Table 3.

Hierarchical multiple regression analyses exploring the relation of FI to ED behaviors.

Step	Binge eating			Laxative/diuretic use			Exercise			Fasting			Vomiting			Dietary restriction
	F	R 2	b	F	R 2	b	F	R 2	b	F	R 2	b	F	R 2	b	OR	R 2	b
1: Covariates	29.6	0.024		3.96	.010		4.20	0.010		8.85	0.011		2.50	0.003			.089
Age
25–34			4.7			−0.1			−1.3			−0.4			0.6			−0.6
35–44			8.9			1.6			−1.4			−2.1			0.2			−0.8
45–54			10.7			1.3			−2.5			−3.3			−0.2			−1.0
55–64			11.7			−0.2			−4.4			−5.5			−1.6			−1.3
65–74			8.9			0.9			−4.6			−4.8			−2.8			−1.5
75–84			12.5			−1.0			−2.5			−5.9			1.3			−1.1
85+			−9.9			0.5			−8.6			−7.9			−2.1			0.2
Gender
Male						−1.5			−1.2			−2.2			−1.6			−0.9
Non-binary						−0.5			−2.2			0.1			−0.8			0.4
Self-described						−1.8			1.1			1.6			7.1			1.4
Prefer not to say						−1.0			1.8			2.1			−2.9			0.0
Income
$20 k–39,999																		0.1
$40 k–59,999																		−0.1
$60 k–79,999																		−0.1
$80 k–99,999																		−0.3
$100 k–149,999																		−0.3
$150 k+																		−0.3
2: FI	33.2	.030	4.9	4.55	.010	0.99	3.95	0.010	−0.45	14.1	0.020	3.4	4.29	0.006	2.07	1.12	.090	0.12

Note: The models examining binge eating included age as a covariate. The models examining laxative use, exercise use, fasting, and vomiting included age and gender as covariates. The model examining dietary restriction included gender, age, and income as covariates. R² in the logistic models refers to Nagelkerke R². The reference group was 18–24 for age, female for gender, and <$20 k for income. Cohen’s f² was calculated as a measure of effect size for linear models as follows: 0.001 for binge eating, 0.002 for laxative/diuretic use, 0.000 for exercise, 0.000 for fasting, and 0.000 for vomiting. f² ≥ 0.02 = small effect size, f² ≥ 0.15 = medium effect size, f² ≥0.35 = large effect size. Bolded statistics indicate significance at p < .05.