Alexandra Yannoutsos1, Davide Agnoletti1, Julie Peroz-Froz1, Camille LY1, Helene Lelong1, Jirar Topouchian1, Jacques Gilquin2, Segolene Boucly2, Hafeda Rostane2, Michel E Safar1, Jean-Paul Viard2,3* and Jacques Blacher1
Received date: July 25, 2014; Accepted date: November 19, 2014; Published date: November 25, 2014
Citation: Yannoutsos A, Agnoletti D, Peroz-Froz J, Camille LY, Lelong H, et al. (2014) Structural and Functional Arterial Parameters, Immunovirological Control and Vitamin D in HIV-Infected Patients. J AIDS Clin Res 5: 375 doi:10.4172/2155-6113.1000375
Copyright: © 2014 Yannoutsos A, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of AIDS & Clinical Research
Objective: HIV-infected patients still experience higher Cardiovascular (CV) mortality rates, even if an adequate viral suppression is achieved. In addition, vitamin D insufficiency, a common condition in HIV-infected patients, is increasingly associated with CV risk. We therefore aim to investigate relationships between immunological parameters, antiretroviral therapy, plasma vitamin D and arterial parameters, including aortic stiffness and wave reflections, in HIV-infected patients who achieved viral suppression but possibly remain at increased CV risk. Methods: We conducted a cross-sectional study including 178 middle-aged HIV-infected patients. HIV infection was controlled in a large number of participants, representative of a real-world setting. In addition to carotid Intima Media Thickness (IMT), central hemodynamic parameters involved aortic Pulse Wave Velocity (PWV), carotid Augmentation index (AIx) and Pulse Pressure Amplification (PPA) measured noninvasively using applanation tonometry. Results: Aortic PWV was slightly but unsignificantly higher than the theoretical values obtained in general population according to age and blood pressure, and was independent of HIV-related parameters. In univariate and multivariate analyses, carotid AIx was positively correlated with current CD4 T-cell count and PPA was positively correlated with vitamin D, independently of other confounders. No HIV-related parameters or vitamin D entered the multivariate analysis of carotid IMT / plaque. Conclusion: In our chronically treated population, HIV infection was not associated with increased aortic stiffness but with a positive correlation between current CD4 T-cell count and degree of AIx, suggesting that patients with higher CD4 T-cell count may have higher wave reflections. The positive correlation between vitamin D and PPA suggests that vitamin D deficiency may be independently associated with altered central hemodynamics in well controlled HIV-infected patients. These findings should be confirmed in prospective studies.
HIV infection; Aortic stiffness; Pulse wave velocity; Arterial wave reflections; Pulse pressure amplification; Cardiovascular disease risk factors; Plasma vitamin D
Combined antiretroviral therapy has led to a substantial decline in AIDS-related morbidity and mortality [1,2]. Mortality and overall incidence of opportunistic infections declined markedly in 1996 and early 1997 with the introduction of combination therapy including protease inhibitor (PI) . In the mid-2000s, despite a near-normal life expectancy, HIV-infected patients presented excess mortality rates compared to the general population; the estimated median remaining lifetime of a 25-year-old HIV-infected individual was 39 years compared to 51 years for a 25-year-old HIV-uninfected person . Patterns of co-morbidities and causes of death have changed. More specifically, cardiovascular (CV) diseases became one of the leading causes of death in this population . The French national survey, describing the causes of death among HIV-infected patients in France in 2010 and their evolution since 2000, has highlighted an increase in the proportion of CV deaths in middle-aged patients, between 40 and 59 years . HIV replication and incomplete immune restoration or persistent immune activation (as reflected by a high CD8 cell count or a CD4/CD8 ratio<1) were independently related to the risk of myocardial infarction . High burden of traditional risk factors and cumulative exposure to Highly Active Antiretroviral Therapy (HAART), especially to PI, may also account for increased CV risk [5,6].
