Low Nadir CD4 Cell Count Predicts Sustained Hypertension in HIV‐Infected Individuals

Abstract

Hypertension is associated with cardiovascular disease in the human immunodeficiency virus (HIV)–infected population. The authors aimed to test the hypothesis whether advanced immunosuppression with low nadir CD4 lymphocyte cell count is a predictor of sustained hypertension in HIV‐infected individuals. In a longitudinal study of an HIV cohort of 434 patients (43±11 years, 72% men, 71% Caucasians), standardized blood pressure was measured in duplicate during 3 clinical visits both at baseline and after 3.4±0.8 years. The lowest CD4 cell count in the individual history was recorded as nadir CD4. Both nadir CD4 cell count <50 cells/μL and duration of antiretroviral therapy (ART) were associated with sustained hypertension, and the highest proportion of hypertensive patients was observed in those who had both nadir CD4 cell count <50 cells/μL and prolonged ART duration. Nadir CD4 cellcount <50 cells/μL was an independent predictor of hypertension (adjusted odds ratio [OR], 2.48; 95% confidence interval [CI], 1.27–4.83), as was ART duration (adjusted OR, 1.13; 95% CI, 1.03–1.24). The predictive power of ART duration was more pronounced in patients with nadir CD4 cell count <50 cells/μL. Delaying ART initiation until a state of advanced immunosuppression might add to and even fuel the cardiovascular risk associated with ART. J Clin Hypertens (Greenwich). 2012; 00:00–00 ©2012 Wiley Periodicals, Inc.

Awareness of hypertension in the human immunodeficiency virus (HIV)–infected population has increased during the past years as hypertension may contribute to the increasing cardiovascular morbidity and mortality in this population. ¹ , ² , ³ , ⁴ To what extent HIV‐related factors contribute to hypertension is not clarified. Cardiovascular risk factors associated with hypertension are prevalent in HIV cohorts. ² , ⁵ Antiretroviral therapy (ART) causes metabolic aberrations that may increase cardiovascular risk, ⁶ , ⁷ and ART has been linked to hypertension ⁸ , ⁹ , ¹⁰ as well as to cardiovascular disease. ¹¹ Moreover, HIV‐related factors could be related to hypertension, and a link between duration of HIV infection and hypertension has been reported. ¹² , ¹³

Translocation of bacterial products from the gut lumen into the systemic circulation has been confirmed in HIV infection. Microbial translocation is suggested as a driver of HIV‐associated chronic immune activation, ¹⁴ which is characterized by increased activation and turnover of immune cells and release of pro‐inflammatory cytokines. Markers of microbial translocation and immune activation are only partly restored with suppressive ART and independently predict disease progression, immune restoration, and mortality in patients with and without ART. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ We have recently reported a potential link between microbial translocation and hypertension. ¹⁹

The level of CD4+ T lymphocytes declines as the HIV infection progress. Viral load (ie, HIV RNA) and CD4 cell count are the main parameters to monitor HIV progression and decide when to start ART. Current European guidelines recommend initiation of ART when the CD4 cell count falls below 350 cells/μL. ²⁰ The therapeutic goal of ART is viral suppression and subsequent normalization of CD4 cell count, ie, complete immune restoration. The nadir CD4 cell count is a critical determinant of CD4 recovery on ART. ²¹ , ²² , ²³ , ²⁴ Incomplete restoration of CD4 cell count despite virologic suppression with ART is associated with both increased non‐AIDS (ie, noninfectious) and AIDS‐related morbidity and mortality. ²⁴ , ²⁵ , ²⁶ , ²⁷ Moreover, incomplete immune recovery is linked to persistent immune activation, ²⁸ , ²⁹ microbial translocation, ¹⁸ and endothelial dysfunction. ³⁰ Association between low nadir CD4 cell count and increased blood pressure (BP) after initiation of ART has been observed ³¹ , ³² ; however, others have failed to find a clear association with hypertension. ³³ , ³⁴ , ³⁵

Given this background, the aim of the present study was to test the hypothesis that low nadir CD4 cell count, reflecting earlier advanced immunosuppression, is related to hypertension in HIV‐infected individuals.

