Hypoxia and Alzheimer disease

[One of the authors responds:]

My coauthors and I thank Sujoy Khan and Ioan Davies for their thoughtful response to our article.1 Before establishing a relationship between hypoxia and subsequent development of Alzheimer disease, one must consider 2 issues. The first is whether the cognitive changes observed during or after hypoxia are due to Alzheimer disease or other cognitive disorders. Recoverable cognitive dysfunction2 can persist for many months after hospital admissions and may simulate a dementia, although it is probably a type of subsyndromal delirium. The second is whether there are factors other than hypoxia (such as chronic inflammation3^,4) that may account for cognitive changes. In addition to hypoxemia, tobacco smoking causes prolonged exposure to carbon monoxide, numerous carcinogens and other potential neurotoxins. Head injury not only produces localized hypoxemia but it may also influence the subsequent development of Alzheimer disease through the effects of hemorrhage, contusion and inflammatory responses. Both hypertension and tobacco smoking increase the risk of stroke, and the synergistic effect of Alzheimer disease and cerebrovascular disease is well known.5 In chronic obstructive lung disease, inflammatory cytokines,6 recurrent infections and conceivably the effects of anticholinergic medications could all contribute to the observed cognitive changes. Heart failure may also be associated with cognitive deficits, but many factors other than hypoxia, such as activation of neuroendocrine pathways, rheologic changes and consumption of numerous medications with known anticholinergic side effects, offer alternative explanations.7 Finally, the existence of anatomic changes pathognemonic of Alzheimer disease does not guarantee a phenotype of dementia.5

Khan and Davies raise an intriguing hypothesis, but our review cannot answer their question as we were limited by the methodologic constraints of longitudinal cohort studies, none of which consistently measured arterial oxygen tension or transcutaneous oxygen saturation levels. Only prospective studies will be able to answer their question.