Skip to main content
. 2020 Jun 18;10:194. doi: 10.1038/s41398-020-00867-6

Table 1.

Weight table showing the population-level parameters in the HDRS total model (top) and the YMRS total model (bottom).

HDRS
μa τb
Predictor Mean (SD) 95% CIc |t|d Mean (SD) 95% CIc
Intercept 6.43 (0.66) 5.13 7.73 9.67 4.10 (0.50) 3.23 5.19
Mood negative −9.11 (1.40) −11.94 −6.43 6.51 0.56 (0.40) 0.02 1.50
Sleep negative −6.48 (1.66) −9.72 −3.19 3.89 0.42 (0.31) 0.02 1.16
Mixed Mood 2.11 (0.67) 0.79 3.42 3.15 0.44 (0.32) 0.02 1.20
Anxiety 2.26 (0.86) 0.58 3.96 2.63 0.38 (0.28) 0.02 1.06
Medicine changed −1.81 (0.71) −3.19 −0.40 2.55 0.35 (0.27) 0.01 0.99
Cognitive difficulty 1.09 (0.73) −0.35 2.48 1.50 0.43 (0.32) 0.02 1.19
Mood positive −2.80 (1.94) −6.59 0.94 1.44 0.42 (0.32) 0.02 1.19
Sleep positive 2.83 (2.05) −1.09 6.90 1.38 0.41 (0.31) 0.02 1.15
Activity 0.53 (0.61) −0.66 1.71 0.88 0.50 (0.35) 0.02 1.29
Stress 0.56 (0.73) −0.86 1.99 0.76 0.50 (0.36) 0.02 1.32
Alcohol 0.59 (1.01) −1.39 2.54 0.59 0.41 (0.31) 0.02 1.15
Medicine omitted 0.52 (0.97) −1.38 2.42 0.54 0.37 (0.28) 0.01 1.04
Irritable 0.05 (0.74) −1.41 1.49 0.06 0.59 (0.42) 0.02 1.57
YMRS
μa τb
Predictor Mean (SD) 95% CIc |t|d Mean (SD) 95% CIc
Intercept 3.10 (0.68) 1.79 4.46 4.59 4.35 (0.50) 3.48 5.40
Mood positive 12.83 (1.90) 9.09 16.53 6.75 0.57 (0.42) 0.02 1.57
Mood negative 3.42 (1.30) 0.87 5.99 2.63 0.66 (0.46) 0.03 1.68
Irritable 1.31 (0.69) −0.05 2.68 1.90 0.71 (0.47) 0.04 1.73
Mixed mood 1.02 (0.62) −0.20 2.21 1.65 0.54 (0.36) 0.03 1.34
Stress 1.15 (0.70) −0.21 2.54 1.65 1.24 (0.53) 0.12 2.18
Sleep positive −2.69 (1.84) −6.30 0.84 1.46 0.40 (0.30) 0.01 1.12
Activity −0.78 (0.56) −1.88 0.30 1.39 0.63 (0.41) 0.03 1.52
Medicine changed 0.46 (0.71) −0.99 1.81 0.64 0.80 (0.48) 0.05 1.76
Cognitive difficulty 0.41 (0.69) −0.92 1.78 0.59 0.94 (0.54) 0.05 1.99
Anxiety 0.18 (0.80) −1.40 1.73 0.23 0.69 (0.48) 0.03 1.76
Sleep negative 0.30 (1.50) −2.63 3.26 0.20 0.43 (0.32) 0.02 1.17
Alcohol 0.05 (0.93) −1.77 1.86 0.06 0.39 (0.30) 0.02 1.11
Medicine omitted −0.02 (0.90) −1.78 1.75 0.02 0.41 (0.31) 0.01 1.15

The population-level regression weight means, μ, are summarized in the leftmost columns and sorted by variable importance computed as the absolute t-statistic of the mean parameter. The corresponding variances, τ, are summarized in the columns to the right and can be interpreted as the amount of pooling of the given variable in the hierarchical model.

aPopulation-level regression weight means.

bPopulation-level variance can be interpreted as the amount of pooling of the given variable in the hierarchical model.

cCredible interval.

dAbsolute t-statistic of the mean parameter indicating variable importance.