Is Alzheimer's disease actually a form of diabetes?

Health and medicine explained.
March 16 2009 1:01 PM

Diabetes of the Brain

Is Alzheimer's disease actually a form of diabetes?

PET scan of a brain with Alzheimer's disease.
PET scan of a brain with Alzheimer's disease

When the brazen James Watson had his genome sequenced, he declined to find out whether he carried a gene variant that would increase his risk for Alzheimer's disease. Ditto for Steven Pinker. There are virtually no treatment or prevention options for those who have or are at risk for Alzheimer's. Nor do scientists fully understand what causes it, though for years, opposing camps have duked it out over hypotheses that have focused largely on brain abnormalities called plaques and others called tangles, neither of which has so far proved a good therapeutic target.

Now some experts are proposing an avant-garde way of approaching Alzheimer's: as a form of diabetes. Some even dub it "type 3 diabetes" or "diabetes of the brain." The idea is that memory loss and cognitive deterioration in at least some Alzheimer's patients may be caused by low insulin or insulin resistance in the brain, much as lack of production or poor response to insulin in the body is central to the pathologies of type 1 and type 2 diabetes. Effective Alzheimer's treatments, then, might aim to boost brain insulin levels or decrease resistance while addressing destructive factors like inflammation and oxidative stress. If the theory holds up, as early research suggests, it could be a boon to a field scarred by disappointments and dead ends.

Advertisement

Alzheimer's researchers have been bitterly divided over what initially causes the disease and where to look for treatments. For years, the dominant view was that plaques—sticky deposits of a protein called beta-amyloid—were the central culprits, destroying neurons and causing cognitive decline. More recently, some researchers in the amyloid camp have begun to focus on toxic, soluble forms of the protein, rather than the plaques themselves, as the real instigator. At the same time, another faction has emphasized abnormal modifications of a protein called tau that results in so-called tangles, which also turn up in brains ravaged by Alzheimer's. But neither the plaques nor the tangles seem to account fully for the onset of the disease. Plaques often appear in the brains of elderly people without Alzheimer's, and some evidence suggests that tangles form later in the disease's progression, rather than triggering it. Nor has either abnormality yet proven to be a fruitful target for new drugs. Meanwhile, the pitched battles have done damage to the field. It's been "one army against another," a prominent researcher told me. "You see them fighting at meetings," she says, and you think, "Oh, shut up. Try to come up with something that fits both."

Which brings us to insulin. Insulin is the hormone that allows cells, including some brain cells, to take up energy in the form of glucose. Proper insulin function in the brain appears necessary to the formation and maintenance of memories. And, crucially, a lack of insulin or insulin resistance is connected both to amyloid protein regulation and to the modification of tau proteins, which can cause tangles. In other words, insulin seems to hold up a conceptual umbrella under which the amyloid and the tangle camps might finally meet. (Type 2 diabetes is also a risk factor for Alzheimer's and cognitive decline. In 2005, researchers at Brown showed that by knocking out insulin production and causing brain insulin resistance in rats, they could create a model of Alzheimer's, complete with plaques and abnormal accumulations of tau. (Suzanne de la Monte, who led this group, was the first to dub Alzheimer's "type 3 diabetes." She reviews the evidence to date on this theory here.) Scientists have also described links between abnormal insulin and other hallmarks of Alzheimer's, such as oxidative damage and inflammation. And last month, Bill Klein at Northwestern found that in an in vitro model using rats' brain cells, insulin could shield the cells from an onslaught of soluble amyloid proteins. That is, he found that when memory-forming cells from the brain's hippocampus were dosed with insulin, their cell connections were not as badly damaged by the amyloids. This suggests insulin might help to preserve or improve memory circuitry in the face of disease.

Insulin-related therapies look preliminarily good in clinical trials, too. In 2007, researchers in Seattle conducted a small, randomized clinical trial in which patients with early Alzheimer's disease received daily puffs of insulin in the nose for 21 days. (The team chose this method of administration so that the insulin would move more directly to the brain without circulating throughout the body, where it might cause an unwanted drop in blood glucose.) The group, led by Suzanne Craft of the VA Puget Sound Medical Center, found that patients who received insulin were better able to pay attention to a story that was read to them and recall details 20 minutes later. Their caregivers also rated their mental functioning more highly. The team is now conducting a larger clinical trial, with results expected this fall. Craft says she is optimistic that insulin may open new doors for Alzheimer's patients. However, she shies away from dubbing the disease a "diabetes of the brain," noting that diabetes is normally diagnosed on the basis of elevated glucose levels, which do not appear to be at play here.

Other promising, early results come from the diabetes drug rosiglitazone, which has a checkered reputation. In patients with type 2 diabetes, rosiglitazone (aka Avandia) acts to increase insulin sensitivity in the body. But it may also raise the risk of heart attacks. The hope is that in some Alzheimer's patients, at least, the drug will have benefits for the brain—without the cardiovascular downside. In a small trial in 2005, Alzheimer's patients who received rosiglitazone for six months showed better attention and better recall than those who received a placebo. A more substantial, phase-two trial, published in 2006 and sponsored by GlaxoSmithKline (which makes the drug), found that patients who received rosiglitazone also fared better after 24 weeks of treatment—but only if they did not carry a particular gene variant. (This variant, called APOE-E4, is known to predispose people to the disease.)

GSK is now sponsoring several multinational phase-three trials, involving roughly 3,000 patients in 22 countries, which are expected to show more definitively whether rosiglitazone helps to improve cognition and functional capacity in patients with mild to moderate disease. Results are expected this summer. GSK researcher Michael Gold says that the Data Safety and Monitoring Board, which oversees these studies and has access to unblinded data, has not signaled a need to change or stop the studies based on safety. Still, researchers like Craft, who has served as a consultant to GSK, suggest that other drugs in the same chemical class as rosiglitazone might ultimately offer benefits to Alzheimer's patients with less cardiovascular risk.

Of course, taking intranasal insulin or a diabetes drug like rosiglitazone is not the only possible way to boost the brain's sensitivity to insulin. Another, even better option may be to do aerobic exercise, says Craft. And this seems to hold for older adults in general, including those without Alzheimer's: Better insulin signaling and glucose uptake in the brain may offer them a cognitive boost as well. One more reason, it seems, to close the medicine cabinet and work up a good sweat.

Amanda Schaffer is a science and medical columnist for Slate.