What is the Ambulatory Stiffness Index and What Is Its Role in Patients With Lacunar Infarcts?

The ambulatory arterial stiffness index (AASI) was introduced in 2006 as an index easily derived from ambulatory blood pressure (ABP) monitoring.1 This index is represented as unity minus the regression coefficient of diastolic on systolic blood pressure (BP). It was found to correlate with pulse pressure (PP)2 and pulse wave velocity (PWV)3 and was thus presumed to be another marker of large artery stiffness, which is readily available from ABP. Moreover, one of the most attractive features of AASI was that it was found to be a significant predictor of stroke,4 cardiovascular disease,1, 4, 5, 6 renal dysfunction,7, 8, 9, 10, 11 hypertensive target organ damage,12, 13 and all‐cause mortality,14 and, in many of these studies, independently from BP or PP. This independence of BP (or PP) brought about investigations of its possible arterial properties. AASI was found to be associated with change in endothelin with salt loading15 and intake16 and to be inversely associated with physical activity.17 AASI has also been associated with the metabolic syndrome18; however, the mechanisms of these arterial properties remain unclear. Moreover, not all studies have persistently shown the previously mentioned association with PWV, considered by many as the “gold standard” of large artery stiffness.19 From early on, the introduction of the AASI has proposed many inconsistencies: AASI seems to be dependent on age and sex20 (not surprising if it represents arterial stiffness) and on nocturnal BP decline,20, 21 and it appears to be inferior to PP in predicting albuminuria in elderly diabetic patients22 and less reproducible than PP in type 1 diabetes patients.23 Gavish and colleagues24, 25 found that some of the problems associated with AASI (for instance its dependence on nocturnal BP decline) can be improved when using a more advanced form of regression of systolic and diastolic BP, the so‐called symmetric regression, which, in this context, provides a slope between that of diastolic on systolic and that of systolic on diastolic BP regression lines. This symmetric AASI (sAASI) has been found to better represent arterial stiffness in patients with HIV,26 to be associated with cystatin C as a measure of renal dysfunction,11 and to be associated with estimated glomerular filtration rate and microalbuminureia.10 Moreover, Gavish and colleagues found that their regression slope was both a mathematic and empiric ratio of the variability of systolic BP over that of diastolic BP, ie, the BP variability ratio (BPVR). This finding may elucidate some of the yet unclear physiologic significance of sAASI as an arterial property that accounts for the BP variability in patients.25 This slope, the BPVR, and the derived bpvrAASI=1‐1/BPVR, was also associated with all‐cause 5‐year mortality beyond that of PP and additional confounders. According to Gavish,27 the sAASI can be best regarded as an index of the stiffening of the arterial wall during systole from the lower diastolic to the higher systolic pressure, because arterial wall stiffness is dependent on the momentary prevailing arterial pressure.

Arguments relevant to the issues above, and others concerning AASI, can be found in a relatively recent comprehensive review of the subject.19 Nevertheless, despite its predictive power, AASI remains ambiguous because of its questionable association with PWV and its physiologic significance.

In this issue of The Journal of Clinical Hypertension, Klarenbeek and colleagues28 describe their findings on the association of AASI with small cerebral vessel disease in a prospective cohort of patients with recent acute lacunar stroke. To focus on small vessel cerebrovascular disease they excluded patients with atrial fibrillation or significant carotid artery disease.

The authors studied a sample of 143 patients from two medical centers in the Netherlands. They found that the highest AASI tertile was associated with white matter hyperintensities but not lacunae, perivascular space widening, or cereberal microbleeds. After adjustment for age and sex, the association of AASI or sAASI with white matter hyperintensities lost its significance. Conversely, 24‐hour systolic and diastolic BP tertiles were significantly associated with lacunae, perivascular space widening, and cereberal microbleeds.

What can we make of these findings? First, we must note the small sample size of the studied population, which may not be representative of the lacunar stroke population at large. Among other reasons, it is surprising that about a third of the patients with symptomatic lacunar stroke had no relevant lesion on magnetic resonance imaging. The authors cite criteria for the diagnosis of lacunar stroke in the absence of an imaged pathology; however, they cite a study published in 1987, long before magnetic resonance imaging was an integral part of stroke imaging.

Nevertheless, 24‐hour ABP was not associated with white matter hyperintensities, whereas it is the most common finding in the general population29 and in patients with stroke.30 It is also a robust finding significantly associated with 24‐hour ABP.31 The association of increasing ABP with small vessel cerebral vascular disease increases on repeat examination.32

Moreover, in a landmark 24‐hour ABP study on the association of stroke and pulsatile PP, or the steady component of BP, mean arterial pressure (MAP), only the latter was associated with stroke.33 Indeed it is evident that across AASI tertiles, MAP decreases as a result of lower BP, especially diastolic BP, which has a major influence on MAP. In this respect, it may not be surprising that in such a population, AASI is less related to small vessel cerebral vascular disease after a lacunar stroke.