Table 3.

Statistical Testing for Differences in HEI-2010 Score by Education, Income, and Food Security

Socioeconomic Indicator	Difference from Reference Group (95% CI)	p-value	Change in HEI-2010 per year (95% CI)	Trend p-value	Interaction Coefficient (95% CI)	Interaction p-value
Education			0.17 (0.04 to 0.31)	0.01
<HS	Ref	n/a			Ref	n/a
HS	2.42 (1.13 to 3.71)	<.001			−0.06 (−0.32 to 0.20)	0.66
>HS	4.14 (2.98 to 5.29)	<.001			−0.07 (− 0.31 to 0.16)	0.56
Poverty to Income Ratio			0.18 (0.04 to 0.33)	0.01
< 100%	Ref	n/a			Ref	n/a
100–200%	0.74 (−0.79 to 2.27)	0.34			−0.10 (− 0.45 to 0.26)	0.59
>200%	3.65 (2.35 to 4.95)	<.001			−0.06 (− 0.33 to 0.21)	0.65
Food Security Status			0.23 (0.09 to 0.37)	0.002
Food Insecure	Ref	n/a			Ref	n/a
Food Secure	2.06 (0.81 to 3.31)	0.002			0.12 (−0.16 to 0.42)	0.39

Abbreviation: HEI-2010 Healthy Eating Index 2010, HS High School; Poverty to Income Ratio represents a ratio of participant’s household income to federal poverty threshold in year of data collection, accounting for household size. Results are from weighted linear regression model adjusted for race/ethnicity, year, age and gender. Models incorporate survey design information for standard errors and use dietary weights for representativeness. P-values are from t-statistics of regression coefficients. NHANES cycle was treated as an ordered categorical variable. For interaction testing, an interaction term between socioeconomic exposure and NHANES cycle was added to the ‘main effects’ model described above