Dementia is an increasingly prevalent condition that intersects worldwide with the epidemic of type 2 diabetes mellitus (DM2). It would seem logical to expect that the occurrence of DM2 increases the likelihood of developing dementia, due to its deleterious effect on the cerebral vasculature and the associated hormonal and metabolic changes. Many reports indicate that it also increases the risk of developing Alzheimer's disease (AD). However, other studies suggest that diabetes might have a relatively strong protective effect against AD, whereas genetically engineered animal models of the condition deteriorate more severely when there is a concomitant insulin resistant brain state (IRBS). Furthermore, IRBS alone is associated with anatomical, behavioral, and molecular changes that justify the proposal that AD may be due to an IRBS. This is explored in the context of accumulating evidence that the IRBS need not be related to peripheral insulin resistance, and that administration of insulin directly to the brain improves selected cognitive parameters targeted in AD. This view is consistent with the Damage Signals hypothesis of AD pathogenesis, which can help unifying the pleiotropic effects of agents toxic to insulin-producing/secreting (eg, pancreatic β) cells, as well as IRBS caused by different mechanisms in AD. Such approach may help tackling the Innovation Gap, which results from a host of factors slowing down progress towards innovative palliation and prevention of AD, as well as dementia due to complications of diabetes distinct from AD, and both conditions combined with their commonly associated metabolic and hormonal alterations.