Insulin resistance · Dementia screening

Insulin Resistance and Dementia: Screening Past the Glucose

Screening for insulin resistance in patients at dementia risk has to go past fasting glucose and A1c. In Whitehall II, insulin sensitivity declined for about five years before a diabetes diagnosis, while fasting glucose accelerated only in the last three.

Glucose-based screening finds the metabolic failure after years of compensation. For the brain, those compensating years are the ones that matter.

The link between insulin resistance and dementia is usually taught as a diabetes statistic, and screening follows the teaching: order a fasting glucose or an A1c and act when it crosses a threshold. The evidence assembled for Chapter 3 of The Starved Brain by Dr. Gurpreet Singh Padda, MD, MBA, MHP, and its accompanying video, argues that this sequence catches the problem at the wrong end. For a practice using Measura [Cardiometabolic and Autonomic Health Analysis], the clinical question is narrower and more useful: which measurements describe the compensating years, and what does a finding change?

How much does insulin resistance raise dementia risk?

The pooled figures are familiar. Across 2,310,330 people and 102,174 dementia cases, type 2 diabetes carried a relative risk of 1.62 in women and 1.58 in men, with a larger estimate for vascular dementia. Whenever a pool is divided by subtype, the vascular number is the bigger one. Measured insulin resistance, pooled across ten prospective cohorts, gives a hazard ratio of 1.11 for all-cause dementia and 1.23 for Alzheimer’s disease: smaller than the diabetes figures, which suggests much of the risk is carried downstream by glycemia, vessels and years of exposure.

The null side deserves equal reading. In the Baltimore Longitudinal Study of Aging, 197 autopsied participants showed no correlation between glucose intolerance or insulin resistance and plaque or tangle scores. Mendelian randomization in 17,008 Alzheimer’s cases found no causal signal from genetically predicted insulin resistance. Type 3 diabetes remains a hypothesis with substantial support, not an established mechanism. Two caveats temper those nulls: peripheral markers in cohorts with little metabolic disease contain little exposure, and a genetic lifetime nudge is not a decade of dietary hyperinsulinemia.

Why is fasting glucose a late screen for insulin resistance?

Whitehall II gives the cleanest timeline. Among the civil servants who progressed to diabetes, insulin sensitivity declined steeply for about five years before diagnosis, beta-cell output climbed between years four and three as compensation, and fasting glucose accelerated only in the last three years. The chart records the failure of compensation, not its onset.

Most of us were trained to reassure the patient whose A1c sat just under the diabetes line. Dr. Padda describes doing exactly that for years before the physiology changed his practice. The error was not carelessness; it was screening with an instrument designed to detect the end state.

The brain data sit on the same clock. In 150 cognitively normal adults with a mean age of 60.7, higher HOMA-IR predicted lower whole-brain glucose uptake and, in the medial temporal lobe, poorer memory. In the Framingham Offspring cohort, with diabetes excluded, fasting insulin was associated with smaller total brain volume where A1c showed no association. In 683 older Manhattan adults, hyperinsulinemia carried a hazard ratio of 2.1 for Alzheimer’s disease, highest among those without diabetes. All observational, and all pointing at insulin rather than glucose as the earlier signal.

What does an insulin resistance finding change in management?

A glucose-only workup leaves a binary: diabetic or not. A broader measurement set produces a gradient, and the gradient changes three decisions.

1. Who gets a cognitive baseline. The insulin-resistant patient with normal cognition is precisely the phenotype in which brain glucose hypometabolism has been shown before symptoms. A documented cognitive assessment converts a future complaint into a comparison, and the annual wellness visit is the natural place to record it.
2. How the vascular route is worked up. The diabetes and dementia association is strongest for vascular dementia, and across 2,365 autopsied brains diabetes tracked infarcts and lacunes rather than plaques or Braak stage. Arterial stiffness and endothelial function belongs in the same encounter as the metabolic workup.
3. What the metabolic conversation is anchored to. Laboratory panels that go past a fasting glucose, read alongside bioimpedance body composition, show the patient the terrain rather than a pass mark.

The obstacle is rarely knowledge. It is a visit built around the numbers the quality dashboard already tracks, a food environment that supplies carbohydrate continuously, and a referral structure that routes one upstream failure to endocrinology, hepatology, gynecology and neurology under four separate names.

Workflow: selection, orders and documentation

Selection does not have to be elaborate. The evidence names its own triggers: prediabetes or a family history of diabetes, central adiposity or midlife obesity, fatty liver, polycystic ovary syndrome, and a family history of Alzheimer’s disease, since the cohorts that showed early hypometabolism were enriched for it. Normal glucose is not an exclusion; it is the population in which the compensating phase hides.