Recently, evidence from large prospective cohort studies has highlighted that HIV-infected patients with viral suppression and who maintained or had recovered CD4 T-cell counts of at least 500 cells/mm3 in the HAART era, had no evidence for a raised risk of death compared with the general population . In contrast, patients with CD4 T-cell counts less than 500 cells/mm3 still experienced higher mortality rates, even if viral suppression was achieved . The commonest causes of death were CV disease and sudden death, highlighting the need for improved CV risk prediction and prevention in this population .
Routine CV risk prediction in HIV-infected patients is estimated with conventional risk prediction models but these tools do not include HIV-specific markers. Furthermore, because of their inability to integrate individual levels of intensity and exposure to all CV risk factors, such models of risk stratification may underestimate CV risk in young and middle-aged HIV-infected patients or those with the metabolic syndrome . Nevertheless, several previous studies have highlighted greater structural and functional arterial damage in HIV-infected patients compared to uninfected individuals [9,10]. Rather than increased carotid intima-media thickness (IMT) [11,12], carotid to femoral pulse wave velocity (PWV), a measure of aortic stiffness, has allowed to gather the largest amount of evidence regarding added predictive value for CV risk, over and above traditional risk factors, and persisting even after adjustment for carotid IMT and carotid stiffness [13-15]. Central augmentation index (AIx), which is considered a transit time-dependent marker of intensity of arterial wave reflections, and pulse pressure (PP) amplification, also bring independent predictive value for CV events and mortality [16,17]. However, such parameters may be difficult to interpret in HIV-infected patients because the association between HIV infection and greater arterial stiffness or wave reflections may be no longer significant after adjustment for classical risk factors, especially smoking and hypertension .
Little and inconsistent evidence exists about the relationship between virological and immunological parameters and arterial stiffness and wave reflections in HIV-infected patients [19-21]. Our aim was to investigate the relationships between immune status and arterial parameters, including aortic stiffness and wave reflections, in middle-aged HIV-infected patients who achieved viral suppression but possibly remain at increased CV risk.
In addition, vitamin D insufficiency is a common condition  and has been associated with CV morbidity  and increased all-cause and CV mortality in the general and HIV-infected populations [24,25]. In order to better assess CV damage and its determinants in HIV-infected patients, we therefore investigated cross-sectional correlations between structural and functional arterial parameters, HIV-related factors and vitamin D status.
Among HIV-infected persons routinely followed up for immunovirological, CV, metabolic, and bone health parameters at the Paris Hôtel-Dieu hospital, 178 consecutive HIV-1-infected patients in primary CV prevention, 143 men and 35 women, were included in this cross-sectional study, between September 2010 and July 2011. Exclusion criteria were age under 18, acute medical conditions and atrial fibrillation. No patient had any active AIDS-related condition at inclusion in the study. No patient had a past or present history of endocarditis and no patient had any active bacterial infection at the time of study. No active intravenous drug user was included in the study. Patients provided informed consent for additional hemodynamic measurements and data collection. The study was approved by our institutional review board.
Clinical and laboratory parameters
We recorded data on HIV clinical classification, comorbidities, CV risk factors, current medications including statin and antihypertensive therapy, past and present ART regimens, duration of ART and nadir CD4 T-cell count. Laboratory parameters, including plasma glucose and HbA1C, cholesterol (total, LDL and HDL) and triglycerides, renal function, 25-hydroxyvitamin D (25OHD), CD4 and CD8 T-cell counts, and viral load (HIV RNA copies/ml) were determined on the day of hemodynamic measurements.
The metabolic syndrome was defined by the presence of at least three out of the five following criteria according to the Third Report of the National Cholesterol Education Program: 1) SBP above 130 mmHg and/or DBP above 85 mmHg, or treated hypertension, 2) waist circumference over 102 cm in men and 88 cm in women, 3) triglyceride concentration above 1.69 mmol/l, 4) HDL-cholesterol concentration below 1.04 mmol/l in men and 1.29 mmol/l in women, and 5) fasting plasma glucose concentration above 6.1 mmol/l.
