Skip to main content

Advertisement

Table 2 Various models with parameter estimates for slopes at each segment along with corresponding correlations and variances

From: Exploring diurnal variation using piecewise linear splines: an example using blood pressure

Parameter Model 1 Model 2 Model 3
Fixed effects Estimate (SE) Estimate (SE) Estimate (SE)
BP at 12.00 134 (0.54) 119.2 (4.6) 119.3 (4.6)
Microalbuminuria 7.57 (1.30)* 5.79 (1.67)*
Slope for spline time period
1. 12.00–18.00 0.02 (0.04) 0.03 (0.04) 0.03 (0.04)
2. 18.00–sleep −1.00 (0.04)* −1.00 (0.04)* −1.01 (0.04)*
3. Sleep–04.00 −1.93 (0.05)* −1.95 (0.06)* −1.99 (0.06)*
4. 04.00–wake 1.69 (0.05)* 1.70 (0.05)* 1.71 (0.05)*
5. Wake–12.00 2.23 (0.07)* 2.21 (0.07)* 2.26 (0.07)*
Microalbuminuria × spline interaction
1. 12.00–18.00 −0.06 (0.14)
2. 18.00–sleep 0.05 (0.13)
3. Sleep–04.00 0.37 (0.18)**
4. 04.00–wake −0.06 (0.16)
5. Wake–12.00 −0.48 (0.22)**
Random effects
Σ 223.6 199.5 200.5
−0.23 0.51 −0.23 0.50 −0.24 0.51
−0.23 −0.10 0.55 −0.25 −0.10 0.54 −0.25 −0.11 0.55
−0.23 −0.45 0.03 1.39 −0.28 −0.46 0.02 1.41 −0.28 −0.44 0.02 1.40
0.46 −0.28 −0.74 −0.05 0.66 0.47 −0.31 −0.74 −0.05 0.65 0.49 −0.33 −0.73 −0.04 0.65
0.34 −0.06 −0.21 −0.78 0.19 2.05 0.42 −0.03 −0.22 −0.80 0.23 2.00 0.42 −0.04 −0.20 −0.81 0.24 1.97
σ 12.3 12.3 12.2
ρ 0.27 0.27 0.27
R 2 0.67 0.68 0.68
Log-likelihood −149,608 −149,505 Model 2 versus Model 1 (p < 0.001) −149,502 Model 3 versus Model 2 (p = 0.12)
  1. Microalbuminuria: albumin:creatinine ratio ≥1.1 mg/mmol
  2. Model 1: Fixed effects (5 linear splines), random effects (5 linear splines)
  3. Model 2: Fixed effects (5 linear splines, microalbuminuria, age, sex, BMI), random effects (5 linear splines)
  4. Model 3: Fixed effects (5 linear splines and interaction with microalbuminuria, age, sex, BMI), random effects (5 linear splines)
  5. Random Effects matrix shown has variances on the diagonal and correlation coefficients on off-diagonals
  6. * p < 0.001; ** p < 0.05