Written as a standing order, the protocol runs the same way regardless of which clinician sees the patient, with the criteria set out under selection criteria. Cognitive and cardiometabolic findings documented in one encounter also support MIPS and quality reporting without a parallel workflow.

What does an insulin resistance finding not prove?

Three limits belong in the chart note as much as in the literature. PET hypometabolism is a research measurement; Measura does not perform imaging, and no peripheral marker is a validated stand-in for regional brain glucose uptake. The cause-versus-consequence question is open, since amyloid oligomers can induce neuronal insulin resistance as well as follow it. And the cleanest direct test of the brain insulin hypothesis, intranasal insulin in the SNIFF trial of 289 participants, returned a null primary result. None of that weakens the case for measuring insulin resistance early. It defines what a finding means: a terrain risk to manage and a baseline to follow, not a diagnosis of neurodegeneration. The full grading, trial by trial, is in the Chapter 3 book companion, and the structural counterpart, cholesterol and cognition on statins, is in statins and dementia. Where fat and muscle sit, not just body weight, tracks neurodegenerative risk; see body fat distribution and dementia risk.

Frequently asked questions

Is type 3 diabetes a recognized diagnosis?

No. It is a research term proposed in 2005 for disrupted insulin and IGF signaling in Alzheimer’s brain tissue, with features of both insulin resistance and insulin deficiency. It has no diagnostic criteria, and whether it is cause or consequence remains open. The measurable clinical correlate is systemic insulin resistance, which can be documented. The broader case is in clinical rationale.

Why not rely on A1c to screen for dementia risk?

Because A1c moves late. Whitehall II showed insulin sensitivity falling for about five years before glucose crossed the diagnostic threshold, and in Framingham Offspring fasting insulin, not A1c, was associated with brain volume. A1c stays necessary; it is not sufficient to describe the compensating phase. The pre-diabetic picture is outlined in insulin resistance before diabetes.

How does cognitive assessment pair with fall prevention here?

The same older, metabolically compromised patients carry both cognitive and fall risk, and both are managed by the same clinician. Recording a cognitive baseline and fall-risk findings in one encounter avoids two separate workups and gives follow-up a single plan. The pairing is described in cognitive assessment and fall prevention.

What changes if a patient with normal glucose shows insulin resistance?

The patient moves from reassured to monitored. A cognitive baseline is documented, vascular measurement is added, and follow-up intervals are set by the finding rather than by the next glucose result. Treatment decisions remain with the treating physician, informed by a clearer picture of the terrain. Reading the findings is covered in interpreting the report.

Is there a patient-facing explanation of this?

Yes. The patient version explains, in plain language, why a normal glucose result may not be reassuring, what the terrain measurements show, and which questions to bring to a visit. It works as a handout before the conversation about testing. It is available as type 3 diabetes for patients.

Does insulin resistance raise the risk of dementia?

Yes. Pooled across ten prospective cohorts, measured insulin resistance carried a hazard ratio of 1.11 for all-cause dementia and 1.23 for Alzheimer’s disease. Type 2 diabetes carried relative risks of 1.62 in women and 1.58 in men. In 683 older Manhattan adults, hyperinsulinemia carried a hazard ratio of 2.1 for Alzheimer’s disease, highest among those without diabetes, which points at insulin rather than glucose as the earlier signal.

Who should be screened for insulin resistance?

The evidence names its own triggers: prediabetes or a family history of diabetes, central adiposity or midlife obesity, fatty liver, polycystic ovary syndrome, and a family history of Alzheimer’s disease. Normal glucose is not an exclusion. It is the population in which the compensating phase hides, and the insulin-resistant patient with normal cognition is the phenotype in which brain glucose hypometabolism has been shown before symptoms.

Which type of dementia is most linked to diabetes?

Vascular dementia. Whenever pooled data are divided by subtype, the vascular number is the bigger one, and across 2,365 autopsied brains diabetes tracked infarcts and lacunes rather than plaques or Braak stage. That is why arterial stiffness and endothelial function belong in the same encounter as the metabolic workup rather than in a separate referral.