Central and peripheral arterial parameters
Hemodynamic measurements were performed in the morning after an overnight fast, in supine position. Peripheral Systolic BP (SBP) and diastolic BP (DBP) were measured at both arms using an automatic BP monitor (OMRON 705 CP II IT) after 5 minutes of rest. Five measurements 2 minutes apart were averaged. Heart rate was recorded.
After BP determination, structural and functional parameters of the arterial wall were performed noninvasively by applanation tonometry: this measure provides an accurate profile of intra-arterial BP curves, with a continuous beat-to-beat monitoring, by applying a piezoelectric sensor, the tonometer, over an artery. The reproducibility of these measurements, in our group and in others, has been previously published in detail [26,27].
Carotid-femoral PWV is considered as the gold standard direct and non invasive test for aortic stiffness assessment . The aortic PWV measurement was performed using an automatic device, the SphygmoCor, which allowed on-line pulse wave recording and automatic calculation of PWV. Aortic PWV was calculated as the distance between carotid and femoral arteries, divided by the time interval between the feet of the pressure waves at the recording sites. Pulse waveforms were obtained transcutaneously using applanation tonometry over the common carotid and femoral arteries. The carotid-femoral pathway was the direct distance measured using a standard compass system between the carotid and femoral measurement sites. Direct distance was multiplied by a scaling factor of 0.8 to obtained "real PWV" as previously described . Transit time was assessed as the time difference between two characteristic points (‘foot of the wave’) on carotid and femoral waveforms at the measurement sites.
Central BP components (SBP, DBP and PP), were estimated by the pulse wave analysis of the common carotid artery. Pulse pressure amplification was defined as the brachial PP-to-carotid PP (B/C) ratio. Amplification of systolic and pulse pressures from central to peripheral sites depends on the timing and amplitude of the reflected waves. Normal values of PPA vary from 1.7 at less than 20 years of age to 1.2 at more than 80 years of age .
Augmentation index is estimated from the pulse waveform measured directly at the carotid artery. This index is defined as the pressure augmentation (which is the difference between the second and first systolic peaks) expressed as a percentage of PP. AIx is frequently used to assess wave reflection, and is affected by both timing and amplitude of the reflected waves. Carotid and aortic AIx are known to be strongly and positively correlated . There is a non-linear relationship between normal vascular ageing and enhanced AIx, more marked in individuals under 50 : in out-of-hospital community-based population considered to represent the normal population, AIx values vary from 1.6% at less than 10 years of age to 24.1% at more than 60 years of age .
Carotid IMT was determined ultrasonographically, in supine position, using Sigma 110 KONTRON and a 7.5-13 MHz linear array probe. Carotid arteries, abdominal aorta and limb arteries were scanned ultrasonographically for the detection of plaques and stenosis. Echocardiography was performed for left ventricular systolic and diastolic function study.
Statistics were performed with SAS software version 9.0 (SAS Institute, Cary, NC). A p ≤ 0.05 was considered as statistically significant. We represented quantitative variables as mean ± standard deviation (±SD), or median (interquartile range, IQR) for non-normal variables. Qualitative variables are expressed as frequency and percentage.
Theoretical aortic PWV values in our study population were calculated by age category on the basis of Mean Arterial Pressure (MAP) according to The Reference Values for Arterial Stiffness' Collaboration study . Mattace-Raso et al. provided normal and reference values for aortic PWV based on a large European population of 11092 non diabetic subjects without overt CV disease and without current treatment for hypertension or for dyslipidemia. The reference value population included subjects with a mean age of 50 years, with high normal or high BP, with dyslipidemia and/or smoking . To test whether aortic PWV measured values in our study population are different from calculated theoretical values, we used the Mann-WhitneyUtest.
For regression analysis non-normal variables were log transformed. Multiple regression analyses were performed on PWV, AIx, PPA, and IMT; logistic regression was done for the presence of carotid plaques. Regressions models were obtained by stepwise selection, containing all variables that were significantly correlated in univariate analysis and based on their physiopathological plausibility. Variables that entered the stepwise selection were age, MAP, heart rate, gender, diabetes, smoking status, waist circumference, plasma 25OHD level, PI treatment, nadir CD4 T-cell count and current CD4 T-cell count. For hemodynamic parameters, age and MAP were forced into the models. For AIx and PPA, heart rate and gender were also forced into the multivariate models [33-34].
