Circadian rhythms alter with ageing and may be aetiologically linked to neurodegeneration. This study explored the association between clinical markers and 1) dim light melatonin onset (DLMO) time and 2) phase angle derived from sleep midpoint, in older adults with varying dementia risks. Participants completed 14 days of actigraphy followed by in-lab measurement of salivary melatonin, from which DLMO time and phase angle were computed. Eighty participants (age = 65.5, SD = 9.6), 44 males (55%), MMSE (28.6, SD = 1.5) were included in the analysis. Sex (t = 2.15, p = .04), sleep onset (r = 0.49, p < .001) and midpoint (r = 0.44, p < .001) also correlated with DLMO time. Multiple linear regression showed chronotype, average actigraphy-derived light exposure during the DLMO window (window 2 h prior to DLMO to 2 h post), early biological day (6-10 h post DLMO time) and late biological day (10-14 h post DLMO time) were predictive of DLMO time (adjusted R2 = 0.75). Sleep offset, depression severity, average light exposure during the early biological night and early and late biological day were shown to be predictive variables in the estimation of phase angle (adjusted R2 = 0.78). The current study highlights the potential use of clinical variables, such as actigraphy-derived light, as circadian markers in ageing which could be easily implemented into existing clinical practice and could yield potential targets focusing on chronotherapeutic interventions.
Keywords: Circadian; actigraphy; dementia; dim light melatonin onset; light; sleep.