According to epidemiological data, those with diabetes have anywhere from a two to five-fold increased risk of Alzheimer’s-type dementia (AD). I’ve known of this link for a while, but have had my doubts about its validity. To me, it was equally plausible that these supposed cases of Alzheimer’s were just misdiagnosed cases of vascular dementia. Diabetes unequivocally leads to widespread atherosclerotic disease, including in the vascular beds supplying the brain, and “vascular” dementia is a clear consequence of this process. In these cases, though, the pathology is distinct from dementia of the Alzheimer’s type, as are the typical clinical features. But few patients receive the type of evaluation that can tease out these differences, and misdiagnoses commonly occur. Any correlations based on epidemiological studies that use this diagnostic data, then, are suspect.
The other problem with this correlation is that our pathogenetic model for Alzheimer’s doesn’t account for it. Ever since Alois Alzheimer presented the case of Auguste D in 1906, senile plaques and neurofibrillary tangles have been recognized as the pathologic signature of Alzheimer’s disease in the brain — the former clumps of beta-amyloid protein, the latter “paired helical filaments” of hyperphosphorylated tau.
Based on this pathology, the prevailing idea was that some intrinsic abnormality in one of these proteins was the driving force behind AD. As such, the central debate within the Alzheimer’s research community in recent years revolved around whether we should be trying to figure out how to keep beta amyloid from clumping or tau from phosphorylating, with the amyloidists and tauists arguing over whose piece of the funding pie should be larger. Nowhere in these competing models was there a mechanism by which impaired glucose clearance (i.e. diabetes) led to the Alzheimer’s pathology. So, if the link is real, it means that either both the amyloidists and tauists are wrong or the connection between diabetes and AD risk is false.
Supposing for a moment that the diabetes-Alzheimers link is legit, just how might one lead to the other? The fundamental problem in diabetes is the inability to adequately clear glucose from the bloodstream, and complications arise through the formation of “Advanced Glycation End Products” (AGEs) — a glucose molecule, left hanging around in the circulation for too long, ends up sticking to proteins in bodily tissues, disrupting their structure and function. Any sugar molecule (i.e. not just glucose), in fact, may react with a protein and form an AGE. The more AGEs, the greater the tissue damage, until you end up with widespread tissue destruction — particularly in places where cell turnover is low (nerves, retina, kidney). Conceivably, the accumulation of AGEs in the brain could result in cognitive decline. But what does this have to do with those plaques and tangles that are the sine qua non of the Alzheimer’s brain? We still don’t have a mechanistic connection between hyperglycemia and the AD pathology.
AGEs and Alzheimers
More recent investigations have revealed that there’s a little more to the Alzheimer’s pathology than we initially thought. As it turns out, when you look a little closer at the contents of both beta-amyloid plaques and neurofibrillary tangles, guess what you find?
Advanced Glycation End Products.
You also find receptors for AGEs (known as RAGEs) on the surface of diseased brain cells — receptors that are now known to play a role in oxidative damage. Strengthening the case even further is the fact that AGEs have been shown to both stimulate beta amyloid production and induce tau phosphorylation.
So, here we have several lines of converging evidence suggesting that AGEs are a significant — if not essential — component of Alzheimer’s pathogenesis. Plaques and tangles may well be the final outcome of a process first set in motion with AGEs. Such a story would provide us with the mechanistic connection we need between hyperglycemia and AD pathology. But, then, what about those plaques and tangles in non-diabetics? Does this AGE-incorporated model have any relevance for those with intact glycoregulation mechanisms?
Glycation without the Glucose
The following study was reported in the October 5th issue of the journal Neurology. It was probably read by a small handful of people, and certainly nobody in the popular media appreciated its potential significance. But it is the first prospective study I know of to investigate the effects of AGEs on cognitive function in non-diabetics.
The study involved 920 people without dementia, roughly half of whom were diabetic. At the start of the study, subjects were divided into three groups according to the levels of pentosidine in their urine. Pentosidine is a well established marker of AGEs — the more pentosidine in the urine, the more protein glycation that’s occurring in the body. Several measures of cognitive function were recorded at the beginning of the study and over the ensuing nine years.
At the start of the study, all groups — low, moderate, and high pentosidine — were equal in their cognitive scores, each averaging around 90 on a 100 point scale. At the nine year mark, however, things were no longer the same. The low pentosidine (surrogate for low AGEs) group had declined 2.5 points over that interval, the moderate group 5.4, and the high group 7.0. The p-value for these between group differences was <0.001. In sum, elevated AGE levels at the onset of the study, irrespective of diabetic status, predicted cognitive decline.
This study suggests two important things. One, that diabetics with only low level protein glycation (presumably from adequate pharmacologic control of blood glucose) don’t have any heightened risk of cognitive decline compared to non-diabetics — the path to cognitive decline (and presumably Alzheimer’s) in diabetics then appears to be through hyperglycemia and AGEs. Second, that non-diabetics with high levels of protein glycation have the same risk of cognitive decline as diabetics. In other words, if we’re trying to avoid cognitive decline and Alzheimer’s, it’s just as important for those without diabetes as it is for those with it to focus on minimizing AGEs.
Well Isn’t That Special???
So, how the heck does a person without diabetes avoid AGE accumulation in the brain? After all, if it wasn’t glucose causing all that glycation in those normoglycemic non-diabetics, then just what exactly was it?
Could it be perhaps, I don’t know…
(That’s right, I’m bringing the Church Lady back. Deal with it.)
As it turns out, there’s more than one way to glycate your bodily proteins. And if you’re not diabetic and thus adequately handling your dietary glucose load, then glucose isn’t likely a major culprit. Fructose, on the other hand, is about ten times more likely than glucose to form AGEs. And, unlike glucose, which can be metabolized in all cells of the body, fructose (like other toxins), can only be metabolized in the liver. The average American consuming an almost unfathomable 150-180 pounds of sugar (which is half glucose, half fructose) a year is easily overwhelming his or her liver’s capacity to metabolize fructose, which leaves it free to glycate like mad.
Not that we really needed another reason to avoid sugar, or, more specifically, fructose. Already implicated as a major contributor to the pathogenesis of obesity, fatty liver disease, insulin resistance, diabetes, and cancer, we’d have to be delusional to still cling to the notion that it’s just an “empty calorie”. But Alzheimer’s, too? Could it really be a primary instigator of this wretched, self-robbing disease of the mind?
Back when Dr. Alzheimer first presented the case of Auguste D in 1906, “presenile dementia” was a medical curiosity, scarce enough to warrant an anecdotal case report. Since that time, the average American’s sugar consumption has more than tripled. And 5.4 million of those Americans now suffer from Alzheimer’s disease.