Type 3 diabetes mellitus (T3DM) is a widely used term that reclassifies Alzheimer’s disease (AD) as a subtype of diabetes mellitus (DM) with pathogenic similarities to types 1 and 2 DM arising from a brain-specific form of insulin resistance. ^{Reference Meng, Zhang and Zhao1} The T3DM model theorizes that AD occurs when the brain’s neurones become unresponsive to insulin, leading to neurotoxic alterations in brain immunometabolism, including neuroinflammation, oxidative stress, β-amyloid (Aβ) toxicity and hyperphosphorylated tau toxicity. ^{Reference Meng, Zhang and Zhao1} In the T3DM theory, diabetic insulin resistance is the root cause of AD.

Analogous to the AD–DM overlap modelled by T3DM, commonalities in biological mechanisms and pathologies between seemingly distinct diseases are relatively common. Multiple other diseases are linked through shared aetiopathogeneses. For example, there is substantial mechanistic overlap between rheumatoid arthritis (RA) and atherosclerosis; indeed, RA is sometimes considered a “natural experiment” in the understanding of accelerated atherogenesis, since the two conditions share converging pathways at the molecular and cellular levels with both involving common pro-inflammatory cytokines. ^{Reference Libby2} The chronic systemic inflammation associated with RA amplifies atherosclerosis and associated cardiovascular comorbidities. Using the logic of T3DM, one could classify atherosclerosis as “arthritis of the arteries.”

The T3DM approach views AD through an endocrinological “diabetic lens.” There is, however, the possibility of looking through this lens via the opposite direction, from a neurological perspective; after all, protein misfolding and aggregation, pathological hallmarks of AD, are also features of diabetes that contribute to inadequate insulin production. Indeed, considerable research suggests extensive bidirectional overlap between diabetic and AD pathologies. A systematic review concerning common pathophysiological mechanisms reports that Aβ and tau protein are key early molecular players in inducing insulin resistance, further strengthening the groundwork for the concept of diabetes as pancreatic Alzheimer’s. ^{Reference Michailidis, Moraitou, Tata, Kalinderi, Papamitsou and Papaliagkas3}

A study by Miklossy et al., analysing pancreatic tissue from autopsy cases of people with noninsulin-dependent diabetes, reported that, similar to the AD brain, amyloid precursor protein (APP) and tau (precursors to fibrillary Aβ and hyperphosphorylated tau deposits) are present in pancreatic tissue, specifically β-cells. ^{Reference Miklossy, Qing and Radenovic4} The study likewise noted that immunoreactivity to Apo-E and Apo(a), proteins implicated in AD, was detectable in pancreatic tissue. Additionally, IB1/JIP-1 and JNK-1 proteins, involved in APP phosphorylation and co-localized with phosphorylated tau in AD tangles, were detected in pancreatic Langerhans islets. This indicates that Aβ deposition and tau hyperphosphorylation may be part of an early widespread host reaction, wherein amyloidosis triggers downstream pathology in non-neuronal as well as neuronal cells.

AD-type amyloid deposition in the pancreas occurs most frequently with amylin, or islet amyloid polypeptide. ^{Reference Miklossy, Qing and Radenovic4} Amylin is produced in the pancreas in concert with insulin, where it inhibits insulin and glucagon secretion. When demand for insulin rises, amylin synthesis correspondingly increases, leading to the accumulation of unprocessed and readily aggregated amylin forms that create intracellular and extracellular cytotoxic oligomers and fibrils. ^{Reference Ferreira, Raimundo, Menezes and Martins5}

The accumulation of misfolded amylin aggregates in Langerhans islets contributes to β-cell apoptosis with decline in β-cell function. In DM, the slowly progressive degeneration of islet β-cells inhibits insulin secretion and insulin action on peripheral tissues. ^{Reference Miklossy, Qing and Radenovic4} Mukherjee and Soto suggest that propagation of misfolded amylin – which may occur through transmission between β-cells or even between islets – contributes to the spread of islet lesions within the pancreas, augmenting disease progression. ^{Reference Mukherjee and Soto6}

T2 DM, alongside AD, Parkinson’s disease and other related neurodegenerative disorders, may thus be classified as a protein-misfolding disease (PMD) with amylin as the primary misfolded protein. ^{Reference Mukherjee and Soto6} Cytotoxic amylin oligomers killing pancreatic β-cells to yield the symptoms of diabetes mirrors neurotoxic Aβ oligomers killing cholinergic neurones to produce the symptoms of dementia. ^{Reference Mukherjee and Soto6} Moreover, cross-seeding and mixed pathology between PMDs is significant; just as Aβ and tau deposits are found in the pancreas, amylin can cross-seed with Aβ ^{Reference Ferreira, Raimundo, Menezes and Martins5} and tau ^{Reference Zhang, Meng and Wang7} in the brain, enhancing the insolubility and toxicity of these misfolded protein aggregates.

Lastly, recognition of commonalities between AD and diabetes has led to research on therapies that target both diseases, like intranasal insulin, which passes the blood-brain barrier to address brain insulin resistance. ^{Reference Meng, Zhang and Zhao1} In addition, Al Adem et al. propose the use of polyphenolic small molecules to inhibit Aβ₄₀ and amylin aggregation as well as their co-aggregation, thus potentially treating diabetes as an AD-like PMD. ^{Reference Al Adem, Shanti and Srivastava8}

In this tongue-in-cheek pushback against the term T3DM, we propose that, rather than conceptualizing AD in endocrinological language, we exploit neurological language to explain diabetes. Though words cannot perfectly describe the complexity of medical phenomena, there are sufficient data to support the notion that diabetes should be considered as pancreatic Alzheimer’s disease.