Clinical, biological and HIV-related parameters
Characteristics of the study population are summarized in Table 1. Patients were mostly men (80%), with a mean (± SD) age of 49 ± 9 years, and 49% were smokers (only 13% current smokers). More than a half (52%) presented with the metabolic syndrome. 5% of patients were diabetic, 22% were on antihypertensive therapy and 20% on a statin. Median (IQR) value for plasma 25 (OH) D level was 24 (15-34) ng/ml. 18% of patients had a history of AIDS-defining events and 76% were treated with PIs. 90% of patients had a viral load below 50 copies/ml and only 6% had a viral load of 400 copies/ml or more, which led to a very skewed distribution. Thus, we could not incorporate this variable in the statistical analysis. CD4 T-cell count was ≥500/mm3, between 200-500/mm3 and <200/mm3 in 55%, 39% and 6% of patients, respectively.
|Clinical and biological parameters|
|Male gender, n(%)||178||143(80)|
|Waist circumference, cm||178||91±11||64||131|
|Metabolic syndrome, n(%)||148||77(52)|
|Current smokers, n(%)||162||21(13)|
|Former smokers, n(%)||162||59(36)|
|Anthihypertensive therapy, n(%)||176||39(22)|
|Statin therapy, n(%)||175||35(20)|
|Total cholesterol, mmol/L||145||5.33±1.01||2.85||8.55|
|HDLb cholesterol, mmol/L||145||1.26±0.44||0.50||4.40|
|25-hydroxyvitamin D*, ng/mL||153||24 (15-34)||5||68|
|Duration of PIctreatment, months*||120||84 (44-171)||0||361|
|Number of patients treated with PI, n (%)||157||120(76)|
|Nadir CD4 count*||169||236 (90-325)||0||889|
|Current CD4 count*||174||560 (402-740)||49||2174|
|Current CD8 count*||172||692 (549-914)||75||2101|
|Nadir CD4 / current CD8*||167||0.339 (0.107-0.553)||0||.55|
|HIV classification: A or B, n(%)||141||116(82)|
|HIV classification: C, n(%)||141||25(18)|
Continuous variable are presented as mean ± standard deviation (SD); discrete variables as number (percent)
*For skewness distribution, median (interquartile range) is presented
aBMI: Body Mass Index
bHDL: High-density Lipoprotein
cPI: Protease Inhibitor.
Table 1: Clinical, biological and HIV-related parameters.
Central and peripheral arterial parameters
Arterial parameters of the study population are summarized in Table 2. Concerning peripheral hemodynamic parameters, median values (IQR) for brachial SBP, DBP, PP and MAP were 125.5(118-135) mmHg, 81(74-88) mmHg, 45.5(40-50) mmHg and 95(88-103) mmHg, respectively. Concerning central hemodynamic parameters, median values (IQR) for carotid SBP, DBP and PP were 113(105-124) mmHg, 81(75-88) mmHg and 33(28-38) mmHg, respectively. Median value (IQR) for carotid-femoral PWV in our study population was 8.3(7.5-9.5) m/sec, which was slightly higher than the theoretical carotid-femoral PWV calculated by age category on the basis of MAP [median (IQR): 8.0(7.4-9.1) m/sec]. However, the difference between measured and calculated theoretical aortic PWV did not reach significance (Mann-WhitneyUtest, p=0.09). The mean value (±SD) for carotid AIx and PPA were 1.0 ± 18.6% and 1.38 ± 0.20%, respectively. Concerning ultrasonographic measurement, the mean value (±SD) for carotid IMT was 0.738 ± 0.152 mm and 11% of patients had at least one carotid plaque.
