Skip to main content

Table 2 Cross-lagged autoregressive models between sleep duration and PA

From: Sleep and physical activity – the dynamics of bi-directional influences over a fortnight

  Regression slopes for
Model 1 (Actual Sleep Time and MVPA)
Regression slopes for
Model 2 (Actual Sleep Time and VPA)
Estimate Standard error z p Estimate Standard error z p
Within-subject dependent variable: Actual Sleep Time
 Autoregressive association: from sleep to sleep 0.01 0.00 3.89  < .0001 0.01 0.00 3.87  < .0001
 Cross-lagged association: From PA to subsequent sleep 0.00 0.00 0.01 .990 -0.00 0.00 -3.65  < .0001
Within-subject dependent variable: PA
 Auto-regressive association: from PA to PA 0.06 0.00 14.34  < .0001 -0.06 0.00 -11.73  < .0001
 Cross-lagged association: from sleep to subsequent PA -1.92 0.17 -10.99  < .0001 0.23 0.07 3.14 .002
Fit Indices
 χ2 16,393.84     16,000.88    
 CFI 0.87     0.81    
 TLI 0.88     0.81    
 RMSEA 0.05     0.05    
 Scaled χ2 (df) 12,912.84 (418)    .000 10,656.40 (418)    .000
  1. PA Physical Activity
  2. MVPA Moderate to Vigorous Physical Activity
  3. VPA Vigorous Physical Activity
  4. CFI Comparative Fit Index
  5. TLI Tucker-Lewin Index
  6. RMSEA Root Mean Square Error of Approximation