Assessing Initial Validity and Reliability of a Beverage Intake Questionnaire in Hispanic Preschool-Aged Children

Abstract

Background: Understanding the relationship between high-calorie beverage consumption and weight gain requires an accurate report of dietary intake. A critical need exists to develop and test the psychometrics of brief quantitative tools for minority pediatric populations.

Objective: To modify the adult beverage intake questionnaire (BEVQ-15) for Hispanic preschool-aged children (BEVQ-PS) and test its validity and test-retest reliability in children aged 3 to 5 years.

Design: Cross-sectional. The modified quantitative 12-beverage category questionnaire assessed consumption of water, fruit juice, sweetened juice drinks, whole milk, reduced-fat milk, low-fat milk, flavored milk, carbonated sweetened drinks, diet carbonated drinks, sweet tea, tea with or without artificial sweetener, and sport drinks consumed during the past month. Hispanic mothers (n=109) recruited from day-care centers provided one 4-day food intake record (FIR) and completed two BEVQ-PS surveys during a 2-week period for their preschool-aged child. Data collection was conducted through one-on-one interviews in Spanish. Validity was assessed by comparing amounts (in grams) and energy intake (in kilocalories) for each beverage category between the first BEVQ-PS and the mean of the FIRs using paired t tests and Pearson's correlation coefficient. Criteria for validity were nonsignificant mean differences in grams and kilocalories from the first BEVQ-PS and mean of the FIRs beverage categories, and significant correlation coefficients between beverage categories. Test-retest reliability was assessed by comparing grams and kilocalories for each beverage category in the first BEVQ-PS with those from the second BEVQ-PS using Pearson's correlation coefficient. The criterion for reliability was a significant correlation coefficient between beverage categories. Significance was set at P<0.05.

Results: Mean differences between the first BEVQ-PS and FIR for water (42.4±23.1 g), sweetened juice drinks (-1.6±11.0 g), whole milk (18.3±9.91 g), sweetened carbonated drinks (-13.0±7.9 g), and total sugar-sweetened beverages (SSB) (1.4±8.9 g) were not significantly different, but were significantly correlated (r=0.20 to 0.37; P<0.05). Thus, validity criteria were met. With the exception of flavored milk and tea with or without artificial sweeteners, the remaining beverage categories-total beverages and SSB-in the first BEVQ-PS were correlated with those from the second BEVQ-PS (r=0.20 to 0.68; P<0.05), meeting reliability criteria.

Conclusions: Researchers and clinicians may use the BEVQ-PS to assess SSB, water, and whole-milk intake in Hispanic children. Additional modifications should be evaluated to assess total beverage intake.

Keywords: Beverages; Hispanic; Preschoolers; Reliability; Validity.