|Brachial SBP*a, mmHg||174||125.5(118-135)||95||198|
|Brachial DBP*b, mmHg||174||81(74-88)||61||121|
|Brachial PP*c, mmHg||174||45.5(40-50)||25||91|
|Heart rate, bpm||177||69±12||42||107|
|Carotid SBP*, mmHg||161||113(105-124)||89||205|
|Carotid DBP*, mmHg||161||81(75-88)||61||120|
|Carotid PP*, mmHg||161||33(28-38)||21||97|
|cf-PWV*e †, m/s||173||8.3(7.5-9.5)||5.8||16.9|
|Theoretical PWV *‡, m/s||174||8.0(7.4-9.1)||5.6||13.5|
|Carotid AIxf, %||161||1.0±18.6||-50.0||42.6|
|Pulse pressure amplification, %||156||1.38±0.20||1.02||1.96|
|Carotid plaque, n(%)||158||18(11)|
Continuous variable are presented as mean ± standard deviation (SD); discrete variables as number (percent);
*For skewness distribution, median (interquartile range) is presented;
†Calculated as direct distance divided by transit time multiplied by 0.8;
‡Calculated by age category on the basis of mean blood pressure; test for difference between cf-PWV and Calculated Theoretical PWV: Mann-Whitney U test; p=0.09;
aSBP: Systolic Blood Pressure;
bDBP: Diastolic Blood Pressure;
cPP: Pulse Pressure;
dMAP: Mean Arterial Pressure;
ecf-PWV: Carotid-femoral Pulse Wave Velocity;
fAIx: Augmentation Index;
gIMT: Intima-media thickness.
Table 2: Central and peripheral arterial parameters.
Determinants of central arterial parameters
Multivariate regression models analysing factors independently correlated with central arterial parameters are summarized in table 3. Age (p<0.0001), MAP (p<0.0001) and diabetic status (p=0.021) were positively correlated with carotid-femoral PWV. Age (p<0.0001), gender (p<0.0001), MAP (p<0.0001), and CD4 T-cell count (p=0.012) were positively correlated with carotid AIx, whereas heart rate (p<0.0001) was negatively correlated. A positive and linear correlation between vitamin D and PPA (n=133, r=0.18, p=0.04) (Figure 1) persisted in the multiple regression analysis independently of the other cofactors as shown in Table 3. Gender (p<0.0001) and MAP (p=0.001) were negatively correlated with PPA, whereas vitamin D (p=0.042) and heart rate (p<0.0001) were positively correlated. Age was not independently associated with PPA in our study population.
|PWV*a||Coefficient||SE||partial R2||P value|
|Mean arterial pressure*||0.745||0.099||0.266||<0.0001|
|AIxb||Coefficient||SE||partial R2||P value|
|Mean arterial pressure*||63.941||9.289||0.244||<0.0001|
|Current CD4 count*||15.559||6.133||0.042||0.012|
|PPAc||Coefficient||SE||partial R2||P value|
|Mean arterial pressure*||-0.445||0.133||0.081||0.001|
aPWV: Pulse Wave Velocity;
bAix: Augmentation Index;
cPPA: Pulse Pressure Amplification;
SE: Standard Error.
Variables in the selection models were age, mean arterial pressure, heart rate, gender, diabetes, smoking status, waist circumference, 25-hydroxyvitamin D level, protease inhibitor treatment, nadir CD4 count, and current CD4 count.
Table 3: Determinants of aortic PWV, carotid AIx, and PPA.
|IMTa||Coefficient||SE||partial R2||P value|
|PLAQUE||OR||95% CL||P value|
|Active or past smoking||2.245||1.096||4.598||0.027|
aIMT: Intima-media Thickness; SBPb: Systolic Blood Pressure.
Variables in the selection models were age, gender, brachial SBP, brachial DBP, diabetes, smoking status, waist circumference, 25-hydroxyvitamin D level, protease inhibitor treatment, nadir CD4 count, current CD4 count
Table 4: Determinants of carotid IMT and carotid plaque, respectively by multivariate linear and logistic regression.
Our study population included middle-aged patients with controlled HIV infection in a large number of participants. Nearly all patients were virologically suppressed at the time of enrollment. Aortic PWV was slightly but unsignificantly higher than the reference values found in the general population according to age and BP. To the best of our knowledge, it is the first time that better immune status, as reflected by the CD4 T-cell count, was independently associated with a transit time-dependent marker of intensity of wave reflections. Plasma vitamin D and PPA were positively and independently correlated, suggesting that vitamin D deficiency may be associated with altered central hemodynamic parameters in well controlled HIV-infected patients.
