Table 3.

Multivariate Associations Between US School Food Item Accessibility and Student Fruit or Green Vegetable Consumption, 2008–2012

		Middle schoolb			High schoolc
Outcome	Food accessibilitya predictors	AOR	(95% CI)	p value	AOR	(95% CI)	p value
Fruit consumption
Regular prevalenced	Candy/regular-fat snackse	0.979	(0.962–0.997)	0.021	0.998	(0.987–1.009)	0.766
	Low-fat snacksf	1.004	(0.980–1.029)	0.735	0.993	(0.978–1.008)	0.343
	Fruits and vegetablesg	0.992	(0.957–1.028)	0.662	1.033	(1.009–1.058)	0.007
	Premade salads	0.952	(0.887–1.023)	0.179	0.981	(0.925–1.040)	0.522
	Salad bar	1.024	(0.962–1.089)	0.458	1.020	(0.974–1.068)	0.406
Daily prevalenceh	Candy/regular-fat snacks	0.986	(0.969–1.003)	0.106	1.001	(0.988–1.015)	0.826
	Low-fat snacks	0.996	(0.973–1.020)	0.750	0.988	(0.972–1.005)	0.167
	Fruits and vegetables	0.994	(0.962–1.028)	0.733	1.039	(1.010–1.068)	0.007
	Premade salads	0.969	(0.911–1.031)	0.320	0.972	(0.907–1.042)	0.423
	Salad bar	0.986	(0.926–1.051)	0.673	1.014	(0.963–1.068)	0.589
Green vegetable consumption
Regular prevalence	Candy/regular-fat snacks	0.986	(0.969–1.002)	0.093	0.989	(0.977–1.001)	0.082
	Low-fat snacks	1.017	(0.991–1.043)	0.214	1.002	(0.985–1.019)	0.835
	Fruits and vegetables	0.981	(0.944–1.020)	0.340	1.034	(1.008–1.061)	0.011
	Premade salads	1.008	(0.936–1.086)	0.827	0.958	(0.897–1.022)	0.195
	Salad bar	1.071	(1.008–1.137)	0.027	1.027	(0.977–1.079)	0.296
Daily prevalence	Candy/regular-fat snacks	0.991	(0.973–1.010)	0.338	0.991	(0.976–1.006)	0.226
	Low-fat snacks	0.996	(0.972–1.021)	0.749	0.998	(0.980–1.017)	0.873
	Fruits and vegetables	0.997	(0.960–1.036)	0.884	1.031	(0.999–1.065)	0.061
	Premade salads	0.990	(0.923–1.063)	0.787	0.977	(0.913–1.045)	0.496
	Salad bar	1.073	(1.006–1.145)	0.033	1.020	(0.962–1.081)	0.506

All food item accessibility measures entered simultaneously. All models controlled for student characteristics (gender, race/ethnicity, parental education, and number of parents in the home), school characteristics (presence of each of the three types of competitive venues and number of venues using competitive venue density category dummy terms, student body majority race/ethnicity, percent of students eligible for free and reduced price lunch), population density, region, and year. High school models also controlled for grade. Missing data for student control measures handled by imputing means (for continuous measures) together with the use of missing data indicators; no substantive differences were found, when compared with results from complete-case analysis. Estimates for missing data indicators not shown. Fruit consumption model Ns: 9836 middle school students in 310 schools; 18,060 high school students in 504 schools. Green vegetable consumption model Ns: 9969 middle school students in 310 schools; 18,208 high school students in 504 schools. Bold font indicates statistical significance (p<0.05 or less).

^aAccessibility indicates total number of school sources for each food type (regardless of venue).

^bMiddle school=grade 8.

^cHigh school=grades 10 and 12.

^dRegular consumption prevalence defined as nearly every day and every day versus other.

^eIncludes candy as well as any regular-fat items: salty snacks; baked goods; and frozen dairy desserts/yogurt.

^fIncludes any low-fat items: salty snacks; baked goods; and frozen dairy desserts/yogurt.

^gIncludes fresh fruit, other fruit, and vegetables (excluding potatoes).

^hDaily consumption prevalence defined as every day versus other.

AOR, adjusted odds ratio; CI, confidence interval.