Skip to main content
. 2014 Apr 29;2:11. doi: 10.1186/2050-2974-2-11

Table 3.

Multivariate models for the longitudinal relationship between maternal and family variables and child eating disorder symptoms

  Model 1   Model 2  
ChEDE brief global
Coefficient (95% CI)
p
Coefficient (95% CI)
p
Parenting scale
0.01 (-0.12, 0.14)
.875
-0.03 (-0.16, 0.10)
.636
Life events scale
0.01 (-0.04, 0.07)
.562
0.02 (-0.03, 0.07)
.402
DASS depression
0.02 (0.01, 0.04)
.033*
0.02 (-0.01, 0.04)
.094
DASS anxiety
-0.01 (-0.03, 0.01)
.431
-0.02 (-0.04, 0.01)
.121
DASS stress
-0.01 (-0.02, 0.02)
.833
0.01 (-0.01, 0.02)
.645
SLSS family satisfaction
-0.27 (-0.38, -0.16)
<.001**
-0.05 (-0.17, 0.07)
.408
Maternal education (ref: completed high school)
 
 
 
 
  < High school
0.19 (0.03, 0.35)
.021*
0.18 (0.02, 0.33)
.023*
Concern about child weight (ref: no concern)
 
 
 
 
  Slight concern
0.01 (-0.19, 0.21)
.931
-.01 (-0.20, 0.18)
.908
  Moderate-high concern
0.67 (0.40, 0.93)
<.001**
0.64 (0.39, 0.88)
<.001**
ChEDE restraint scale
Coefficient (95% CI)
p
Coefficient (95% CI)
p
Parenting scale
0.10 (0.02, 0.18)
.021*
0.07 (-0.01, 0.16)
.100
Concern about child weight (ref: no concern)
 
 
 
 
  Slight concern
-0.01 (-0.13, 0.13)
.968
-0.02 (-0.16, 0.11)
.745
  Moderate-high concern
0.27 (0.11, 0.44)
.001**
0.25 (0.08, 0.41)
.003**
CARES emotional eating
Coefficient (95% CI)
p
Coefficient (95% CI)
p
Life events scale
0.05 (0.02, 0.08)
.001**
0.04 (0.01, 0.07)
.017*
DASS depression
0.01 (-0.01, 0.02)
.056
0.01 (-0.01, 0.02)
.292
Loss of control eating
Odds ratio (95% CI)
p
Odds ratio (95% CI)
p
DASS anxiety
1.06 (1.02, 1.14)
.150
1.03 (0.95, 1.12)
.470
SLSS family satisfaction 0.44 (0.29, 0.72) <.001** 0.56 (0.35, 0.90) .026*

Note. Models include variables that predicted the outcome significantly in univariate analyses. Model 1 is adjusted for child body mass index (BMI) z-score only. Model 2 is adjusted for child BMI z-score and child depressive symptom and global self-esteem scores. Coefficients show the predicted change in the dependent variable per unit change in the predictor variable, as identified through linear mixed models. Generalised estimating equations were used to determine odds ratios for the categorical outcome of loss of control eating.

*p < .05 **p < .001.