Methods

Study Participants

In this prospective, longitudinal study on BP, 542 patients attending the outpatient clinic at the Department of Infectious Diseases at Oslo University Hospital, Ullevål, finished the baseline study and 434 HIV‐infected patients (80%) completed the follow‐up study, as detailed previously. ³⁶ Of the 108 patients lost to follow‐up, more were normotensive compared with those who completed the study (80.6% vs 65.2%, P=.002). Inclusion required written consent, and the study was approved by the South‐Eastern Regional Committee for Medical and Health Research Ethics, and concession was obtained from the National Data Inspectorate.

Clinical Data

A total of 434 patients were followed for a period of 3.4±0.8 (range 1.7–5.7) years. BP was examined during three clinical visits, both at baseline and at follow‐up. As BP measurements were performed during clinical visits at the outpatient clinic, the interval between BP measurements was most often 3 months. BP was measured in duplicate 2 minutes apart at each visit using a semiautomatic oscillometric device (Omron M4, Matsusaka Co. Ltd, Matsusaka, Japan) with appropriate cuff size according to the upper arm circumference. BP measurements were performed by well‐trained nurses with the patient in a relaxed sitting position in a quiet room after minutes of rest. The average of 6 systolic BP (SBP) and diastolic BP (DBP) measurements were used for statistical analysis both at baseline and follow‐up. Hypertension was defined as SBP ≥140 mm Hg and/or DBP ≥90 mm Hg or use of antihypertensive therapy. Sustained hypertension was characterized as persistent hypertension both at baseline and at follow‐up. Because considerable dynamics in the hypertension status has been found in this population, ³⁶ sustained hypertension was chosen as the main focus in this study.

As previously detailed, demographic, clinical, and laboratory data were obtained from hospital records as well as the hospital’s HIV database, in which clinical and laboratory data are updated continuously. Height and body weight were measured at the first clinical visit at baseline and body mass index (BMI) was calculated as kg/m². Obesity was defined as BMI ≥30 kg/m². Blood samples were taken as per clinical practice before each clinical visit. Data were recorded from the date nearest to the date of inclusion both at baseline and at follow‐up. ⁸ CD4 cell count was determined by routine flow cytometry using TriTest CD4/CD8 with TruCount Tubes (Becton Dickinson Biosciences, San Jose, CA). Nadir CD4 cell count was obtained from the HIV database and recorded as the lowest CD4 cell count ever until the first date of inclusion at baseline, regardless of ART status (nadir CD4 reached as ART‐naïve [79.5%], during ART interruption [9.2%], or on ART [11.3%]). “Low” nadir CD4 cell count refers to cell count <50 cells/μL. HIV RNA was recorded close to the date of nadir when available (ie, from 1996), as well as at inclusion at baseline. Hyperlipidemia was defined as total cholesterol ≥6.2 mmol/L. Glomerular filtration rate (GFR) was estimated according to the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) formula eGFR. ³⁷ Urinary albumin excretion rate was measured in triplicate as albumin‐to‐creatinine ratio (ACR) in spot urine both at baseline and follow‐up. Abnormal urinary albumin excretion was defined as the presence of microalbuminuria (ie, ACR ≥2.5–30 mg/mmol in at least 2 of 3 samples) or ACR 30 to 150 mg/mmol, ie, low‐grade proteinuria. Duration of HIV infection was estimated as the period from HIV diagnosis to inclusion at baseline, while duration of ART was calculated as the cumulative exposure to a triple class regimen at inclusion.

