abbreviations: CI 5 confidence interval; HR 5 hazard ratio. As expected, hypertension was associated with a
a
HR for interaction between HIV and being female was 2.07 (95% CI 1.10–3.89, p 5 higher hazard of ICH in the overall model, and there
0.024).
b
For each year increase in age, the effect of HIV decreased by 0.96 (95% CI 0.94–0.99,
was a trend towards an association with endocarditis
p 5 0.005). and CKD. These comorbidities represent potentially
modifiable risk factors that should be targeted in
known epidemiology of ICH in the general popula- HIV-infected individuals, as in the general population.
tion up to 75 years of age,21,e9 was absent or less strong The decrease in the hazard of ICH associated with dia-
for HIV-infected women compared with uninfected betes and statin use may reflect better engagement in
women (multivariable model p value for interaction the health care system of these individuals (and thus,
of HIV and sex, p 5 0.024). Prior studies have dem- for example, better control of associated comorbidities
onstrated a greater risk of myocardial infarction and including hypertension). A similar explanation may
ischemic stroke in HIV-infected compared to unin- underlie the protective effect of antiplatelet use.
fected women.2,3,5 Women have been shown to have In addition to unmeasured confounding that can
lower rates of adherence and higher rates of viral occur in observational studies, limitations relate to the

Table 4 Hazard ratios for incident intracerebral hemorrhage assessing age-specific differences by HIV status

Men (HIV-infected compared with Women (HIV-infected compared with

uninfected control) uninfected control)

Age, y HR (95% CI) p Value HR (95% CI) p Value

20 3.92 (1.73–8.89) 0.001 8.11 (3.36–19.58) ,0.001

30 2.71 (1.49–4.92) 0.001 5.60 (2.79–11.24) ,0.001

40 1.87 (1.23–2.84) 0.003 3.86 (2.18–6.85) ,0.001

45 1.55 (1.08–2.24) 0.019 3.21 (1.86–5.55) ,0.001

50 1.29 (0.90–1.84) 0.16 2.67 (1.54–4.62) ,0.001

60 0.89 (0.56–1.41) 0.62 1.84 (0.97–3.49) 0.062

70 0.61 (0.32–1.19) 0.15 1.27 (0.57–2.85) 0.56

Abbreviations: CI 5 confidence interval; HR 5 hazard ratio.

Neurology 83 November 4, 2014 1709

 infected individuals should remain a priority in the
Table 5 Hazard ratios for predictors of incident intracerebral hemorrhage in the
HIV cohort
setting of an aging HIV-infected population.

