Relative validation of a beverage frequency questionnaire in children ages 6 months through 5 years using 3-day food and beverage diaries

Objective To determine the relative validity of a quantitative beverage frequency questionnaire in assessing beverage, calcium, and vitamin D intakes using 3-day food diaries for reference. Design Parents were asked to complete questionnaires for the preceding week and diaries for the following week for their children. Calcium and vitamin D intakes were estimated from human milk, infant formulas, and cow's milk ("beverages") for questionnaires and diaries and from "all foods and beverages" for diaries. Data collected at 6 and 12 months and 3 and 5 years of age as part of the Iowa Fluoride Study (N = 700) were analyzed cross-sectionally. Subjects Children (N = 240); 60 randomly selected from each quartile of energy intake at 6 months of age. Statistical analyses Spearman correlation coefficients, weighted K statistics, and percentages of exact agreement were used to assess associations between tools. Results Correlations between mean daily beverage intakes estimated from questionnaires and diaries ranged from 0.95-0.99 for human milk, 0.84-0.85 for infant formula, 0.63-0.86 for cow's milk, 0.54-0.69 for juice/drinks, 0.26-0.59 for liquid soft drinks, 0.35-0.74 for powdered soft drinks and 0.54-0.70 for water. Correlations between mean daily nutrient intakes estimated from questionnaires and diaries "beverages" ranged from 0.64-0.74 for calcium and 0.60-0.80 for vitamin D; and between questionnaires and diaries "all foods and beverages" ranged from 0.41-0.63 for calcium and 0.43-0.80 for vitamin D. Applications A quantitative beverage frequency questionnaire can provide a relative estimate of beverage, calcium, and vitamin D intakes.