Reply to Kawada: Obstructive Sleep Apnea and Cognitive Decline in Older Adults

From the Authors:

After the publication of our review (1), Dr. Kawada raised important comments in a letter to the editor, highlighting that serum insulin and insulin activity in the brain could play a role in the relationship between obstructive sleep apnea (OSA) and cognitive decline. OSA has been associated with a higher risk of insulin resistance (IR) and type II diabetes mellitus (T2DM) (2) in studies that included mostly middle-aged subjects. IR occurs when tissues do not respond effectively to insulin, and thus higher levels are required to keep blood glucose levels normal. When an individual cannot keep up with this increased endogenous insulin production/secretion, glucose levels start to rise and diabetes ensues. Many potential mechanisms have been proposed to explain this relationship between OSA and IR/T2DM. Intermittent hypoxemia and sleep fragmentation caused by OSA lead to inflammation, oxidative stress, and increased sympathetic nervous system activity (2), all of which have the potential to increase IR and lead to T2DM (2). OSA may play a causal role in the pathophysiology of IR/T2DM, as insulin sensitivity was reported to be improved by a treatment for OSA (continuous positive airway pressure therapy) (2), although not all studies have shown that effect. In middle-aged individuals compared with the elderly, OSA is more strongly associated with metabolic syndrome, a condition characterized by central obesity, hypertension, dyslipidemia, and impairments in glucose control, including IR (3). Therefore, it is important to consider the effects of OSA in middle age versus late life on metabolic, vascular, and cognitive outcomes.

Altered cerebral insulin activity can be observed when there is IR in the brain, or when insulin transporters at the blood–brain barrier are downregulated as a result of the compensatory hyperinsulinemia that follows peripheral IR (4). Altered insulin activity has been hypothesized to affect cognitive health. In addition to its role in glucose metabolism, insulin is involved in a signal transduction cascade that ultimately affects synaptic plasticity (4). In fact, insulin is involved in long-term potentiation by modulating the expression of N-methyl-d-aspartate receptors. Moreover, insulin modulates neurotransmitter and nitric oxide levels, and thereby may play a role in vascular and cognitive functions (4). IR has been associated with cognitive deficits in both human and animal studies (4). Interestingly, in a recent meta-analysis, Cao and colleagues investigated the effects of several antidiabetic agents and concluded that these agents were associated with improved cognitive performance in patients with mild cognitive impairment or Alzheimer’s disease (AD) in randomized clinical trials (5). In fact, IR/T2DM is associated with cognitive impairment, cognitive decline, and increased risk of dementia (4). However, the connection between IR and AD does not appear to involve classic AD pathology, as humans with T2DM do not show increased plaques and tangles postmortem (6). An alternative mechanism proposed for the added risk of dementia with impaired insulin function is based on the metabolic and cerebrovascular abnormalities shared by the two conditions.

In summary, although OSA is a risk factor for IR/T2DM in middle-aged individuals, which in turn could increase the risks of cognitive decline and dementia, there is currently not enough evidence to conclude that IR/T2DM might be one of the mechanisms that explain these associations. That said, OSA and IR/T2DM present similar pathological mechanisms, and thus the concomitant presence of these conditions may lead to an additive insult to the brain. However, given the interaction between OSA and IR, insulin regulation may be an interesting avenue to explore when investigating how OSA increases the risk of dementia and cognitive decline. Because OSA and IR/T2DM are treatable, clarifying these associations has the potential to lead to possible interventions to reduce or even prevent neurodegeneration; however, the current state of the literature lacks the appropriate evidence.