Comparison of women's diet assessed by FFQs and 24-hour recalls with and without underreporters: associations with biomarkers

BACKGROUND/AIMS: Women's diet can be especially difficult to assess, as women tend to underreport their intakes more often than men and are more likely to do so if they think they are overweight or obese. The aim was to compare two methods to assess women's diet and how well they associate with biomarkers. The influence and frequency of underreporting was also investigated. METHODS: Diet of 53 women was assessed by two 24-hour recalls and a food frequency questionnaire (FFQ). Blood was analyzed for retinol, beta-carotene, vitamin C and serum ferritin, and 24-hour urine for nitrogen, potassium and sodium. Underreporting was evaluated with nitrogen excretion vs. intake, and energy intake vs. basal metabolic rate. RESULTS: Energy percent (E%) from macronutrients was similar from FFQ and 24-hour recalls, but total intake was higher from 24-hour recalls (9,516 +/- 2,080 vs. 8,183 +/- 2,893 kJ, p < 0.01). Intakes of vitamin C and potassium from both methods correlated with their respective biomarkers (r = 0.316-0.393). Underreporters had higher body mass index (BMI) than others (27.7 +/- 5.5 vs. 23.8 +/- 3.7 kg/m2, p < 0.05). They reported lower E% total fat (32 +/- 5 vs. 38 +/- 6 E%, p < 0.01) and higher E% carbohydrate (49 +/- 4 vs. 45 +/- 7 E%, p < 0.05). Correlation between intake and biomarkers increased after exclusion of underreporters. CONCLUSION: For women, FFQ and 24-hour recalls give similar E% and most nutrients correlate, but FFQ gives lower intake. Underreporters have higher BMI and diminish the correlation between calculated intake and biomarkers. This has to be considered when intake data are associated with weight management, disease and lifestyle factors.