Analysis Of Methods To Assess Fruit And Vegetable Intake Among An Ethnically Diverse Sample In Hawai‘I

Abstract

Background Studies indicate that the vast majority of adult Americans do not regularly consume recommended daily servings of fruits and vegetables. One major issue with research in this area is how to measure fruit and vegetable intake, especially in ethnically diverse populations.

Objectives The specific aims of the study were to a) compare the percentage of participants categorized as regularly consuming five or more („5 or more‟) daily servings of fruits and vegetables using two commonly used instruments, b) assess if percentages varied by race/ethnicity, and c) assess if the different instruments interacted with race/ethnicity.

Methods The source for the current study was the Healthy Hawai„i Initiative (HHI) which collected baseline data from adult residents of Hawai„i in 2002. The data used in this study are baseline data from the HHI longitudinal study. „Five or more‟ was calculated using two instruments: a single question instrument and a multi-item food frequency questionnaire (FFQ) 19-item instrument. The latter allowed for variations in how „5 or more‟ was calculated (e.g. not including fried potatoes). Percentages were compared overall and by race/ethnicity.

Results The percentage meeting „5 or more‟ criteria varied greatly depending on how „5 or more‟ was calculated ranging from 20.9% with the single question instrument to 60.8% when all items on the multi-item FFQ instrument were used. Caucasians were iii significantly more likely to consume „5 or more‟ than were Japanese and Filipinos. With the single question instrument the results for Filipinos were exceedingly low and inconsistent with results using the multi-item FFQ and with State of Hawai„i survey data for 2002. Female gender and older age were also associated with „5 or more‟ while education was not. No specific food items explained differences by race/ethnicity.

Conclusions The percentage of participants meeting „5 or more‟ criteria varied significantly depending on how „5 or more‟ was calculated. Race/ethnicity was significantly associated with „5 or more‟. The single question method for determining „5 or more‟ categorization appeared to interact with race/ethnicity, greatly underestimating intake for some groups relative to Caucasians and therefore it should not be used in studies in Hawai„i.