Exercise for health

Why blood sugar control matters even when you are not diabetic

Chronic glucose dysregulation causes measurable damage before a diabetes diagnosis arrives — including to the brain. The evidence on euglycemia, insulin resistance, and dementia risk is worth understanding well before problems develop.

6 min read · by · educational content, not medical advice

What euglycemia means and why the spectrum matters

  • Euglycemia refers to normal blood glucose regulation: fasting glucose between 70–99 mg/dL (3.9–5.5 mmol/L), an HbA1c below 5.7%, and postprandial glucose returning to baseline within two hours of eating.
  • Above these ranges, the classifications are: impaired fasting glucose (100–125 mg/dL), prediabetes (HbA1c 5.7–6.4%), and type 2 diabetes (fasting glucose ≥126 mg/dL or HbA1c ≥6.5%). But these thresholds are administrative cutoffs designed for clinical diagnosis — not biological inflection points where harm begins.
  • The damage from elevated blood glucose occurs on a continuum. Advanced glycation end products (AGEs) — formed when glucose molecules bind irreversibly to proteins and lipids — accumulate at levels well below the diabetic threshold, causing structural damage to blood vessels, nerve tissue, kidneys, and the lens of the eye.
  • A 2013 New England Journal of Medicine study (Crane et al.) found that among adults without diabetes, each incremental increase in average glucose — even within the normal and high-normal range — was independently associated with increased dementia risk. There was no safe floor in the non-diabetic range: lower was better, across the board.

Chronic glucose dysregulation damages the body before symptoms appear

  • Postprandial glucose spikes — the sharp rise in blood sugar after eating — generate brief but repeated episodes of oxidative stress and endothelial inflammation. Sustained over years, this cumulative injury to blood vessel walls is a primary driver of cardiovascular disease, independent of cholesterol.
  • Advanced glycation end products cross-link structural proteins throughout the body, stiffening arterial walls, impairing kidney filtration membranes, and degrading nerve myelin. AGE accumulation begins in the prediabetic and even high-normal glucose range, long before a formal diagnosis.
  • Insulin resistance — the condition in which cells respond less efficiently to insulin signals — often precedes a diabetes diagnosis by 10–15 years. During this silent period, compensatory hyperinsulinemia (the pancreas producing excess insulin to overcome resistance) drives inflammation, promotes fat storage, and begins impairing the insulin-dependent functions of non-metabolic organs, including the brain.
  • HbA1c in the upper range of 'normal' (5.5–5.6%) has been associated in large cohort studies with meaningfully elevated cardiovascular and cognitive risk compared to HbA1c in the lower-normal range (below 5.0%). The implication is that optimizing within the normal range matters.

Type 3 diabetes: insulin resistance in the brain

  • In 2005 and 2008, neuropathologist Suzanne de la Monte and colleagues proposed the term 'Type 3 diabetes' to describe a pattern they identified in Alzheimer's disease brains: severely impaired insulin signaling and insulin receptor responsiveness in brain tissue, distinct from and overlapping with — but not identical to — systemic type 2 diabetes.
  • The brain is an insulin-dependent organ, but it uses insulin differently than peripheral tissues. In the hippocampus and prefrontal cortex — regions critical to memory and executive function — insulin signaling supports the production of brain-derived neurotrophic factor (BDNF), regulates tau protein phosphorylation, and facilitates the clearance of amyloid-beta peptides, the primary component of the plaques found in Alzheimer's brains.
  • When brain insulin signaling is impaired, BDNF levels fall (impairing synaptic plasticity and new memory formation), tau becomes hyperphosphorylated (forming the neurofibrillary tangles characteristic of Alzheimer's pathology), and amyloid-beta clearance slows. These are not incidental findings — they are the core molecular events in Alzheimer's disease progression.
  • The 'Type 3 diabetes' designation remains an active area of research rather than an established consensus diagnosis. What is established is that the mechanistic overlap between systemic insulin resistance and Alzheimer's pathology is substantial — and that epidemiological evidence consistently shows individuals with type 2 diabetes have approximately 1.5–2× the dementia risk of euglycemic adults.

What the epidemiological evidence shows

  • The Crane et al. 2013 NEJM study followed 2,067 older adults without diabetes for up to 7 years. Among those who developed dementia, average glucose levels in the years prior to diagnosis were significantly higher than in those who did not — with risk increasing continuously across the non-diabetic range. This study was significant because it demonstrated risk at levels well below any clinical threshold.
  • The Rotterdam Study found that insulin resistance measured by HOMA-IR — a calculation using fasting glucose and fasting insulin — was associated with dementia risk in non-diabetic adults independent of cardiovascular risk factors. Insulin resistance, not just high glucose, appears to be the operative risk signal.
  • Midlife glucose regulation appears to matter more than late-life levels for dementia risk. Several large longitudinal studies suggest that the critical window for intervention is in the 40s and 50s — before structural brain changes become irreversible. Late-life glucose control may slow progression but does less to prevent onset.
  • A 2019 analysis of UK Biobank data found that even prediabetes — not full diabetes — was associated with increased risk of dementia, depression, and cognitive decline, reinforcing that the relevant risk is distributed across the entire glucose spectrum.

The modifiable levers for maintaining euglycemia

  • Resistance training is the highest-leverage single intervention for improving insulin sensitivity. It increases GLUT4 transporter density in muscle cells, improves glucose disposal acutely for 24–72 hours post-exercise, and builds lean mass — the primary tissue responsible for blood glucose clearance at rest. The chronic adaptation from consistent resistance training directly addresses the insulin resistance pathway.
  • Aerobic exercise independently improves insulin sensitivity through different mechanisms — primarily mitochondrial biogenesis and improved oxidative metabolism in muscle and liver. The combination of resistance training and aerobic activity produces better glucose outcomes than either alone.
  • Sleep is an underappreciated driver of glucose regulation. A single night of partial sleep deprivation (4–5 hours) can reduce insulin sensitivity by 20–25% in healthy adults. Chronic poor sleep is one of the most reliable drivers of insulin resistance in otherwise healthy populations.
  • Meal composition affects postprandial glucose amplitude: lower glycaemic load, higher fibre, and adequate protein all blunt the postprandial glucose spike for a given caloric intake. These are nutritional complements to the exercise-based approach, not replacements for it.
  • The practical implication is that euglycemia is not primarily a pharmaceutical problem — it is a lifestyle infrastructure problem. The interventions that protect long-term blood sugar regulation are largely the same ones that improve body composition, cardiovascular health, and physical function: structured exercise, sufficient lean mass, and adequate sleep.