HIV-infected (n 5 2,278) AUTHOR CONTRIBUTIONS

F.C. Chow participated in the study concept and design, performed the
HR (95% CI) p Value data analysis and interpretation, and drafted the manuscript. W. He per-
Female sex 1.79 (0.78–4.09) 0.17 formed the data analysis and provided critical revisions on the paper.
P. Bacchetti performed the data analysis and interpretation and provided
CD4 count <200 3 106 cells/L 4.61 (2.09–10.17) ,0.001 critical revisions on the paper. S. Regan performed the data analysis and
Anticoagulation use 4.44 (1.57–12.53) 0.005 interpretation and provided critical revisions on the paper. S.K. Feske
participated in the study concept and design and provided critical revi-
ART use 0.46 (0.18–1.15) 0.098 sions on the paper. J.B. Meigs participated in the study concept and
Statin use 0.14 (0.02–1.13) 0.066 design and provided critical revisions on the paper. S.K. Grinspoon par-
ticipated in the study concept and design and provided critical revisions
Abbreviations: ART 5 antiretroviral therapy; CI 5 confidence interval; HR 5 hazard ratio. on the paper. V.A. Triant participated in the study concept and design,
interpreted the data, supervised the study, and provided critical revisions
on the paper.
use of ICD-9-CM codes. Rates of ICH were higher in
our cohort than previously published estimates, a find- STUDY FUNDING
ing that may reflect misclassification by ICD-9-CM Supported in part by NIH grants 5T32MH090847 (F.C.C.), K24
codes or referral center bias. However, misclassification DK080140 (J.B.M.), P30 DK040561 (S.K.G.), and K01 AI073109
(V.A.T.).
should be equally likely to occur in HIV-infected and
uninfected cohorts (i.e., nondifferential) and thus
would bias the result toward the null. Due to limitations DISCLOSURE
The authors report no disclosures relevant to the manuscript. Go to
associated with ICD-9-CM codes, we could not assess Neurology.org for full disclosures.
ICH size, location, or severity, or their relationship
with hypertension stage or duration. Furthermore, Received February 18, 2014. Accepted in final form June 23, 2014.
we could not distinguish between ICH and hemor-
rhagic conversion of ischemic strokes. We lacked data REFERENCES
1. Currier JS, Taylor A, Boyd F, et al. Coronary heart disease
on duration of HIV infection and illicit drug use; how-
in HIV-infected individuals. J Acquir Immune Defic
ever, we included several clinically relevant covariates,
Syndr 2003;33:506–512.
including smoking status, which is unavailable for 2. Lang S, Mary-Krause M, Cotte L, et al. Increased risk
many observational studies. Our study is also poten- of myocardial infarction in HIV-infected patients in
tially limited by defining various fixed predictors using France, relative to the general population. AIDS
information from the entire observation period. For 2010;24:1228–1230.
example, an individual who does not incur ICH at a 3. Triant VA, Lee H, Hadigan C, Grinspoon SK. Increased
acute myocardial infarction rates and cardiovascular risk
given time has the continued opportunity to be started
factors among patients with human immunodeficiency virus
on ART or develop hypertension, while those who disease. J Clin Endocrinol Metab 2007;92:2506–2512.
incur ICH at that time have no such opportunity, 4. Obel N, Thomsen HF, Kronborg G, et al. Ischemic heart
leading to bias toward a protective effect. Finally, the disease in HIV-infected and HIV-uninfected individuals: a
Partners HIV cohort may not entirely reflect the population-based cohort study. Clin Infect Dis 2007;44:
greater population of HIV-infected individuals in 1625–1631.
5. Chow FC, Regan S, Feske S, Meigs JB, Grinspoon SK,
the United States, or in resource-limited settings
Triant VA. Comparison of ischemic stroke incidence in
worldwide, although data on other vascular outcomes HIV-infected and non-HIV-infected patients in a US
from this cohort have been consistently corroborated health care system. J Acquir Immune Defic Syndr 2012;
by those from other large HIV cohorts. Although the 60:351–358.
Partners HIV cohort has a greater percentage of whites 6. Rasmussen LD, Engsig FN, Christensen H, et al. Risk of
than that reported for the United States,e20 nonwhite cerebrovascular events in persons with and without HIV.
AIDS 2011;25:1637–1646.
race/ethnicity and women are, overall, well-represented
7. Silverberg MJ, Leyden WA, Xu L, et al. Immunodefi-
(48% and 31%, respectively). ciency and risk of myocardial infarction among HIV-
In this large US clinical care cohort, HIV con- positive individuals with access to care. J Acquir Immune
ferred an independent risk of ICH, adjusting for Defic Syndr 2014;65:160–166.
established ICH-related risk factors. This increased 8. Evers S, Nabavi D, Rahmann A, Heese C, Reichelt D,
risk was relatively more pronounced for young pa- Husstedt I-W. Ischaemic cerebrovascular events in
HIV infection: a cohort study. Cerebrovasc Dis 2003;
tients and women. Further investigation into the
15:199–205.
mechanism for this relatively greater increase in risk
9. Corral I, Quereda C, Moreno A, et al. Cerebrovascular
of ICH among these traditionally low-risk demo- ischemic events in HIV-1-infected patients receiving
graphic subsets is merited. Focusing preventive efforts highly active antiretroviral therapy: incidence and risk fac-
on modifiable ICH risk factors among all HIV- tors. Cerebrovasc Dis 2009;27:559–563.

