The blood-brain-barrier is a powerful border that carefully regulates the entry of fuel substrates and nutrients from the periphery. However, it is not capable of protecting the brain from the deleterious effects of an onslaught of refined carbohydrates, oxidized vegetable oils, and nutritionally empty processed foods. The brain is an intensely energy-hungry organ, and anything that impedes its use of glucose—such as peripheral and/or central insulin resistance—will have disastrous consequences for cognitive function. Alzheimer’s disease is the end stage manifestation after a significant number of neurons have “starved to death” due to a loss of their ability to metabolize glucose.

Although the outward manifestations of AD—such as memory loss, confusion, and disturbing behavioral changes—are easy to observe, there are also physiological factors that can be measured and quantified. One of the earliest and most profound observable biochemical changes in the AD brain is a reduction in the rate at which the brain uses glucose, called the cerebral metabolic rate of glucose (CMRglu). This can be measured in vivo, with AD patients showing upwards of 45% reduction in CMRglu compared to healthy, age-matched controls. Some researchers see this decline in glucose usage by the brain as the predominant abnormality in AD

Interestingly, the decline in CMRglu can be observed in people at risk for AD (based on family history or genotype) as early as in their 30s or 40s, long before overt signs of AD have manifested. Thus, the decreased CMRglu can be seen as a kind of “canary in the coal mine”—an early warning sign that something is going awry in the brain. The extent of the reduction in CMRglu is tied to AD severity. A longitudinal study using PET scan to measure CMRglu.  in people age 50-80 showed that reduced hippocampal CMRglu at baseline predicted progression from normal cognitive function to AD, with the greatest reductions at baseline correlating with the quickest development of full-blown AD.

At baseline, hippocampal glucose metabolism in people who progressed from healthy to AD was 26% below that of people who did not develop AD, and the annual rate of decline averaged 4.4%. In people who progressed from normal to mild cognitive impairment (a precursor to AD), CMRglu was 15% reduced at baseline, with an annual rate of decline at 2.4%. The rate of decline for people who had normal CMRglu at baseline and did not develop AD was just 0.8%. Assuming the rates of decline were somewhat constant, extrapolating backward indicates that the decline may have started as early as 20 years before overt signs of AD were present. At baseline, despite the already decreased CMRglu in some subjects, all subjects were cognitively normal. This suggests that a starting point of reduced glucose usage in the brain and a stronger rate of continued decline might be one of the first triggering events in AD. The brain may be able to compensate for years before damage is so widespread that overt symptoms are observable. The normal forgetfulness and foibles we associate with “just getting older”—Where did I leave my keys? Don’t I have an appointment somewhere this week?—might be the earliest indicators that the brain is struggling to fuel itself.

An interesting potential contributor to the reduced CMRglu is peripheral and/or central insulin resistance. Plasma concentration of insulin is positively correlated with AD severity.  When neurons become insulin resistant, they are afflicted by the same pathology that occurs in the periphery—an inability to properly metabolize glucose, causing glucose to accumulate in extracellular spaces for an extended period of time. This results in rampant glycation and the formation of advanced glycation end products (AGEs). These AGEs add insult to injury by forming cross-linkages with each other that may alter the shape of neuronal synapses and impede cellular communication and nerve impulse transmission in the brain, with cognitive abnormalities being an obvious consequence. With hyperinsulinemia affecting 40% of people over age 80, it’s no surprise to find a link between insulin dysregulation and a condition that preferentially strikes older individuals. Moreover, hyperinsulinemia has been found to be and independent risk factor for AD.

The beta-amyloid (Aβ) plaques often implicated as a cause of AD may, in fact, be a result of peripheral hyperinsulinemia. In addition to the reduced CMRglu, the presence of insoluble Aβ plaques is one of the defining signatures of AD pathology. However, Aβ is a normal product of protein degradation, and there is no evidence that AD patients overproduce Aβ. Rather, the problem seems to be that Aβ isn’t cleared away as it should be, which results in these small, otherwise soluble peptide fragments aggregating into insoluble plaques. (These plaques are then subject to glycation and blocking synapses, adding yet another obstruction to neuronal communication.)

A fascinating thing to note is that what is responsible for clearing away Aβ in a timely manner—before it dwells long enough to form plaques—is insulin degrading enzyme (IDE), the same enzyme that clears away insulin. However, the affinity of IDE for insulin is so high that even small amounts of insulin completely the degradation of AB. One study demonstrated that peripheral infusion of insulin in older subjects increased the level of AB in cerebrospinal fluid within 120 minutes, and this also correlated to decreased memory function. Thus, the formation of Aβ plaques is facilitated by hyperinsulinemia. Adding yet another piece of evidence to the theory that Aβ plaques are an effect of AD pathology, rather than its cause, is the fact that the decline in CMRglu precedes the formation of the plaques. Therefore, the presence of Aβ plaques is not likely the triggering factor. (They may exacerbate disease severity, but they are not the initial event in its initiation.)

Considering the connections between impaired glucose metabolism, chronically elevated insulin, and Alzheimer’s disease, the phrase “type 3 diabetes” is viable.