References

• Chatterjee, S., Peters, S. A. E., Woodward, M., Mejia Arango, S., Batty, G. D., Beckett, N., Beiser, A., Borenstein, A. R., Crane, P. K., Haan, M., Hassing, L. B., Hayden, K. M., Kiyohara, Y., Larson, E. B., Li, C.-Y., Ninomiya, T., Ohara, T., Peters, R., Russ, T. C., … Huxley, R. R. (2016). Type 2 diabetes as a risk factor for dementia in women compared with men: A pooled analysis of 2.3 million people comprising more than 100,000 cases of dementia. Diabetes Care, 39(2), 300–307. https://doi.org/10.2337/dc15-1588
• Xu, X., Li, J., Pan, H., Zhao, B., & Wang, D. (2026). Insulin resistance and the risk of dementia: A systematic review and meta-analysis of prospective cohort studies. BMC Neurology, Article in press (DOI assigned); PMID 42387430. https://doi.org/10.1186/s12883-026-05079-x
• Thambisetty, M., Jeffrey Metter, E., Yang, A., Dolan, H., Marano, C., Zonderman, A. B., Troncoso, J. C., Zhou, Y., Wong, D. F., Ferrucci, L., Egan, J., Resnick, S. M., & O’Brien, R. J. (2013). Glucose intolerance, insulin resistance, and pathological features of Alzheimer disease in the Baltimore Longitudinal Study of Aging. JAMA Neurology, 70(9), 1167–1172. https://doi.org/10.1001/jamaneurol.2013.284
• Østergaard, S. D., Mukherjee, S., Sharp, S. J., Proitsi, P., Lotta, L. A., Day, F., Perry, J. R. B., Boehme, K. L., Walter, S., Kauwe, J. S., Gibbons, L. E., … Scott, R. A. (2015). Associations between potentially modifiable risk factors and Alzheimer disease: A Mendelian randomization study. PLoS Medicine, 12(6), e1001841. https://doi.org/10.1371/journal.pmed.1001841
• Tabák, A. G., Jokela, M., Akbaraly, T. N., Brunner, E. J., Kivimäki, M., & Witte, D. R. (2009). Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: An analysis from the Whitehall II study. The Lancet, 373(9682), 2215–2221. https://doi.org/10.1016/S0140-6736(09)60619-X
• Willette, A. A., Bendlin, B. B., Starks, E. J., Birdsill, A. C., Johnson, S. C., Christian, B. T., Okonkwo, O. C., La Rue, A., Hermann, B. P., Koscik, R. L., Jonaitis, E. M., Sager, M. A., & Asthana, S. (2015). Association of insulin resistance with cerebral glucose uptake in late middle-aged adults at risk for Alzheimer disease. JAMA Neurology, 72(9), 1013–1020. https://doi.org/10.1001/jamaneurol.2015.0613
• Tan, Z. S., Beiser, A. S., Fox, C. S., Au, R., Himali, J. J., Debette, S., DeCarli, C., Vasan, R. S., Wolf, P. A., & Seshadri, S. (2011). Association of metabolic dysregulation with volumetric brain magnetic resonance imaging and cognitive markers of subclinical brain aging in middle-aged adults: The Framingham Offspring Study. Diabetes Care, 34(8), 1766–1770. https://doi.org/10.2337/dc11-0308
• Luchsinger, J. A., Tang, M.-X., Shea, S., & Mayeux, R. (2004). Hyperinsulinemia and risk of Alzheimer disease. Neurology, 63(7), 1187–1192. https://doi.org/10.1212/01.wnl.0000140292.04932.87
• Abner, E. L., Nelson, P. T., Kryscio, R. J., Schmitt, F. A., Fardo, D. W., Woltjer, R. L., Cairns, N. J., Yu, L., Dodge, H. H., Xiong, C., Masaki, K., Tyas, S. L., Bennett, D. A., Schneider, J. A., & Arvanitakis, Z. (2016). Diabetes is associated with cerebrovascular but not Alzheimer’s disease neuropathology. Alzheimer’s & Dementia, 12(8), 882–889. https://doi.org/10.1016/j.jalz.2015.12.006
• Craft, S., Raman, R., Chow, T. W., Rafii, M. S., Sun, C.-K., Rissman, R. A., Donohue, M. C., Brewer, J. B., Jenkins, C., Harless, K., Gessert, D., & Aisen, P. S. (2020). Safety, efficacy, and feasibility of intranasal insulin for the treatment of mild cognitive impairment and Alzheimer disease dementia: A randomized clinical trial. JAMA Neurology, 77(9), 1099–1109. https://doi.org/10.1001/jamaneurol.2020.1840

Related reading

• Fasting Insulin Screening: The Gate on the Brain’s Second Fuel
• Statins and Dementia: Measuring Cognition in Patients on Statins
• Cognitive Assessment

Medically reviewed by Dr. Gurpreet Singh Padda, MD, MBA, MHP, medical director of Measura. Last reviewed September 12, 2026.