Table 3 Circadian rest-activity metrics and change in frailty with adjustment for covariates

	Estimate (SE) p value	Estimate (SE) p value	Estimate (SE) p value	Estimate (SE) p value	Estimate (SE) p value	Estimate (SE) p value	Estimate (SE) p value
Intercept	−0.118 (0.027) < 0.001	−0.125 (0.027) < 0.001	−0.123 (0.027) < 0.001	−0.122 (0.027) < 0.001	−0.121 (0.027) < 0.001	−0.114 (0.027) < 0.001	−0.127 (0.027) < 0.001
Time	0.089 (0.004) < 0.001	0.088 (0.004) < 0.001	0.088 (0.004) < 0.001	0.088 (0.004) < 0.001	0.088 (0.004) < 0.001	0.090 (0.004) < 0.001	0.088 (0.004) < 0.001
Amplitude^a	−0.050 (0.016) 0.002
Amplitude × time^b	−0.005 (0.002) 0.030
Acrophase^a		0.039 (0.015) 0.010
Acrophase × time^b		−0.005 (0.002) 0.022
Variation of cycle length^a			0.039 (0.015) 0.010
Variation of cycle length × time^b			−0.002 (0.002) 0.423
IS^a				−0.034 (0.015) 0.026
IS × time^b				−0.002 (0.002) 0.389
IV^a					0.052 (0.016) 0.001
IV × time^b					0.003 (0.002) 0.255
M10^a						−0.057 (0.018) 0.001
M10 × time^b						−0.005 (0.002) 0.026
L5^a							0.020 (0.022) 0.351
L5 × time^b							0.001 (0.003) 0.861

Linear mixed-effects models were used to examine the associations. Models were adjusted for age, sex, education, sleep duration, sleep fragmentation, vascular disease burden, vascular risk factors, and their interactions with time.
SD standard deviation, SE standard error, IS interdaily stability, IV intradaily variability, M10 the average activity during the most active 10-h period, L5 the average activity during the least active 5-h period.
^aResults for 1-SD change.
^bResults for 1-SD ×1-year change.

Quick links

Search