Alzheimer's: Type 3 Diabetes?

Neurodegeneration research turns to insulin for answers

Kerry Grens
Kerry Grens
Oct 31, 2007

In 2005, while testing the effects of impaired insulin signaling on the brain, Suzanne de la Monte at Brown University and her colleagues observed several unexpected phenomena in her experimental mice. Hallmarks of neurodegenerative disease had surfaced: oxidative stress, amyloid fibrils, and cell loss. "It was the craziest thing," de la Monte says. Glucose metabolism and Alzheimer's had been linked previously, says de la Monte, and perhaps her findings explained why.

Looking in the brains of patients diagnosed with Alzheimer's disease, de la Monte found reductions in insulin, insulin-like growth factor, and downstream elements such as tau, insulin receptor substrate, and kinases.1 Type 1 diabetes is a deficiency in insulin production, and type 2 is a resistance to insulin, where there is plenty of insulin but cells don't respond to it. Her group coined the term "type 3 diabetes" to explain their observations. "In Alzheimer's you have both things going on. That's why we called [what we saw] type 3, because it resembles both of them," de la Monte says.

De la Monte's findings made a splash in the media, appearing in the BBC and numerous Alzheimer's news outlets. However, in the scientific community, "I wouldn't say that the term type 3 diabetes has caught on," says Greg Cole at the University of California, Los Angeles.

The term might not be popular because what is happening in the brain is still unclear. Cole says accumulated evidence suggests that insulin and insulin-like growth factor signaling is impaired in patients with Alzheimer's disease. "It looks like in Alzheimer's disease you end up having a defect in these kinds of pathways, which are similar to the pathways for insulin-resistant diabetes," says Cole. But, it's unknown as to which comes first, the disease or the insulin resistance, although de la Monte is confident that the signaling defects precede the disease. As for a decrease in insulin in the brain, "if that deficit is important, we don't know," says Cole.

In July of this year, Morris White at Children's Hospital Boston found that insulin might actually be bad for the brain. He knocked out insulin signaling in mice and found that they live nearly 20% longer.2 Although he didn't conduct a cognitive assay, the animals appeared more resistant to oxidative stress, which should be protective against neurodegeneration. "Attenuated insulin signaling in the brain is probably a good thing," White says.

Excess insulin is also thought to compete with amyloid plaques for degradation, thereby contributing to their nefarious accumulation in the brains of Alzheimer's patients. White says that while these results appear to oppose de la Monte's findings that insulin deficiency is the problem, he agrees that approaching Alzheimer's as a diabetes-like disorder is a good direction to follow.

While scientists continue to work out the insulin-neurodegeneration link, drug companies are zooming ahead to capitalize on it. GlaxoSmithKline has begun Phase III clinical trials on its type 2 diabetes drug, Avandia, for treatment of Alzheimer's. White says the approach makes sense, as it would increase insulin sensitivity and reduce insulin in the brain. Suzanne Craft at the University of Washington is trying the opposite approach, supplying the brain with additional insulin. So far she's found promising results on memory, and clinical trials are in Phase II.

The potential success of both approaches suggests that de la Monte's idea that too little insulin and White's idea of too much insulin might both be operating Alzheimer's disease. "It's not just too much or too little," says Cole, "but proper regulation is what you need."


1. E. Steen et al., "Impaired insulin and insulin-like growth factor expression and signaling mechanisms in Alzheimer's disease - Is this type 3 diabetes?" J Alzheimer Dis, 7:63-80, 2005. 2. A. Taguchi et al., "Brain IRS2 signaling coordinates lifespan and nutrient homeostasis," Science, 317:369-72, 2007