Reproducibility and relative validity of a questionnaire to assess intake of black tea polyphenols in epidemiological studies

Epidemiological studies suggest that tea drinking may reduce the risk of ca rdiovascular diseases and cancers, Although tea is an important source of a ntioxidant phytochemicals, variation in preparation techniques may translat e to variation in antioxidant capacity. However, most large-scale epidemiol ogical studies use regular food frequency questionnaires to estimate tea in take, and nationally available nutrient analysis databases do not include l evels of black tea polyphenols. The Arizona Tea Questionnaire (ATQ) was des igned as a tool for collecting more complete dietary tea consumption inform ation, and a database was developed after analyzing 40 black tea samples (b rewed, instant, and sun tea) for polyphenols, This study assesses the relia bility and relative validity of the ATQ and polyphenol database. Relative v alidity of estimates of black tea consumption was tested by comparing the A TQ with the traditional Arizona Food Frequency Questionnaire and four days of food records. The ATQ was tested for reproducibility of estimates of bla ck (hot and iced) tea consumption and levels of black tea polyphenol intake . Correlations between two measures of intake taken 2 months apart ranged f rom 0.72 for black hot tea to 0.86 for black sun tea. Mean intakes (range) of total flavonoids for black tea consumers were 80.8 (3.0-588.0) mg/day at the first ATQ and 102.4 (4.5-802.3) mg/day at the second ATQ (r = 0.83, P < 0.001). The ATQ provided highly reproducible estimates of both total tea consumption and individual tea polyphenol intake. This instrument may be a useful tool in studies of the associations between tea consumption, tea pol yphenols intake, and risk for chronic disease.