Statistical Analyses

Baseline demographic, clinical, and laboratory characteristics were used in statistical analyses. Patients were categorized as Caucasians or non‐Caucasians. BMI and baseline CD4 cell count was dichotomized with cut‐off at 25 kg/m² and 200 cells/μL, respectively. Nadir CD4 cell count was divided into 2 (<50 and ≥200 cells/μL) or 3 groups (<50, 50–200, and ≥200 cells/μL). In the logistic regression analyses, nadir CD4 cell count was dichotomized with 50 cells/μL as cut‐off as this was the threshold for increased prevalence of sustained hypertension (Table I). Duration of ART was analyzed both as a continuous variable and a categorical variable divided into 4 groups, namely ART‐naïve, duration <2, 2 to 5, and >5 years. Data are presented as mean±standard deviation, number or percentage, or if skewed, as median and 25th and 75th percentiles (interquartile range [IQR]). For between‐groups comparisons either t test or the Mann‐Whitney was used. Categorical data were compared using chi‐square test and linear trend tested using chi‐square linear‐by‐linear association.

Table I.

Baseline Demographic, Clinical, and Laboratory Data in a Prospective, Longitudinal Study of 434 HIV‐Infected Individuals

Characteristics	Sustained Hypertensives (n=123)	Others (n=311)	P Value
Age, y	9.5±10.2	40.5±9.6	<.001
Male sex, %	88.6	65.6	<.001
Caucasian ethnicity, %	81.3	67.5	.004
Smoking, %	35.8	44.7	.090
Diabetes mellitus, %	4.1	1.9	.306
BMI, kg/m2	25.4±3.6	23.5±3.6	<.001
BMI >25 kg/m2	49.6	29.6	<.001
Cholesterol, mmol/L	5.5±1.1	5.0±1.1	<.001
eGFR, mL/min	104.7 (94.2–110.8)	111.7 (100.1–121.8)	<.001
SBP, mm Hg	151.5±15.4	123.1±11.6	<.001
DBP, mm Hg	93.6±8.7	76.8±7.4	<.001
Abnormal albumin excretion, %	17.1	6.4	.001
HIV RNA, copies/mL	<50 (0–20,000)	300 (0–37,000)	.140
CD4 ≥200 cells/μL, %	92.7	83.0	.009
Nadir CD4, cells/μL, %
<50	27.6	14.5	.006
50–200	35.8	41.8
≥200	36.6	43.7
HIV RNA at nadir, copies/mLa	155,000 (33,750–430,000)	100,000 (24,000–315,000)	.326
HIV duration, y	8.9 (4.7–14.2)	5.2 (1.7–10.7)	<.001
ART‐naïve, %	15.4	27.7	.007
Duration of ART, y	5.1 (1.4–7.0)	1.6 (0–5.2)	<.001

Abbreviations: ART, antiretroviral therapy; BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate (according to the chronic kidney disease epidemiology collaboration formula); HIV, human immunodeficiency virus; SBP, systolic blood pressure. Data are given as percentages, mean±standard deviation, or medians (interquartile range) for skewed distributions. ^aAvailable data for 321 pre‐ART patients.

In a multiple, logistic regression model with sustained hypertension as the dependent variable, independent variables were selected based on significant associations (P<.1) with sustained hypertension in bivariate analyses. As HIV duration and duration of ART were highly correlated (Spearman’s rho=0.65), HIV duration was entered separately into the model.

Results

Baseline Characteristics of the HIV Cohort

The majority of the study population (n=434) were men (72.1%), most were of Caucasian ethnicity (71.4%), and the mean age was 43.1±10.5 years. The men were older than the women (45.4±10.4 vs 36.9±8.2 years, P<.001). One third (35.3%) of the patients had BMI >25 kg/m², and 5.7% were obese (ie, BMI ≥30 kg/m²). Hyperlipidemia was present in 16.4% of the patients, and 42.2% of the patients were daily smokers. Abnormal urinary albumin excretion was found in 9.4%. Hypertension was present in 34.8% (n=151) of the cohort at baseline, of whom 123 patients (81%) had sustained hypertension. Sustained hypertension was more prevalent in men than women (34.8% vs 11.6%, P<.001). Nadir CD4 cell count <50 cells/μL and ≥200 cells/μL was observed in 18.2% (n=79) and 41.7% (n=181), respectively. Median CD4 cell count was 375 (interquartile range [IQR], 250–570) cells/μL, the majority (85.7%) of the study population had baseline CD4 cell count ≥200 cells/μL, and 44.5% had an HIV RNA level <50 copies/mL. One fourth (24.2%) of the patients were ART‐naïve at baseline, whereas 56.6% were taking current ART, of whom 64.5% (n=169) used protease inhibitors. Patients taking current ART were older (43.9±10.5 vs 41.7±10.4 years, P=.040) and more often had hyperlipidemia compared with all the others (21.8% vs 8.1%, P<.001). No differences were observed regarding sex, BMI, eGFR, proportion of smokers, or diabetes mellitus in patients with or without current ART.