1710 Neurology 83 November 4, 2014

10. Tirschwell DL, Longstreth WT. Validating administrative 24. Dennis MS. Outcome after brain haemorrhage. Cerebro-
data in stroke research. Stroke 2002;33:2465–2470. vasc Dis 2003;16(suppl 1):9–13.
11. Triant VA, Meigs JB, Grinspoon SK. Association of 25. Cole JW, Pinto AN, Hebel JR, et al. Acquired immuno-
C-reactive protein and HIV infection with acute myocar- deficiency syndrome and the risk of stroke. Stroke 2004;
dial infarction. J Acquir Immune Defic Syndr 2009;51: 35:51–56.
268–273. 26. Ortiz G, Koch S, Romano JG, Forteza AM,
12. Zeng QT, Goryachev S, Weiss S, Sordo M, Murphy SN, Rabinstein AA. Mechanisms of ischemic stroke in HIV-
Lazarus R. Extracting principal diagnosis, co-morbidity infected patients. Neurology 2007;68:1257–1261.
and smoking status for asthma research: evaluation of a 27. Mochan A, Modi M, Modi G. Stroke in black South
natural language processing system. BMC Med Inform African HIV-positive patients: a prospective analysis.
Decis Mak 2006;6:30. Stroke 2003;34:10–15.
13. Bergersen BM, Sandvik L, Dunlop O, Birkeland K, 28. Ovbiagele B, Nath A. Increasing incidence of ischemic
Bruun JN. Prevalence of hypertension in HIV-positive stroke in patients with HIV infection. Neurology 2011;
patients on highly active retroviral therapy (HAART) com- 76:444–4505.
pared with HAART-naïve and HIV-negative controls: re- 29. Durand M, Sheehy O, Baril J-G, LeLorier J,
sults from a Norwegian study of 721 patients. Eur J Clin Tremblay CL. Risk of spontaneous intracranial hemor-
Microbiol Infect Dis 2003;22:731–736. rhage in HIV-infected individuals: a population-based
14. Seaberg EC, Muñoz A, Lu M, et al. Association between cohort study. J Stroke Cerebrovasc Dis 2013;22:e34–e41.
highly active antiretroviral therapy and hypertension in a 30. Hsue PY, Deeks SG, Hunt PW. Immunologic basis of
large cohort of men followed from 1984 to 2003. AIDS cardiovascular disease in HIV-infected adults. J Infect
2005;19:953–960. Dis 2012;205:S375–S382.
15. Palacios R, Santos J, García A, et al. Impact of highly 31. Triant VA. HIV infection and coronary heart disease: an
active antiretroviral therapy on blood pressure in HIV‐ intersection of epidemics. J Infect Dis 2012;205:S355–S361.
infected patients: a prospective study in a cohort of naive 32. Strategies for Management of Antiretroviral Therapy
patients. HIV Med 2006;7:10–15. SMART Study Group, El-Sadr WM, Lundgren JD,
16. De Socio GV, Ricci E, Maggi P, et al. Prevalence, aware- et al. CD41 count-guided interruption of antiretroviral
ness, treatment, and control rate of hypertension in HIV- treatment. N Engl J Med 2006;355:2283–2296.
infected patients: the HIV-HY study. Am J Hypertens 33. Triant VA, Regan S, Lee H, Sax PE, Meigs JB,
2014;2:222–228. Grinspoon SK. Association of immunologic and virologic
17. Sturgeon JD, Folsom AR, Longstreth WT, Shahar E, factors with myocardial infarction rates in a US healthcare
Rosamond WD, Cushman M. Risk factors for intracere- system. J Acquir Immune Defic Syndr 2010;55:615–619.
bral hemorrhage in a pooled prospective study. Stroke 34. Hsue PY, Lo JC, Franklin A, et al. Progression of atheroscle-
2007;38:2718–2725. rosis as assessed by carotid intima-media thickness in patients
18. Feldmann E, Broderick JP, Kernan WN, et al. Major risk with HIV infection. Circulation 2004;109:1603–1608.
factors for intracerebral hemorrhage in the young are mod- 35. Kaplan RC, Kingsley LA, Gange SJ, et al. Low CD41 T-
ifiable. Stroke 2005;36:1881–1885. cell count as a major atherosclerosis risk factor in HIV-
19. Woo D, Sauerbeck LR, Kissela BM, et al. Genetic and infected women and men. AIDS 2008;22:1615–1624.
environmental risk factors for intracerebral hemorrhage: 36. Lichtenstein KA, Armon C, Buchacz K, et al. Low CD4
preliminary results of a population-based study. Stroke 1T Cell count is a risk factor for cardiovascular disease
2002;33:1190–1195. events in the HIV outpatient study. Clin Infect Dis 2010;
20. Howard G, Cushman M, Howard VJ, et al. Risk factors 51:435–447.
for intracerebral hemorrhage: the Reasons for Geographic 37. Subramanian S, Tawakol A, Burdo TH, et al. Arterial
and Racial Differences in Stroke (REGARDS) study. inflammation in patients with HIV. JAMA 2012;308:
Stroke 2013;44:1282–1287. 379–386.
21. Ariesen MJ, Claus SP, Rinkel GJE, Algra A. Risk factors 38. Vinikoor MJ, Napravnik S, Floris-Moore M, Wilson S,
for intracerebral hemorrhage in the general population: a Huang DY, Eron JJ. Incidence and clinical Features of
systematic review. Stroke 2003;34:2060–2065. cerebrovascular disease among HIV-infected adults in the
22. Centers for Disease Control and Prevention. Disparities in Southeastern United States. AIDS Res Hum Retroviruses
diagnoses of HIV infection between blacks/African Americans 2013;29:1068–1074.
and other racial/ethnic populations: 37 states, 2005–2008. 39. Sabin CA, Ryom L, De Wit S, et al. Associations between
MMWR Morb Mortal Wkly Rep 2011;60:93–98. immune depression and cardiovascular events in HIV
23. Bhalla A, Wang Y, Rudd A, Wolfe CDA. Differences infection. AIDS 2013;27:2735–2748.
in outcome and predictors between ischemic and intrace- 40. Tien PC, Choi AI, Zolopa AR, et al. Inflammation and
rebral hemorrhage: the South London stroke register. mortality in HIV-infected adults: analysis of the FRAM study
Stroke 2013;44:2174–2181. cohort. J Acquir Immune Defic Syndr 2010;55:316–322.

Neurology 83 November 4, 2014 1711