The study aim was to develop a tool (software and ruler) to assess the dietary calcium and vitamin D intakes in Portugal, and evaluate the usefulness of non-dietary variables as intake predictors. Our findings indicated that is possible to estimate both using three and six food items, respectively, and non-dietary predictors.
Introduction: The study aim was to develop a tool to assess the dietary calcium and vitamin D intakes in Portugal, and evaluate the usefulness of non-dietary variables as predictors.
Methods: Trained interviewers collected information of 2,414 adults of Porto, Portugal, using a structured questionnaire and a validated semi-quantitative food frequency questionnaire (FFQ). Food items with the highest contribution to the total intake and non-dietary predictors (gender, age and body mass index (BMI)) were selected for the tool. Different statistical approaches were used to predict the intake. A Bland-Altman plot compared the predictions from the tool and the full FFQ.
Results: The items selected to predict intake were milk (38%), cheese (12%), yogurt (10%) and gender for calcium and oily fish (39%), canned fish (9%), white fish (7%), eggs (5%), red meat (5%), age and BMI for vitamin D. The Bland-Altman plot showed that the mean differences were 0.0 (limits of agreement = [-220.67; 220.77]) mg/day and 0.0 (limits of agreement = [-1.03; 1.05]) microg/day, respectively for calcium and vitamin D.
Conclusion: The equations estimated by the best statistical model to predict the calcium and vitamin D intake allowed for the design of a software and a circular ruler useful in clinical settings.