Table 3 Results of the repeated measures analysis of variance with linear mixed model to assess the relationship between parametric circadian rhythm measures, collected during first week of treatment, and change in depression after that first week of treatment.

From: Circadian rhythm biomarker from wearable device data is related to concurrent antidepressant treatment response

Dependent Variable: ΔHDRS1 (N = 40)

Effects (DF = 33)

Estimate

Standard

Error

F

P

SSRI (vs. Placebo)

0.17

0.09

3.48

0.07

Sex: F (vs. M)

−0.16

0.11

2.00

0.17

Age (years)

−0.01

0.03

0.05

0.82

Age_sq

0.00

0.00

0.05

0.82

CQ_std

0.07

0.05

1.95

0.17

Acrophase_std

0.05

0.05

0.98

0.33

Week

−0.44

0.39

1.28

0.27

SSRI*Week

−0.14

0.07

3.93

0.06

Female*Week

−0.07

0.09

0.63

0.43

Age*Week

0.02

0.02

1.12

0.30

Age_sq*Week

−0.00

0.00

1.00

0.32

CQ_std*Week

0.11

0.04

7.01

0.01*

Acrophase_std*Week

−0.01

0.04

0.04

0.85

  1. ΔHDRS1: change in depression severity after first week of treatment, ln(Week1-HDRS/ Week0-HDRS), with a higher value indicating more depression symptoms and a lower value indicating improvement in depression. HDRS Hamilton depression rating scale. Week1-HDRS: Hamilton Depression Rating Scale score after first week of treatment. Week0-HDRS: Hamilton Depression Rating Scale score at baseline. N: sample size. DF: degrees of freedom = 33 for all the predictors in the analysis. SSRI: selective serotonin reuptake inhibitor. F Female. M Male. CQ_std: standardized values of circadian quotient (CQ = amplitude/MESOR), which reflects robustness of the circadian rhythm. Acrophase_std: standardized values of acrophase reflecting the time of peak activity. Week: treatment week (Week 0 vs. Week 1, repeated measure). The P-values that are statistically significant are indicated in bold for *P < 0.05 at α = 0.05.
Back to article page