Characteristics of Patients With Nadir CD4 Cell Count <50 cells/μL

All but 1 of the 79 patients with nadir CD4 cell count <50 cells/μL were ART exposed at baseline. Patients with nadir CD4 cell count <50 cells/μL had similar age (44.7±8.8 vs 42.7±10.9 years, P=.138), similar proportion of patients with diabetes mellitus (3.1% vs 0%, P=.228), higher BP (135.5±18.5/84.3±11.8 mm Hg vs 130.2±17.9/81.0±10.6 mm Hg, P=.018/.016), more often hypertension (43.6% vs 25.1%, P=.001), less often BMI >25 kg/m² (21.5% vs 38.3%, P=.005), higher cholesterol (5.5±1.1 mmol/L vs 5.1±1.2 mmol/L, P=.009), lower GFR (105±19 mL/min vs 110±19 mL/min, P=.025), and longer duration of ART (6.0 [IQR 2.5–7.7] years vs 1.6 [0–5.2] years, P<.001), while a history of AIDS was more prevalent (62.0% vs 9.6%, P<.001) compared with those with a nadir CD4 cell count ≥50 cells/μL. There were no differences in the proportion of men or abnormal albumin excretion between the two groups (data not shown).

A minority (14.5%) of the patients with a baseline CD4 cell count ≥200 cells/μL had a nadir CD4 cell count <50 cells/μL, whereas two thirds (68.4%) of the patients with nadir CD4 cell count <50 cells/μL had gained CD4 cell count ≥200 cells/μL at baseline. There was a stepwise increase in patients with nadir CD4 cell count <50 cells/μL within groups of increased duration of ART (ART‐naïve, <2, 2–5, and >5 years); namely, 1%, 14%, 17%, and 34%, respectively (P<.001).

Predictors of Sustained Hypertension

The baseline characteristics of the HIV‐infected individuals with and without sustained hypertension are given in Table I. The two groups differed in age, cholesterol, eGFR, BP, and HIV duration as well as in the proportion of men, Caucasians, smokers, BMI >25 kg/m², CD4 ≥200 cells/μL, abnormal albuminuria, and nadir CD4 cell count <50 cells/μL. Furthermore, the hypertensive patients had longer duration of HIV as well as of ART, and fewer of the hypertensive patients were ART‐naïve (Table I). As many as 85.3% (29 of 34) of the hypertensive patients with a nadir CD4 cell count <50 cells/μL had reached CD4 cell counts ≥200 cells/μL at baseline due to ART‐mediated immune reconstitution. No difference was observed in the proportion of patients with obesity (8.9% vs 4.5%, P=.074) or in viral load between patients with and without sustained hypertension (Table I).

Within the three strata of nadir CD4 cell count, the proportion of patients with hypertension increased significantly across groups of increased ART duration (Figure). The highest proportion of hypertensive patients was observed in patients having both nadir CD4 cell count <50 cells/μL and duration of ART >5 years (Figure). In univariate logistic analyses with sustained hypertension as the outcome, the association between ART duration and hypertension increased by decreased levels of nadir CD4 cell count, with the highest odds ratio (OR) per year of longer ART duration in the stratum of nadir CD4 cell count <50 cells/μL (OR, 1.48; 95% CI, 1.19–1.83, P<.001).

Figure.

