Author response: Sleep architecture and the risk of incident dementia in the community

We thank Dr. Kim for the letter regarding our article.1 In this important first step, we demonstrated an association between lower REM sleep percentage and an increased risk of dementia. It is now important for future studies to explore why REM sleep is related to dementia risk. We concur with Dr. Kim that it would be interesting to examine whether the suppression of REM sleep through exposure to antidepressants is associated with dementia incidence. However, as only 11 (3%) of our patients were taking antidepressants, such a study would require a much larger sample. It would also be challenging to overcome indication bias given that depression may also increase risk of dementia.2 Focusing on antidepressants with nonpsychiatric indications is one potential approach to help limit such bias. We agree with the sentiments of Dr. Kim in that the intriguing relationships between sleep and dementia require further study.