To the Editor,

Arterial stiffness predicts cardiovascular morbidity and mortality, and can be evaluated with: pulse wave velocity (PWV),1 ambulatory arterial stiffness index (AASI),2, 3 and augmentation index (AI).4

There are studies in Spain on arterial stiffness in various subpopulations,5 but there are no studies that have jointly analyzed PWV, AASI, and AI in diabetic and hypertensive patients. The aim of this study is to report average values for these parameters in diabetic and hypertensive patients compared to controls, and to analyze determining factors.

From 2006 to 2010, we recruited 373 patients from consultations in 2 health centers. Those who had medical histories of cardiovascular disease were excluded. The sample size was estimated for detecting differences of 1 m/s in PWV between groups. By assuming an alpha risk of 0.05, a beta risk of 0.2 and a standard deviation of 2.12 m/s, 95 subjects per group were needed.

The protocol was approved by the research ethics committee and all patients signed an informed consent form.

The PWV and AI were estimated with the SphymgoCor System. We performed the aortic pulse wave analysis with a sensor on the radial artery with the patient seated. Using mathematical transformation based on the radial wave, we estimated the central blood pressure and the central and peripheral AI. We evaluated reliability by intraclass correlation (r=0.974; 95% confidence interval [CI], 0.936-0.989) in repeated measurements of 22 subjects. Carotid-femoral PWV was determined with the patient lying down, estimating the pulse wave delay at the carotid and femoral level with respect to the electrocardiogram. We defined AASI as 1 - the regression slope for the diastolic blood pressure (DBP) and the systolic blood pressure (SBP) of 24-hour readings, and the Sym-AASI as 1 - (1- AASI) / r.

We assessed the association between qualitative variables with the χ2 test, and the difference of the means with ANOVA. We controlled for age with ANCOVA. We analyzed the variables related to the parameters of arterial stiffness with stepwise multiple regression analysis, adjusted for age and sex. PWV, AASI and central augmentation index (cAI) were used as dependent variables. The independent variables were smoking, SBP, DBP, heart rate (HR), body mass index, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, baseline glycemia, C-reactive protein and fibrinogen. We used the SPSS/PC+18.0 statistical program.

Table 1 shows the cardiovascular risk factors, average values of the stiffness parameters, and average antihypertensive and lipid-lowering drugs per group. After adjusting for age, the differences between diabetic patients and controls were: PWV, 1.13m/s (95% CI, 0.51-1.74) (P<.01); AASI, 0.01 (95% CI, –0.01- 0.03); and cAI, 2.87 (95% CI, –0.82-6.56). Differences between hypertensive patients and controls were: PWV, 0.67m/s (95% CI, 0.14-1.21) (P<.05); AASI, –0.01 (95% CI, –0.03-0.01); and cAI, 5.45 (95% CI, 2.26-8.66) (P<.05). In diabetic, hypertensive, and control patients, the correlation between PWV and AASI was 0.42 (P<.01), 0.38 (P<.01) and 0.19; between PWV and cAI, 0.15, –0.12 and 0.33 (P<.01) and between AASI and cAI, 0.1, 0.03 and 0.02, respectively.

Table 1. Clinical Characteristics, Cardiovascular Risk Factors, and Values for Arterial Stiffness Parameters.

BMI, body mass index; CRP, C-reactive protein; DBP, diastolic blood pressure; HDL-C, high-density lipoprotein cholesterol; HR, heart rate; LDL-C, low-density lipoprotein cholesterol; OBPM, outpatient blood pressure monitoring; PP, pulse pressure; SBP, systolic blood pressure; Sym-AASI, symmetrical ambulatory arterial stiffness index.
Data is presented as mean±standard deviation or no. (%).

a Differences between diabetic and control patients.
b Differences between hypertensive and control patients.
c Differences between diabetic and hypertensive patients (P<.05).

The factors that best explain the arterial stiffness parameters in PWV, adjusted for age and sex, are SBP and waist circumference. They achieve the best fit (r2=0.6) in diabetic patients. AASI is explained by SBP and DBP, with the best fit in diabetic patients (r2=0.61). For cAI, the variable that is maintained in all cases is HR, with the best fit (r2=0.55) in controls (Table 2).

Table 2. Multiple Regression Analysis of Arterial Stiffness Parameters in Diabetic, Hypertensive, and Control Patients.

AASI, ambulatory arterial stiffness index; BMI, body mass index; cAI, central augmentation index; DBP, diastolic blood pressure; HR, heart rate; PWV, pulse wave velocity; SBP, systolic blood pressure.

The data obtained allow us to report the values and factors associated with PWV, AASI and AI in hypertensive, diabetic and control patients. PWV, but not AASI or AI, has a similar behavior in diabetic and hypertensive patients, after adjusting for age.

The values found for PWV, adjusted for age, are similar to those of the Framingham study.1 When these values are compared with reference values from Europe, according to age and blood pressure readings, they are higher in hypertensive patients and the same in controls.6 As in previous studies, the variability of PWV is greatly influenced by age and blood pressure.

AASI values were lower than those of the Dublin Outcome study2 (AASI of 0.41, and 0.46 in diabetic patients) and the Ohasama study3 (AASI of 0.45 in nonhypertensive patients, and 0.46 in hypertensive patients) with no differences between groups, as in our study. The differences may be due to the fact that neither age nor distribution of other risk factors is the same. AASI variability is influenced by age, sex, blood pressure, and HR.

cAI varies according to the subjects’ characteristics, as shown in 11 studies that analyzed central blood pressure values.6 Heart rate, age, sex, and blood pressure are some of the variables that are maintained in the multiple regression.

In conclusion, stiffness parameters do not behave uniformly. Thus, while PWV adjusted for age is greater in diabetic patients than in hypertensive ones and is greater in hypertensive patients than in controls, the same does not happen with the other evaluated parameters. These results would support systematic evaluation of PWV in diabetic and hypertensive patients.

This article includes data from the “Target organ damage and outpatient blood pressure monitoring” study (Lesión de órganos diana y monitorización ambulatoria de la presión arterial, LOD-RISK) and has been funded in its various phases and subprojects by the regional health management entity (GRS/47-05, GRS/167/A/07 and GRS/254/A/08) and the regional department of health of Castile and Leon (SAN/196/SA36/07) and by ISCIII-RETICS, cofinanced with FEDER funds (RD06/0018).

None declared.

.

Corresponding author: magomez@usal.es