 Proportional increase in hypertensive human immunodeficiency virus human immunodeficiency virus–infected patients across groups of increased duration of antiretroviral therapy (ART) in different strata of nadir CD4 cell count. P values refer to chi‐square linear‐by‐linear association.

In a multivariate logistic regression analyses with sustained hypertension as the outcome, nadir CD4 cell count <50 cells/μL was an independent predictor after adjustment (Table II). Furthermore, older age, male sex, BMI >25 kg/m², baseline CD4 cell count ≥200 cells/μL, and longer duration of ART were significant predictors of sustained hypertension (Table II) after adjustment for ethnicity, smoking, eGFR, cholesterol, and presence of abnormal urinary albumin excretion.

Table II.

Predictors of Sustained Hypertension in a Longitudinal Study of 434 HIV‐Infected Individuals: Univariate and Multivariate Logistic Regression Analyses

Characteristics at Baseline	Unadjusted OR (95% CI)	P Value	Adjusted OR (95% CI)	P Value
Age, per 1 year	1.09 (1.07–1.12)	<.001	1.08 (1.05–1.12)	<.001
Male sex	4.08 (2.23–7.47)	<.001	2.55 (1.23–5.29)	.012
Caucasian ethnicity	2.09 (1.25–3.49)	.005	0.76 (0.37–1.56)	.455
Smoking	0.69 (0.45–1.06)	.091	0.78 (0.47–1.31)	.352
Body mass index >25 kg/m2	2.34 (1.53–3.60)	<.001	2.71 (1.62–4.55)	<.001
Cholesterol, per mmol/L	1.50 (1.24–1.82)	<.001	1.10 (0.88–1.38)	.407
eGFR, per mL/min	0.97 (0.96–0.98)	<.001	1.01 (0.99–1.03)	.441
Abnormal albumin excretion, yes vs no	3.00 (1.56–5.75)	.001	1.80 (0.82–3.91)	.141
CD4 ≥200 cells/μL	2.60 (1.24–5.46)	.011	2.58 (1.09–6.12)	.031
Nadir CD4 <50 cells/μL	2.26 (1.36–3.75)	.002	2.31 (1.17–4.56)	.015
Duration of ART, per year	1.25 (1.16–1.35)	<.001	1.13 (1.03–1.24)	.011

Abbreviations: ART, antiretroviral therapy; CI, confidence interval; eGFR, estimated glomerular filtration rate (according to the chronic kidney disease epidemiology collaboration formula); HIV, human immunodeficiency virus; OR, odds ratio.

Substituting duration of ART with HIV duration in the regression model, nadir CD4 cell count remained an independent predictor of sustained hypertension with odds ratio practically unchanged, while HIV duration did not reach statistical significance.

Discussion

The main and novel finding in this study was that a nadir CD4 cell count <50 cells/μL, reflecting advanced immunodeficiency in the past, was an independent predictor of sustained hypertension in HIV‐infected individuals. One previous study has reported a nadir CD4 cell count <200 cells/μL to be associated with prevalent hypertension in univariate analysis, but this association did not persist after multivariate adjustment. ¹² In the cross‐sectional studies that failed to reveal a link between prevalent hypertension and nadir CD4 cell count, ³³ , ³⁴ , ³⁵ nadir was analyzed either as a continuous variable ³⁴ or with the cut‐off level at 100 or 200 cells/μL, ³³ , ³⁵ which makes comparison with our study difficult. Furthermore, the outcome in these studies was prevalent hypertension ³³ , ³⁴ , ³⁵ as opposed to that of sustained hypertension in the present study.

Close to three quarters of the patients in our study population were men, and they were older than the women and they had more frequently sustained hypertension. This is in keeping with data from the general population in Norway. ³⁸ In addition to low nadir CD4 cell count and ART duration, traditional risk factors for hypertension such as age, male sex, and BMI were independent predictors of hypertension in this study. The number of obese patients in the present study was low.