Central arterial parameters and HIV infection
Aortic PWV: In our study population, mean ± SD age was 49 ± 9 years and the median (IQR) for carotid-femoral PWV was 8.3(7.5-9.5) m/s, which was slightly, but not significantly higher than the theoretical value which was 8.0(7.4-9.1) m/s according to the same age category, on the basis of MAP. Gold standard of normal and reference values for carotid-femoral PWV have been established previously based on an extensive data set obtained from 13 centers distributed across Europe, taking into account different methodological approaches for the determination of PWV . Advancing age and distending pressure appear to be the most important determinants of altered buffering function resulting from aortic wall stiffening . Aortic PWV reference values represent reliable estimates according to age and BP, allowing comparison with carotid-femoral PWV in our study population.
Consistent with our results, similar values of aortic PWV in HIV-infected patients in the HAART era compared to uninfected individuals were recently described [18,34]. Age, MAP and diabetic status were independent factors associated with increased aortic PWV in our study, with a strong correlation with age and MAP, findings highly consistent with the literature [33,35]. Contribution of other CV risk factors was nonsignificant and, interestingly, the presence of the metabolic syndrome, found in more than half of patients, did not enter the multiple regression analysis. In diabetic patients, longer past exposure to metabolic abnormalities may have a more marked impact on arterial stiffness.
A relationship between increased aortic stiffness and PI treatment in HIV-infected patients has been described in many observational studies, independently of conventional CV risk factors [9,20,36,37]. In our study population, neither HIV-related parameters or PI treatment entered the explicative model of aortic PWV, consistent with recently published results . It should be noted that there is some controversy on whether PIs or other antiretroviral classes are indeed associated with CV risk . Furthermore, interruption of ART has been associated with an increased risk of death and CV diseases comparatively to continuous treatment , suggesting that immunological benefits of drugs exceed the CV risk of metabolic side effects. In addition, we studied a group of middle-aged patients with extremely good virological control in a large number of participants and significant CV prevention implementation (only 13% were active smokers, 20% and 22% were on a statin and anti-hypertensive drugs, respectively). This may contribute to explain the discrepancy with previous studies showing higher PWV values in HIV-positive than in HIV-negative persons [9,36,37]. As age (in association with MAP) is the major determinant of aortic stiffness, the longitudinal follow-up of our study population would provide further evidence about early vascular aging and its determinants in HIV-infected patients with viral suppression.
Carotid AIx : CD4 T-cell count was the only independent HIV-related factor modulating carotid AIx, with a positive relation: increased carotid AIx was significantly associated with higher CD4 T-cell counts. The other independent factors modulating positively carotid AIx were age, MAP and female gender, consistent with the literature [34,40]. As expected, heart rate was negatively correlated with AIx .
The positive correlation found with current CD4 T-cell count seems in contrast with previous studies linking increased arterial stiffness with HIV infection. Little evidence and inconsistent results exist about the relationship between immunological parameters and arterial wave reflections among heterogeneous study populations. Ho et al. found, in a cross-sectional study of 134 middle-aged HIV-infected men, that increased aortic stiffness and wave reflections were independently related in multivariate models to a nadir CD4 T-cell count <350/mm3 at HAART initiation . In a cohort of 32 young HIV-infected patients treated with PIs and 32 HIV-uninfected subjects, Schillaci et al.  found that HIV infection was independently associated with increased aortic PWV and central AIx in multiple regression models. The authors suggested that immunodeficiency may account for vascular damage with earlier return of reflected waves at central level. However, the number of studied subjects was small. In contrast, Lazar et al. reported no association between HIV infection and increased arterial wave reflections in a homogeneous group of 276 HIV-infected and 67 HIV-uninfected Rwandan young women. In addition, current CD4 T-cell count among HIV-infected women did not correlate with central AIx or central PP .
