Table 3.
Coefficient estimates from regression analyses examining associations of children's eating behaviours with family conflict, cohesion, and chaos (N = 548†)
| Dependent variables‡ | |||||
|---|---|---|---|---|---|
| Sugar‐sweetened beverages (servings/week) | 100% fruit and vegetable juice (servings/week) | Emotional eater§ | Non‐self‐regulated eater§ | Adventurous eater¥ | |
| Independent variable | β (SE), [95% CI] | β (SE), [95% CI] | β (SE), [95% CI] | β (SE), [95% CI] | β (SE), [95% CI] |
| Low family conflict | −0.05 (0.05) | −0.04 (0.21) | −0.68* (0.41) | −0.58** (0.24) | −0.15 (0.23) |
| [−0.14, 0.05] | [−0.45, 0.36] | [−1.49, 0.13] | [−1.04, −0.11] | [−0.59, 0.29] | |
| High family conflict | −0.01 (0.07) | 0.21 (0.28) | 0.85** (0.40) | −0.26 (0.31) | −0.11 (0.30) |
| [−0.14, 0.12] | [−0.33, 0.75] | [0.07, 1.62] | [−0.87, 0.34] | [−0.70, 0.49] | |
| Low family cohesion | 0.01 (0.07) | −0.32 (0.28) | −0.19 (0.43) | −0.15 (0.32) | −0.11 (0.31) |
| [−0.12, 0.14] | [−0.87, 0.24] | [−1.03, 0.65] | [−0.77, 0.48] | [−0.71, 0.50] | |
| High family cohesion | 0.04 (0.05) | 0.32* (0.19) | −0.03 (0.39) | 0.26 (0.22) | −0.14 (0.21) |
| [−0.06, 0.13] | [−0.06, 0.71] | [−0.79, 0.73] | [−0.18, 0.70] | [−0.56, 0.27] | |
| Low household chaos | 0.03 (0.05) | 0.58*** (0.22) | 0.47 (0.43) | 0.25 (0.25) | −0.32 (0.24) |
| [−0.08, 0.13] | [0.16, 1.01] | [−0.37, 1.30] | [−0.24, 0.73] | [−0.79, 0.14] | |
| High household chaos | 0.10* (0.05) | 0.13 (0.22) | −0.02 (0.39) | 0.33 (0.24) | 0.19 (0.24) |
| [−0.00, 0.20] | [−0.30, 0.55] | [−0.77, 0.74] | [−0.14, 0.80] | [−0.28, 0.65] | |
| R 2 (pseudo R 2) | 0.117 | 0.132 | 0.144 | 0.057 | 0.042 |
Note. For the regression of each outcome variable, the intermediate group for family conflict, cohesion, and chaos is used as a reference group. To avoid perfect collinearity, the dummy variables for the intermediate group are excluded in the regression. Linear regression analysis was performed in all models, except for Emotional Eater, Non‐Self‐Regulated Eater, and Adventurous Eater variables where logit regression analyses were performed.
Participants excluded from analyses due to missing child height and weight data (n = 2).
Models include family conflict, cohesion, chaos, and the following covariates: mother's education level, paid hours of employment/week, race/ethnicity, general health status, weight status, family affluence, household composition (single vs. dual parents, number of children <18 years in household), total number family meals/week, family meal atmosphere, food insecurity risk, and child's sex, age, weight status, and general health status.
Variable scores were dichotomized to represent two groups (0 and 1) in the regression. Scores representing nonexpression of the characteristic were coded as 0 (score < 3) or coded as 1 (score ≥ 3) to represent expression of the characteristic.
Variable score was dichotomized to represent two groups (0 and 1) in the regression. Scores representing nonexpression of the characteristic were coded as 0 (score ≤ 3) or coded as 1 (score > 3) to represent expression of the characteristic.
p < .10,
p < .05,
p < .01.