Another finding in our study was that ART duration independently predicted sustained hypertension, in accordance with prior studies linking ART duration to prevalent hypertension ⁸ , ⁹ as well as to increased risk of cardiovascular disease. ¹¹ This effect of ART was evident even after adjustment for metabolic risk factors, which may be exaggerated by use of ART. ⁶ , ⁷ Notably, low nadir CD4 cell count has been independently predictive of increased BP 1 year after initiation of ART, using a cut‐off of 50 and 100 cells/μL, respectively. ³¹ , ³² In the present study, the predictive power of ART duration was more pronounced in patients with low nadir CD4 cell count, suggesting that the harmful effects of ART might be enhanced by factors related to severe immunodeficiency. Since the majority of the patients with low nadir CD4 cell count and sustained hypertension had a baseline CD4 cell count ≥200 cells/μL, the combination of severe immunodeficiency in the past, and subsequent ART‐mediated immune reconstitution might be a contributing factor in these patients. Thus, although ART is likely to contribute to development of hypertension and cardiovascular risk in the HIV‐infected population, delaying ART initiation until advanced immunosuppression may increase the risk of future hypertension and possibly cardiovascular disease.

Low nadir CD4 cell count has been associated with chronic immune activation and persistent microbial translocation. ¹⁸ , ²⁹ , ³⁹ In an exploratory substudy of 42 patients recruited from the present study, we have recently demonstrated a strong association between microbial translocation, measured by increased levels of lipopolysaccharide (LPS) before initiation of ART, and later sustained hypertension. ¹⁹ An association between LPS and atherosclerosis has been reported both in the general population ⁴⁰ and in HIV‐infected patients. ⁴¹ Moreover, low‐grade inflammation has been linked to elevated BP and hypertension, both as a potentially causal factor and as a consequence of hypertensive stimuli such as angiotensin II and salt. ⁴²

Furthermore, LPS is a known trigger of the innate immune response and might cause endothelial activation through stimulation of Toll‐like receptor 4 on endothelial cells. ⁴³ HIV‐infected individuals have endothelial dysfunction compared with HIV‐uninfected individuals, ⁴⁴ , ⁴⁵ and low nadir CD4 cell counts have been independently associated with endothelial dysfunction, ³⁰ progression of subclinical atherosclerosis, ⁴⁶ and arterial stiffness ⁴⁷ in HIV‐infected individuals. Direct toxic effects of viremia on endothelial cells ⁴⁴ , ⁴⁵ could also be a potential contributor to the association between low nadir CD4 count, endothelial dysfunction, and subsequent hypertension, although we observed no link between viral load at the time of nadir and hypertension in this study.

Limitations and Strengths

Unknown hypertension status at the time of nadir CD4 cell count is a limitation, and the classification of hypertension status according to the average of several BP measurements with variable time intervals is not optimal. A moderate sample size increases the risk of statistical type II errors. On the other hand, the risk of type I errors is less probable, and the significant association between a low nadir CD4 cell count and hypertension is likely to be reliable. Our study also has several strengths, including use of sustained hypertension over a period of several years, strict standardized repeated BP measurements performed at a single‐study center, and the recruitment of patients from an unselected HIV cohort.

Conclusions

The major finding in our study is that low nadir CD4 cell count was an independent and strong predictor of sustained hypertension, which even seemed to enhance the predictive power of ART duration. Our data indicate that delaying ART initiation until a state of advanced immunosuppression might add to and even fuel the cardiovascular risk associated with ART.

Acknowledgments and statement of financial disclosure:  We acknowledge the staff at the Outpatient Clinic of Infectious Diseases at Oslo University Hospital, Ullevål for their skilled assistance. Professor Leiv Sandvik, Unit of Epidemiology and Biostatistics at Oslo University Hospital, Ullevål has given invaluable advice on the statistics presented in this paper. Sources of funding: The HIV fund of the Department of Infectious Diseases at Oslo University Hospital, Ullevål, and the South‐Eastern Norway Regional Health Authority. The authors report no specific funding in relation to this research and no conflicts of interest to disclose.