Back to Top
IPA Bulletin: Featured Article

Diabetes and Alzheimer’s disease: Is there a link? (Mary Sano) June 2014

Diabetes and Alzheimer's disease: Is there a link?

Mary Sano, Director and ADRC Professor, Department of Psychiatry, Mount Sinai School of Medicine, New York, United States

Many studies have reported that Diabetes, particularly Type 2 Diabetes (T2D) is a risk factor for Alzheimer’s disease (AD). Such an idea is intriguing because it may offer new approaches to the treatment of cognitive loss and dementia. Further, both of these conditions are global epidemics and their consequences are likely to grow as lifestyles in industrialized societies tends to foster T2D and technological demand requires highly maintained cognition.

As early as 1996 the connection between T2D and dementia was reported from the Rotterdam study of elders in the Netherlands (Ott et al.). More recent findings support this link with evidence that the risk of incident dementia increases with glucose values over the previous five to eight years. These results came from more than 2,000 subjects over the age of 65 who were members of the Group Health Cooperative, a health care system in Washington state (United States). Of note, this relationship was apparent in both diabetics and non-diabetics (Crane et al., 2013). These early studies depended on a clinical evaluation of dementia and often proposed a link to Alzheimer's disease, the most common form of dementia.

Today, with the advent of amyloid imaging and a large number of neuropathological studies, we can determine if the association is with AD pathology or perhaps is due to other types of dementia.

In fact, studies that measure the defining pathology of Alzheimer's disease tell a different story. A 2005 report examined specimens from a brain bank of individuals who were well characterized as having a clinical diagnosis of AD and the presence of T2D found an unexpected result. Cases with diabetes and AD pathology had greater dementia severity as measured by the Clinical Dementia Rating (CDR) than those with the same amount of amyloid pathology and no T2D. In other words, the presence of T2D required less AD pathology to express the same amount of clinical impairment (Beeri et al., 2005).

This suggests that while T2D might worsen the clinical and cognitive picture it did not do it through increasing AD pathology. Several other studies support this notion. For example, the Baltimore Longitudinal Study, which has followed aging individuals for several decades has examined both anti-mortem, in-vivo brain amyloid burden measured with carbon 11–labeled Pittsburgh Compound B (11C-PiB), and post mortem neuropathology data. A report from this study found no significant correlations between measures of brain AD pathology or 11C-PiB β-amyloid load and glucose intolerance or insulin resistance in this population (Thambisetty et al.).

With these results in mind a closer look at the clinical picture of the aging individual with T2D would seem in order. In one study of non-demented individuals who had received an extensive cognitive assessment, T2D individuals did more poorly on a global cognitive composite score (Nandipati et al., 2012). However, the impairment was found on tests of attention, executive function and psychomotor speed but not on measures of memory.

Another study reported on subjects from a diabetes clinic, and found that half of the diabetic cohort had undiagnosed cognitive impairment with 10% having frank dementia and 30% having a diagnosis of some type of MCI. When compared to a demographically matched non-diabetic cohort, those with diabetes did more poorly on the Trail Making and Digit Symbol tests, measures of attention and executive function, while no difference was found on paragraph recall. When the diabetic cohort was compared to a demographically similar group who had amnestic MCI of the Alzheimer's type, the diabetic cohort performed significantly better on memory tests.

A report that examined data from the Alzheimer’s Disease Neuroimaging (ADNI) study demonstrated that impaired glycemic control among individuals with MCI was associated with greater progression of clinical and cognitive measures than seen in those with normal glycemic control (Morris et al., 2014). ADNI measures a range of biomarkers and this report described significantly greater whole brain volume loss over time than in the normal glycemic group but no difference in progression of hippocampal volume loss.

Other studies have reported that T2D is associated with poorer uptake of FDG PET, a measure of brain activation (Roberts et al., 2014). A study from Malaga, Spain, examined the relationship between glucose metabolism and imaging in those with T2DM (García-Casares et al., 2014). In this cohort, reduced gray matter densities and reduced cerebral glucose metabolism was reported in several fronto-temporal brain regions after controlling for various vascular risk factors. Furthermore, within the T2DM group, longer disease duration, and higher HbA1c levels and HOMA-IR were associated with lower gray matter density and reduced cerebral glucose metabolism in fronto-temporal regions.

Several studies have examined treatment of T2D for its benefit in cognitive impairment and dementia. Some but not all observational studies have suggested that insulin may reduce amyloid burden and more work is underway to address this focusing on those with AD. Randomized trials of other anti-diabetic agents have not shown systematic benefits.

These studies indicate the need for further research to characterize the associated biology of cognitive and clinical impairment seen in T2D and aging to determine how it might be mitigated to improve the clinical picture.

In summary, T2D and AD are both common diseases that increase with aging. These entities have been reported to co-occur in some but not all studies but there are limitations to drawing conclusions. For example, while the overlap is commonly reported when AD is assessed clinically and when T2D is self-reported, when the profile of cognitive impairment is examined, T2D tends to be characterized by non-amnestic deficit. Further, pathological studies report equal or less amyloid pathology. The role of vascular pathology and inflammatory markers are not well-assessed and may assist in our understanding of the clinical overlap.

It is important to remember that T2D remains highly associated with significant cognitive impairment, probably progressive in nature. The deficits are in cognitive areas such as executive function and attention which are associated with functional disability. There is little evidence that therapeutic control of diabetes improves cognition. Because T2D is one of the most complicated diseases to manage, the cognitive impairment in aging that is associated with it is a major public health problem.


  1. Beeri, M.S., Silverman, J.M., Davis, K.L., et al. Type 2 diabetes is negatively associated with Alzheimer's disease neuropathology. Journals of Gerontology Series A: Biological Sciences and Medical Sciences. 2005; 60: 471-5.
  2. Crane, P.K., Walker, R., Hubbard, R.A., et al. Glucose Levels and Risk of Dementia. New England Journal of Medicine. 2013; 369: 540-8.
  3. García-Casares, N., Berthier, M.L., Jorge, R.E. et al. Structural and functional brain changes in middle-aged type 2 diabetic patients: a cross-sectional study. Journal of Alzheimer’s Disease. 2014; 40: 375-86.
  4. Morris, J.K., Vidoni, E.D., Honea, R.A., Burns, J.M. Impaired glycemia increases disease progression in mild cognitive impairment. Neurobiology of Aging. 2014; 35: 585-589.
  5. Nandipati, S., Luo, X., Schimming, C., Grossman, H.T., Sano, M. Cognition in Non-Demented Diabetic Older. Current Aging Science. 2012; 5(2): 131–135.
  6. Ott, A., Stolk, R.P., Hofman, A., van Harskamp, F., Grobbee, D.E., Breteler, M.M. Association of diabetes mellitus and dementia: the Rotterdam Study. Diabetologia. 1996; 39: 1392-7.
  7. Roberts, R.O., Knopman, D.S., Geda, Y.E., et al. Association of diabetes with amnestic and nonamnestic mild cognitive impairment. Alzheimer’s Dementia. 2014; 10: 18-26.
  8. Thambisetty, M., Metter, E.J., Yang, A, et al. Glucose Intolerance, Insulin Resistance, and Pathological Features of Alzheimer Disease in the Baltimore Longitudinal Study of Aging. JAMA Neurology. 2013; 70(9): 1167-1172.


Back to main page


VCambridge University Press