In the present study, we found a strong positive association between carotid AIx and current CD4 T-cell count which persisted in multiple regression analysis independently of age, MAP, gender and heart rate. From a physiopathological viewpoint, AIx is a dependent measure of reflected wave transit time which may be related either to the compliance of the aorta or to increased heart rate. Increased heart rate is associated with a reduction in the time required for the backward pressure wave to return toward the heart without any change in aortic stiffness. Furthermore, central AIx may be an unreliable indicator of aortic stiffness in middle-aged or even older patients, relative wave reflections remaining unchanged or even reduced in elderly individuals whereas aortic stiffness is increased . This positive and independent correlation between CD4 T-cell count and a transit time-dependent marker of intensity of wave reflections does not imply a causal relationship between higher CD4 T-cell counts and altered wave reflections. In HIV-infected patients without overt increased aortic stiffness, this positive correlation, which persisted independently of heart rate, might suggest that patients with higher CD4 T-cell count have simply a more distensible aorta. Indeed, impedance mismatch at the junction between a normally compliant aorta and carotid artery represents a protective mechanism which facilitates pressure wave reflection and limits the transmission of excessive pulsatile energy into the microcirculation. Of course, this interpretation must be confirmed in longitudinal follow-up with a matched control group. Finally, differences in clinical characteristics between our study population and the previously published cohorts have to be considered: we studied a group of middle-aged HIV-infected men and women, chronically treated and virologically suppressed at the time of enrollment, and without overt increased aortic stiffness.
Carotid IMT and carotid plaque
Carotid plaque is a marker of the presence and extent of coronary artery disease and an independent predictor of CV events and mortality. Prevalence of ultrasonographic carotid plaque was 11% in our study population of middle-aged patients. A similar prevalence of carotid atherosclerosis has been observed in the general population  although a direct comparison with our results is difficult, because of the absence of an age-matched control group. Consistent with a recent meta-analysis of observational studies , we found that parameters of HIV infection or PI exposure were not independently associated with carotid plaque. Age and smoking have emerged as leading risk factors of carotid atherosclerosis in the literature [42,44]. As expected, age and smoking status were independently and positively correlated with the prevalence of carotid plaque in our study population.
A number of studies in HIV-infected patients used carotid IMT as a screening tool for risk assessment but this marker of early vascular aging may fail to identify patients at increased CV risk [11,12]. Carotid IMT represents a structural quantitative analysis but does not account for qualitative components. The IMT reflects both intima damages (related to the atherosclerotic process) and media damages related to an adaptive process associated with age and distending pressure . As expected, age and BP were independently associated with carotid IMT. No HIV-related parameters or PI treatment were independently associated with carotid IMT in our study population. We found that waist circumference, which is widely applied to estimate visceral obesity, was positively and independently associated with carotid IMT in our study population. In accordance with the present results, previous studies in middle-aged HIV-infected patients have highlighted that exposure to PI therapy or HIV infection might not be associated with an increased carotid IMT [46,47]. Indeed, in young and middle-aged individuals, traditional risk factors may overshadow impact of HAART exposure and HIV-infection in determining premature vascular lesions.
Pulse pressure amplification and plasma vitamin D
In our study population, vitamin D insufficiency (25OHD<30 ng/ml), and deficiency (25OHD<20 ng/ml) were prevalent. Plasma vitamin D level was positively associated with PPA in our study population, and persisted in multiple regression analysis independently of MAP, gender and heart rate. Age was not independently related to PPA, although age and female gender are considered the most important non-modifiable determinants of decreased PPA . However, the independent value of age may be minimized in the present study: in our population, gender may represent a stronger predictor than age (since the age range was quite narrow), which could possibly attenuate the independent value of aging. As expected, heart rate was positively correlated with PPA .
