Résumé
ABSTRACT BACKGROUND: Few food frequency questionnaires (FFQ) have been validated for pregnant women, particularly those in small- and medium-sized cities in different regions of Brazil. OBJECTIVES: To validate and calibrate a semiquantitative FFQ for pregnant women. DESIGN AND SETTING: The study was validated with a sample of 50 pregnant women (≥ 18 years) enrolled in Brazilian prenatal services. METHODS: An FFQ and a 24-hour recall were used to evaluate dietary intake. Dietary variables were tested for normality and log-converted when asymmetrical. Pearson's Correlation Coefficient was used to validate the questionnaire. Linear regression was applied to extract calibration factors. All variables underlying the consumption analysis were adjusted for energy. RESULTS: The mean age of the pregnant women was 26 years ± 6.2 years; 58% were in their first trimester, and 30% were identified as overweight/obese. The Pearson correlation analysis results indicated that the FFQ overestimated energy and nutrient intake, whose coefficients ranged from −0.15 (monounsaturated fat) to 0.50 (carbohydrate). Adjusting for energy reduced the mean values of intake coefficients, which now ranged from −0.33 (sodium) to 0.96 (folate). The calibration analysis results indicated variation in the coefficients from −0.23 (sodium) to 1.00 (folate). Calibration produced satisfactory coefficients for the FFQ compared with the reference standard for energy, macronutrients, monounsaturated fat, cholesterol, vitamins B12/C, folate, sodium, iron, and calcium. CONCLUSIONS: After validating and calibrating tests, we observed that the FFQ was adequately accurate for assessing the food consumption of the pregnant women in this study.