Nutrition of the brain: macronutrient supply

Abstract

Unit for Metabolic Medicine, 4th Floor, Hunt's House, Guy's Hospital, London SEl 9RT GLUCOSE It has long been accepted that the human brain uses glucose as its only metabolic fuel and, thus, is entirely dependent on glucose for its function. This is in spite of the fact that the brain contains many enzyme systems theoretically capable of metabolizing non-glucose substrates such as glycerol, fatty acids, lactate, ketones and amino acids. Nevertheless, it is true that the brain is the major consumer of glucose in the resting state and about 10% of the blood glucose is extracted by the brain. Of that glucose, over 90% is fully oxidized to C02 and water with the generation of high-energy phosphates. Perhaps about 5% of brain glucose is metabolized through the hexose monophosphate shunt and the remainder through glycolysis to lactate and pyruvate, and only a very small quantity is synthesized into glycogen. The glycogen stores of the brain are very small and do not provide a useful reservoir of glucose in times of glucose lack. It is undoubtedly true that failure of the blood glucose supply to the brain produces significant loss of brain function (Amiel et al. 1991b). Indeed, the body is well designed to prevent such falls in glucose. If blood glucose does begin to fall, in response to prolonged starvation or extreme exercise or to insulin administration, a counter-regulatory response is initiated which acts both to restore blood glucose levels and produce a symptom complex that stimulates the subject to eat (Amiel, 1991). In terms of blood glucose restoration, the most important part of the endogenous response is probably the cessation of pancreatic insulin secretion and the stimulation of pancreatic glucagon. These hormone changes stimulate hepatic glucose production causing blood sugars to rise again. Almost equally effective and clinically very important is the stimulation of adrenaline secretion and of the sympathetic nervous system. Together with cortisol and growth hormone secretion, these responses not only stimulate and support hepatic glucose production, but also lower the rate at which peripheral tissue such as muscle and fat can take up glucose from the circulation. Furthermore, the adrenaline and sympathetic nervous system stimulation is associated with symptoms of anxiety, sweating, tremor and palpitations and hunger, stimulating eating, the most effective way of correcting the situation (Hepburn et al. 1991). Only if these mechanisms fail and/or blood glucose is forced below 3 mmoY1, do the higher cerebral functions of 402 S . A . AMIEL psychomotor coordination and cognitive function become impaired (Widom & Simonson, 1990; Maran et al. 1993). It is believed that cerebral metabolism is involved not just in cognitive processes but also in the recognition of developing hypoglycaemia and the initiation of much of the counter-regulatory response. Evidence for this theory comes from the selective catheterization experiments of Cherrington's group in Nashville (Biggers et al. 1989). They catheterized dogs in order to be able to perfuse separately the body and the brain. When the animals were rendered bodily hypoglycaemic but brain glucose levels were maintained, the hormonal and glucose kinetic responses of counter-regulation were virtually obliterated, suggesting that cerebral hypoglycaemia was necessary to initiate these protective responses. In studies in man, Nagy et al. (1992) have renovated an older procedure of measuring cerebral metabolism by measuring cerebral blood substrate arterio-venous differences across the brain. They have been able to calculate rates of cerebral glucose metabolism across the whole brain and dernonstrate quite clearly a decrement in cerebral blood glucose uptake in normal individuals at a blood glucose level of approximately 3.8 mmoUl, an arterial blood glucose around or slightly above that necessary to stimulate adrenergic counter-regulatory responses and certainly higher than that usually believed to be associated with cognitive dysfunction. The rate-limiting step in cerebral glucose metabolism is cerebral glucose uptake. McCall et al. (1986) have demonstrated that the process can adapt itself to prevailing glucose levels. Acute hypoglycaemia produces no alteration in the rate at which brain tissue can extract glucose in vitro, but when the experiments are repeated after 4 d of moderate hypoglycaemia, rates of brain glucose extraction are significantly increased. The mechanism of adaptation is by means of an increase in the number of membrane glucose transporters. Correlations can be made between humans and studies in animals. Some patients with insulin-dependent diabetes mellitus (particularly those with long-term disease and/or very tight metabolic control) exhibit a phenomenon now known as loss of awareness of hypoglycaemia. Underlying this is a failure to generate adrenergic and symptomatic counter-regulatory responses to hypoglycaemia until blood sugar has fallen much lower than the level that would normally be required (Amiel et al. 1988). For example, adrenaline responses to hypoglycaemia do not begin in such individuals until blood sugar has reached about 2.2 mmoVl, whereas in non-diabetic or less-well-controlled diabetic patients, such responses begin at a blood sugar level of 3.6 mmol/l. Sincle in all subjects cognitive function (measured by four-choice reaction time) begins to deteriorate at a mean blood glucose of about 2.8 mmoUl, in the hypoglycaemia-unaware patient cognitive function deteriorates at a higher blood glucose level (i.e. earliler) than that at which symptomatic responses occur. By the time glucose is low enough to generate responses that should produce symptoms, the patient may be too cognitively impaired to respond. Boyle's group (Nagy et al. 1993) have shown that in normal individuals a 2 d period of moderate insulin-induced hypoglycaemia results in a maintenance of blood glucose utilization rates during induced hypoglycaemia, until blood gluclose reaches 2.5 mmoUl. In the hypoglycaemia-naive subjects, brain glucose utilization has fallen considerably by the time blood glucose is 3.6 mmoUl. While in this study, hypoglycaemia-naive subjects lost motor function when their blood glucose fell to 2.5 mmoU1, after 2 d of hypoglycaemia, these subjects were able to retain motor function in the face of a similar blood glucose level. However, intellectual function ,assessed by the