The effect of hypercapnia on the directional sensitivity of dynamic cerebral autoregulation and the influence of age and sex

Abstract

The cerebral circulation responds differently to increases in mean arterial pressure (MAP), compared to reductions in MAP. We tested the hypothesis that this directional sensitivity is reduced by hypercapnia. Retrospective analysis of 104 healthy subjects (46 male (44%), age range 19-74 years), with five minute recordings of middle cerebral blood velocity (MCAv, transcranial Doppler), non-invasive MAP (Finometer) and end-tidal CO₂ (capnography) at rest, during both poikilocapnia and hypercapnia (5% CO₂ breathing in air) produced MCAv step responses allowing estimation of the classical Autoregulation Index (ARI_ORIG), and corresponding values for both positive (ARI_+D) and negative (ARI_-D) changes in MAP. Hypercapnia led to marked reductions in ARI_ORIG, ARI_+D and ARI_-D (p < 0.0001, all cases). Females had a lower value of ARI_ORIG compared to males (p = 0.030) at poikilocapnia (4.44 ± 1.74 vs 4.74 ± 1.48) and hypercapnia (2.44 ± 1.93 vs 3.33 ± 1.61). The strength of directional sensitivity (ARI_+D-ARI_-D) was not influenced by hypercapnia (p = 0.46), sex (p = 0.76) or age (p = 0.61). During poikilocapnia, ARI_+D decreased with age in females (p = 0.027), but not in males. Directional sensitivity was not affected by hypercapnia, suggesting that its origins are more likely to be inherent to the mechanics of vascular smooth muscle than to myogenic pathways.

Keywords: Cerebral blood flow; age; cerebral autoregulation; hypercapnia; sex.

Figure 1.

Representative recordings and MCAv step responses from a 52 year old female subject during poikilocapnia (left column plots) and hypercapnia (right column plots). (a,b) middle cerebral artery velocity (MCAv, continuous line) and model predicted output (symbols); (c,d) mean arterial blood pressure (MAP); (e,f) integrated positive derivative of MAP; (g,h) integrated negative derivative of MAP; (i,j) End-tidal CO₂; (k,l) normalised MCAv step responses for MAP_+D (continuous line) and MAP_-D (dashed line). The poikilocapnia and hypercapnia, model fittings had R_mod of 0.363 and 0.934, respectively. Corresponding values of ARI_+D/ARI_-D were 4.46/3.27 and 0.00/0.64, whilst the NMSE were 0.034/0.024 and 0.093/0.159, respectively.

Figure 2.

Population average MCAv step responses during poikilocapnia (a,c,e) and hypercapnia (b,d,f) for male (continuous line) and female (dashed line) subjects. (a,b) classical step responses for the mean arterial blood pressure (MAP) input; (c,d) step responses for the integrated positive derivative of MAP; (e,f) step responses for the integrated negative derivative of MAP. The error bars correspond to the largest ± 1 SE at the point of occurrence. Note the different amplitude scale for (f).

Figure 3.

Difference of autoregulation index between the integrated positive (Deriv+) and negative (Deriv−) mean arterial blood pressure (MAP) inputs for the ARMA model of the middle cerebral artery velocity response during spontaneous fluctuations in BP for male (n = 46, continuous line) and female (n = 58, dashed line) subjects. The error bars correspond to the 95% confidence limits.

Figure 4.

Autoregulation index as a function of age during poikilocapnia in female subjects (n = 53). The symbols and continuous regression line (p = 0.027) correspond to estimates from the integrated positive derivative input to the ARMA model of cerebral blood velocity in the middle cerebral artery. The dashed regression line (p = 0.060) correspond to estimates from the integrated negative derivative input. Corresponding linear regressions for male subjects had p-values of 0.21 and 0.93, respectively.