Objective: The reproducibility and validity of coffee, decaffeinated coffee and tea intake has not been adequately studied, particularly in Italy, where coffee drinking is peculiar in terms of type and amount of coffee consumed.
Design: We compared coffee and tea consumption, measured by two interviewer-administered food frequency questionnaires (FFQ), with average intake derived from two 7-day dietary (DD) records (the reference method) on 395 volunteers. The Pearson correlation coefficients (r) were used to assess both reproducibility and validity of information on coffee intake.
Results: A satisfactory level of reproducibility and validity of the pattern consumption was observed for coffee, decaffeinated coffee and tea. The reproducibility for both sex combined showed r of 0.74-0.78 for coffee, 0.57-0.65 for decaffeinated coffee and 0.61-0.67 for tea. The validity was about 0.70 for coffee, around 0.58 for decaffeinated coffee and 0.56-0.60 for tea intake.
Conclusions: The FFQ is a satisfactorily reliable and valid instrument for collecting information on coffee, decaffeinated coffee and tea intake.