In response to the Letter to the Editor regarding “Earlier chronotype in midlife as a predictor of accelerated brain aging: a population-based longitudinal cohort study.”: Earlier chronotype in midlife and brain aging

Dear Editor,
We thank Dr. Smagula for the commentary with insightful perspectives [1], which provides a precise assessment, including both the strengths and the limitations of our study [2]. We appreciate the recognition of the need for further studies on changes in sleep-wake rhythm/phase during the aging process besides the quality/quantity of sleep and its impact on neurodegenerative disorder risk in older adults. Dr. Smagula commented on the measures of predictive utility for detecting an outcome likely to be clinically meaningful. We agree that we could have more clearly demonstrated the value of chronotype as a prediction/screening tool for neurodegenerative disorder by including a prediction model for detecting outcomes such as mild cognitive impairment (MCI) or dementia. However, we think our investigation was a sufficient basis for determining whether the chronotype at around 60 years old, a relatively stabilized age for sleep phase advancement in the community-based cohort without definite cognitive impairment, can gauge the direction of change in cognitive function and brain structure in the future. Moreover, since we excluded data from participants who scored less than 24 on the Mini-Mental State Examination screening at baseline, few participants were diagnosed with MCI or dementia. Our cohort introduced a more comprehensive standard neuropsychological test battery under the supervision of neurologists to diagnose MCI or dementia after the third follow-up investigation (2019-2022). Therefore, we will be able to figure out soon which sleep behavior changes from middle-aged are the most potent factors in predicting/screening neurodegenerative disorders through over 20 years of follow-up as Dr. Smagula’s suggestions.
The second comment concerns the lack of conceptual specificity in the exposure measure and consideration for related factors. Our study started from the preliminary analysis of our cohort Page 3 of 9 Sleep (N=1800, supplementary tables), which confirmed that sleep phase-related factors had more significant associations with cognitive function and gray matter volume (GMV) rather than sleep quantity-related characteristics. In addition, MSFsc (mid-sleep time on free days corrected for sleep debt on workdays) is a measurement developed to estimate chronotype in consideration of the time of bedtime and wake-up time on weekdays and weekends, which vary depending on social activities and is known to be highly relevant to the conventional circadian process markers such as dim-light melatonin onset or core body temperature [3]. Furthermore, as we previously have confirmed that our MSFsc metric derived from the modified Pittsburgh Sleep Quality Index (PSQI) has relatively little dependence on other subcomponents of PSQI or the total index itself (r < 0.1) [4], we proceeded with the analysis after confirming MSFsc as an independent exposure variable according to the purpose of the study. As most studies on cognitive function concerning sleep of the elderly so far have mainly focused on sleep homeostasis by investigating sleep duration or slow-wave activity, our study definitely added evidence to an emerging picture that the sleep phase is related to markers of neurodegeneration. As we need a more detailed understanding of the respective impact of circadian rhythm and homeostatic deteriorations with age that regulate neurobehavioral function, we should further investigate the association between various sleep characteristics, including chronotype, and the structural/functional brain outcome [4,5].