Skip to main content
. Author manuscript; available in PMC: 2021 May 8.
Published in final edited form as: Biol Psychol. 2020 May 8;153:107886. doi: 10.1016/j.biopsycho.2020.107886

Table 1.

Multilevel Models with Parasympathetic Flexibility (Respiratory Sinus Arrhythmia Decreases in Response to Sad Film Relative to Neutral Film) and Person-Centered Stressful Life Event Exposure at Each Wave Predicting Real-World Rumination

Model (Predictor) B/γ SE p ΔR2
Predicting Real-World Brooding .75
Within-Subject Level (Level 1)
 RRS Brooding (t-1) 0.18 0.07 <.01 .05
 BDI (t) 0.18 0.04 <.001 .13
 Life Events Scale (LES) Slope 0.05 0.04 .24 .01
Associations with Levels of Brooding 1 (Level 2)
 RSA (Sad Film) 0.65 0.32 .04 .03
RSA (Neutral Film) −0.27 0.31 .39 .01
RRS Brooding 2.02 0.17 <.001 .52
T1 BDI 0.06 0.02 .02 .06
LES Mean 0.05 0.02 .04 .05
ΔNA 0.18 0.20 .35 .01
Associations between RSA and Brooding Responses to Stress (Level 2)
 RSA (Sad Film) × LES 0.17 0.08 .04 .03
RSA (Neutral Film) × LES 0.18 0.08 .03 .04
ΔNA × LES −0.06 0.03 .11 .03
Random Effects (Level 2)
RRS Brooding (Random Intercept) 6.49 0.74 <.001 .37
LES (Random Slope) <0.01 0.01 .25 <.01
Predicting Real-World Pondering .51
Within-Subject Level (Level 1)
RRS Pondering (t-1) 0.09 0.06 .13 .02
BDI (t) 0.13 0.04 <.001 .07
Life Events Scale (LES) Slope 0.16 0.05 <.01 .07
Associations with Levels of Pondering (Level 2)
RSA (Sad Film) −0.34 0.40 .40 .01
RSA (Neutral Film) 0.43 0.41 .29 .01
RRS Pondering 2.31 0.20 <.001 .50
T1 BDI 0.03 0.03 .36 .01
LES Mean 0.09 0.04 .02 .04
ΔNA 0.19 0.19 .31 .01
Associations between RSA and Pondering Responses to Stress (Level 2)
 RSA (Sad Film) × LES −0.07 0.07 .36 .01
 RSA (Neutral Film) × LES 0.01 0.07 .87 <.01
 ΔNA × LES −0.14 0.05 <.01 .06
Random Effects (Level 2)
 RRS Pondering (Random Intercept) 7.68 0.74 <.001 .45
 LES (Random Slope) <0.01 0.01 .19 <.01

Note. RSA = Respiratory Sinus Arrhythmia; R2 for full model represents pseudo R2, the proportion of variance explained by the model relative to an unrestricted (intercept-only) model containing no predictors (Kreft & Leeuw, 1998; Singer, 1998). ΔR2 = partial R2 of given predictor (proportion of variance in rumination predicted) after accounting for covariates. Significant hypothesized predictors of interest are bolded.