Skip to main content

Table 2 Cluster analysis results

From: The association between mood state and chronobiological characteristics in bipolar I disorder: a naturalistic, variable cluster analysis-based study

Cluster Variable Standardized scoring coefficients R-squared with 1−R2 Variation explained by clusters Proportion explained by clusters
Own Next Ratio
Cluster Closest  
Cluster 1 PSQI 1.00 1.00 0.08 0.00* 6.12 0.36
Cluster 2 IS 0.23 0.84 0.25 0.22* 8.6 0.51
  RA 0.21 0.69 0.16 0.37   
  GOF 0.24 0.86 0.48 0.28   
  24-h correlation 0.22 0.74 0.40 0.43   
  CQ 0.20 0.61 0.21 0.49   
  SRM-5 0.11 0.18 0.05 0.87   
Cluster 3 Acceleration index 0.53 0.90 0.01 0.10* 10.36 0.61
  Activity index − 0.53 0.90 0.04 0.11   
Cluster 4 Sleep efficiency 0.31 0.38 0.07 0.66 11.87 0.70
  Sleep percent 0.45 0.82 0.21 0.23   
  Sleep time 0.45 0.80 0.05 0.21*   
Cluster 5 IV 0.24 0.44 0.28 0.77 12.77 0.75
  Mesor 0.23 0.95 0.15 0.05*   
  Amplitude 0.23 0.88 0.39 0.20   
  Activity mean 0.23 0.91 0.13 0.11   
  Activity median − 0.16 0.88 0.16 0.14   
  1. The results of the principal components-based cluster analysis (proc varclus) of chronobiological variables. Five cluster groups were identified. Summarized are the variables that comprise the clusters, the standardized scoring coefficients, and the representative chronobiological variables for each cluster (smallest 1−R2 within each cluster). * and italic font denotes most representative variable in the cluster. Presented are the values for the variation and proportion of the variation explained in the sample based on the number of chronobiological clusters considered. Demonstrated are the variation and proportion of variation explained in the sample when considering successive clusters. The total variation and total proportion of the sample explained when including all five clusters were 12.77 and 75.00%, respectively
  2. IS interdaily stability, IV intradaily variability, RA relative amplitude, CQ circadian quotient, GOF goodness-of-fit, SRM-5 5-Item Social Rhythm Metric, PSQI Pittsburgh Sleep Quality Index