Rate of cerebral blood flow (CBF). CBF rates are largely determined by the cerebral metabolic rate for oxygen consumption; there is exquisite coupling of blood flow and metabolism on a regional basis. The global blood flow to the brain remains fairly stable for a given physiologic state (e.g., awake adult). Anesthetics and hypothermia tend to decrease metabolism throughout the brain and thereby reduce global CBF.
The global CBF in adults is approximately 50 mL/100g/minute. This global measure is composed of flow from two very different regions: the gray matter, which is where neuronal cell bodies and synapses are located and has a blood flow of 75 mL/100 g/minute, and the white matter, which consists mainly of fiber tracts and has a blood flow of 20 mL/100 g/minute. The higher blood flow to gray matter is primarily due to its greater metabolic rate.
The global CBF of children is approximately 95 mL/100 g/minute, which is higher than that of adults. In contrast, infants have a slightly lower CBF than adults (40 mL/100 g/minute).
Spinal cord blood flow has been less extensively studied; the gray matter has a rate of 60 mL/100 g/minute and the white matter a rate of 20 mL/100 g/minute.
Regulation of CBF. Regional flow and metabolism coupling depends on the buildup of metabolites that cause local dilatation of the microvessels.
The precise mechanism of this coupling is unknown. It may be due to the buildup of K or H in the extracellular fluid surrounding the arterioles. Other agents that may mediate flow and metabolism coupling include nitric oxide (NO), calcium, adenosine, and eicosanoids such as thromboxane and prostaglandin. A combination of these factors likely contributes to coupling.
NO is a vasodilator that is released locally by the vascular endothelial cells. This vasodilatation is important for vascular regulation throughout the body, but the precise role of NO in the control of CBF remains to be determined. It is likely to be an important regulator of local CBF, perhaps by affecting arterioles upstream of the microvessels dilated by metabolic factors.
Carbon dioxide (CO2) enhances vasodilatation and increases CBF. CO2 is hydrated with the help of carbonic anhydrase leading to an acidification. This reduction in pH is thought to cause the vasodilatation. When CO2 is halved from 40 to 20 mm Hg, the CBF is reduced by approximately half (Figure-1).
Hyperventilation leads to a reduction in CO2, an increase in pH, and thereby a reduction in CBF. If hyperventilation is maintained over a period of 6 to 8 hours, the pH returns to normal due to bicarbonate transport and CBF returns to its prehyperventilation levels. Thus, hyperventilation is useful for only short periods of time.
(1) If hyperventilation is discontinued abruptly, the increase in CO2 to the normal level, in the presence of reduced HCO3, leads to acidosis and an above normal increase in CBF. This can be a critical problem when hyperventilation is used to reduce intracranial pressure (ICP).

Figure-1. Relationship between cerebral blood flow and arterial carbon dioxide tension (PaCO₂) in the normocapnic adult, the hypercapnic adult, and the newborn.

 
(2) It is possible that extreme hyperventilation to levels below 20 mm Hg can lead to ischemia due to vasoconstriction; thus, hyperventilation below 30 mm Hg is not recommended clinically.
Autoregulation allows CBF to remain constant if the cerebral perfusion pressure (CPP) varies between 50 and 150 mm Hg. Signs of cerebral ischemia are seen below 50 mm Hg; above 150 mm Hg, disruption of the blood-brain barrier (BBB) and cerebral edema may occur. The adjustment of flow to abrupt changes in pressure requires 30 to 180 seconds (Figure-2).
The mechanism of autoregulation is not completely understood but is likely to be a combination of effects including myogenic and metabolic factors.
(1) Myogenic activity of the vessel wall musculature occurs in response to increased distending pressure. In isolated vessels, when the vessel wall is stretched, as it would be by increased blood pressure, the smooth muscle contracts, causing a vasoconstriction that reduces flow. This balances out the increase in blood flow due to the increased pressure, resulting in little net change in blood flow.

Figure-2. Autoregulatory curve of the cerebral vasculature in the normotensive adult, the hypertensive adult, and the newborn.

 
(2) The metabolic theory states that reduced pressure leads to reduced flow and the buildup of metabolites. This buildup in metabolites and the decrease in local pH lead to a local vasodilatation and thereby an increase in blood flow back to normal.
(3) Autoregulation can be impaired by hypoxia, ischemia, hypercapnia, trauma, and certain anesthetic agents.
(4) Patients who are chronically hypertensive or have a high sympathetic tone have a shift in the autoregulatory curve to the right. They may demonstrate signs of ischemia due to reduced blood flow at pressures above the lower limit for normotensive individuals.
The partial pressure of oxygen (PaO₂) has little effect on global CBF until it falls below 50 mm Hg. At this point, a dramatic increase in blood flow with further reductions in PaO₂ occurs. Since the oxygen-carrying capacity of blood is high, a critical reduction in the oxygen content of the blood might not occur until the Pao2 falls below a threshold of 50 mm Hg (Figure-3).
Neurogenic factors including adrenergic, cholinergic, and serotonergic systems also influence CBF. Their greatest influence is on larger blood vessels.
The hematocrit alters blood viscosity and thereby can affect blood flow. A low hematocrit can increase blood flow by decreasing blood viscosity.
Hypothermia decreases neuronal metabolism and thereby reduces CBF; hyperthermia has the opposite effect.

Figure-3. Relationship between cerebral blood flow and PaO₂, arterial oxygen tension.

The metabolic rate of neurons surrounding the microvasculature regulates regional blood flow. If the activity of these neurons increases, their metabolic rate increases, and the blood flow to that region increases to meet the demand for oxygen and glucose. This property has been exploited using positron emission tomography to functionally map brain activity.