Autoregulation of the cerebral circulation during sleep in newborn lambs

Abstract

Autoregulation is a vital protective mechanism that maintains stable cerebral blood flow as cerebral perfusion pressure changes. We contrasted cerebral autoregulation across sleep-wake states, as little is known about its effectiveness during sleep. Newborn lambs (n= 9) were instrumented to measure cerebral blood flow (flow probe on the superior sagittal sinus) and cerebral perfusion pressure, then studied during active sleep (AS), quiet sleep (QS) and quiet wakefulness (QW). We generated cerebral autoregulation curves by inflating an occluder cuff around the brachiocephalic artery thereby lowering cerebral perfusion pressure. Baseline cerebral blood flow was higher (P < 0.05) and cerebral vascular resistance lower (P < 0.05) in AS than in QW (76 +/- 8% and 133 +/- 15%, respectively, of the AS value, mean +/-s.d.) and in QS (66 +/- 11% and 158 +/- 30%). The autoregulation curve in AS differed from that in QS and QW in three key respects: firstly, the plateau was elevated relative to QS and QW (P < 0.05); secondly, the lower limit of the curve (breakpoint) was higher (P < 0.05) in AS (50 mmHg) than QS (45 mmHg); and thirdly, the slope of the descending limb below the breakpoint was greater (P < 0.05) in AS than QS (56% of AS) or QW (56% of AS). Although autoregulation functions in AS, the higher breakpoint and greater slope of the descending limb may place the brain at risk for vascular compromise should hypotension occur.

Figure 1. Physiological recordings during induced hypotension in sleep

Physiological recording from a single lamb collected before, during (inflation) and after decreasing cerebral perfusion pressure by inflating the brachiocephalic cuff in each of the three behavioural states studied. During the control periods, cerebral perfusion pressure is equal in each state yet cerebral blood flow is higher in active sleep than it is in quiet wakefulness, which in turn is higher than it is in quiet sleep. During the inflation of the brachiocephalic occluder cuff, cerebral perfusion pressure was lowered rapidly to approximately the same level (40 mmHg) in each state. In both QW and QS, the effects of autoregulation are rapidly evident as cerebral blood flow begins to recover towards the control level, a recovery that is underpinned by cerebral vasodilatation. The recovery towards control levels of cerebral blood flow is more rapid in QS than it is in QW. Moreover, both the magnitude and the speed of this recovery are substantially less in AS than in either of the other states. EMGn = electromyogram of the nuchal (neck) muscles, EOG = electrooculogram, ECoG = electrocorticogram.

Figure 2. Average data recorded during the control period in quiet wakefulness (QW), active sleep (AS) and quiet sleep (QS)

Although carotid artery pressure and intracranial pressure were lower in AS than in QW and QS, cerebral perfusion pressure did not differ between the three behavioural states. Nevertheless, cerebral blood flow was greater in AS than in either QS or QW, and greater in QW than in QS. Also, cerebral vascular resistance was lower in AS than in either QS or QW and lower in QW than in QS. Values = mean ± s.d., n = 9. Bars indicate significant differences: *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 3. Average autoregulation curves derived for each behavioural state

Bilinear regression analysis was used to determine the breakpoint position of each autoregulation curve. The breakpoint pressure was identified as the point on the pressure axis at which the combined sum of the residuals from the two least square regressions (residual sums of squares) was minimal (lower panel). Note that a plateau phase in each autoregulation curve (upper panels) was underpinned by cerebral vasodilatation (middle panel) in response to lowering of cerebral perfusion pressure. Also note that the breakpoint in cerebral blood flow coincides with a breakpoint in the cerebral vascular resistance (CVR) values.

Figure 4. Average autoregulation curves generated from the nine lambs studied in each of the behavioural states

Data were normalized to the control value recorded for each lamb in quiet sleep and replotted to reveal the relationship that exists between sleep state and autoregulation. Note that the sleep state-dependent differences in cerebral blood flow are maintained across all levels of cerebral perfusion pressure. Note also that cerebral blood flow in active sleep (AS) exceeds that in quiet wakefulness (QW) which in turn exceeds that in quiet sleep (QS). Moreover, the position of the breakpoint of the autoregulation curve is shifted to the right in AS (50 mmHg) relative to that in QW (46 mmHg) and QS (45 mmHg). As a result of the higher cerebral blood flow in AS, the slope of the descending limb of the autoregulation curve in AS (y= 2.5x+ 34) is also steeper than in QS (y= 1.4x+ 36) or QW (y= 1.4x+ 52). Both these characteristics (higher breakpoint and greater slope) may place the brain in danger of ischaemia and hypoxia should hypotension occur during AS.