Figure 1.
Exclusion criteria and sample sizes for the individual- and neighborhood-level analyses.
The largest data loss occurs between study enrollment in 2000–2002 and Exam 5, in which this study occurs, in 2010–2012.
Table 1.
Demographic and socioeconomic characteristics of the Multi-Ethnic Study of Atherosclerosis cohort at Exam 5 (2010–2012), by organic produce consumption habits.
Table 2.
Prevalence ratios and 95% confidence intervals for the association between organic food consumption and individual-level demographic and socioeconomic characteristics in adjusted models.
Table 3.
Frequency of organic food consumption in relationship to measures of the local food environment.
Figure 2.
Associations of organic food consumption with neighborhood food accessibility.
Food accessibility is estimated by a) density of supermarkets (per increase in one supermarket per mile); b) self-report of fruit and vegetable selection in a participant's neighborhood (per one point increase on the Likert scale); and c) Aggregated Neighborhood Survey (per interquartile difference, represented by a 0.5 increase on the Likert scale). Models are adjusted for sex, age, education, income, metropolitan area, and race/ethnicity.