Perceived and Observed Food Environments, Eating Behaviors, and BMI.

INTRODUCTION
This study examines relationships between perceived and observed nutrition environments, diet, and BMI, in order to examine the criterion validity of the Nutrition Environment Measures Survey-Perceived (NEMS-P).

METHODS
In a cross-sectional study, perceived nutrition environments were assessed (NEMS-P) among 221 adults from four neighborhoods in the Philadelphia area in 2010 and 2011. A total of 158 food store environments were observed using the NEMS-Stores. Data analyses were conducted in 2016. Bivariate Spearman rank correlations were used to examine relationships between perceived and observed availability, quality, and price of fruits and vegetables in respondents' neighborhoods. Linear regression models were used to examine relationships between perceived neighborhood and home food environments and daily fruit and vegetable consumption and BMI.

RESULTS
A significant, positive relationship was found between perceived and observed availability of fruits and vegetables in the neighborhood (r = 0.36, p<0.001). A similar relationship was seen between perceived and observed quality of fruits and vegetables (r = 0.34, p<0.001). Perceived availability and quality of fruits and vegetables in the neighborhood, and availability and accessibility of fruits and vegetables in the home, were significantly related to daily fruit and vegetable consumption. Perceived price of food in the neighborhood was significantly associated with BMI.

CONCLUSIONS
Responses to a self-reported survey to assess perceived food environments related to fruits and vegetables were significantly associated with observed nutrition environments, fruit and vegetable consumption, and BMI. The perceived prices of fruits and vegetables were modestly associated with BMI and warrant further testing in prospective studies. When observations of food environments are not feasible, residents' survey responses are an acceptable indicator, with reasonable criterion validity.