Amplification of systolic and pulse pressures from central to peripheral sites depends on the timing and amplitude of the reflected waves and is thus closely related to aortic PWV and AIx. The decrease or disappearance of PPA has been shown to be predictive of an increased risk of all-cause and CV mortality in large community-based cohorts [48,49]. Although the strength of the correlation between vitamin D and PPA may appear weak in our study population, this positive relation did persist in the multiple regression analysis independently of the other cofactors. Our findings suggest that higher vitamin D levels may be associated with better central hemodynamic profile in HIV-infected patients, which is consistent with epidemiological evidence in healthy individuals. In fact, vitamin D levels have been inversely related to aortic stiffness, central AIx, central aortic systolic and diastolic BP, independently of age [50,51]. Mortality has been inversely and independently related to vitamin D deficiency in large prospective cohort studies [24,52]. In HIV-infected patients, vitamin D deficiency has been associated with endothelial dysfunction , and coronary lesions . Recently, low vitamin D level has been shown to predict short term mortality in HIV-positive persons, in relation with exacerbated inflammation .
Mechanisms underlying the link between vitamin D status and CV disease may be related to central hemodynamic profile associated with aortic stiffness and/or wave reflections. However, these associations are derived from observational and cross-sectional studies and do not necessarily imply a causal relationship. Nevertheless, the growing body of observational data linking vitamin D insufficiency with the risk of mortality and consistent findings of high rates of low vitamin D levels in HIV-infected patients warrant further interventional studies to investigate the relationship between vitamin D, arterial parameters and CV health.
Our study has limitations common to cross-sectional design, especially in establishing cause-effect relationships between central hemodynamics, immunological parameters and vitamin D status. As our population was made mainly of well controlled infected patients, we could not investigate relationship between central hemodynamic parameters and overt viral replication, poorer immune status or poorer immune restoration. Immune restoration was fairly good in this population with a median ofcurrent CD4 T-cell count approaching 500/mm3 and with 94% of patients having a CD4 T-cell count above 200/mm3. Virological control is demonstrated by the fact that 90% of patients had a plasma viral load below the technique's detection limit, 4% had a viral load that was detectable below 400 RNA copies/ml (400 copies being the threshold for defining virological failure) and only 6% had a viral load above 400 copies/ml. The variable "viral load" showed a very skewed distribution. Thus, we could not incorporate this variable in the statistical analysis because of the small number of patients with a viral load of 400copies/ml or more.
The relatively small size of our study population may have limited the strength of the correlations between HIV-related parameters, vitamin D status and central hemodynamics. Finally, the lack of HIV-negative control group may weaken some of our conclusions on the impact of immune status on arterial parameters. However, we believe that European reference values established for PWV in an extremely large population represent an important step towards the definition of normal and reference values, to which our population can be compared. Mattace-Raso et al. noted that diabetic subjects and subjects treated for hypertension and dyslipidaemia had significantly elevated PWV values, compared with untreated patients, even after correction for age and MAP and this is why the authors decided not to include these patients . In our study population, 5% of patients were diabetics and 20% and 22% were on statin and anti-hypertensive drugs, respectively. Statistical analysis comparing measured values and calculated theoretical values for aortic PWV has included all patients of our study population, which may strengthen our results. In the absence of reference and widely used values for PPA and carotid AIx, comparisons between HIV-infected and uninfected populations could not be undertaken and only determinants of these vascular parameters could be assessed in our study population.
In contrast to previous studies, our group of well-controlled HIV-infected patients did not show overtly increased aortic stiffness, when compared to the theoretical values. To the best of our knowledge, it is the first time that better immune status was associated with a transit time-dependent marker of intensity of wave reflections, in relation with the fact that patients under treatment with higher CD4 T-cell count may have unaltered buffering function. The positive correlation between vitamin D and amplification of PP suggests a possible beneficial impact of adequate vitamin D level on central hemodynamic parameters. Amplification of PP, which is modulated by vascular properties and inversely related to large artery stiffness and increased wave reflections, may provide further information about the relationship between vitamin D, arterial parameters and CV health in HIV-infected patients.
The authors thank Agnès Cros and Aline Maignan for data collection and monitoring. This study was performed with the help of the grant from the French Society of Hypertension.
Make the best use of Scientific Research and information from our 700 + peer reviewed, Open Access Journals