Validity of a web-based dietary questionnaire designed especially to measure the intake of phyto-oestrogens

A diet questionnaire (DQ) designed to assess habitual diet and phyto-oestrogen intake was developed. This study aimed to examine the validity of the DQ in men, with and without having prostate cancer. The DQ was validated against alkylresorcinol metabolites measured in urine as objective biomarkers of whole grain wheat and rye (WG) intake, and a 4-d estimated food record (FR) was used for relative comparison. Participants (n 61) completed both methods and provided spot urine samples. We found a statistically significant correlation between the DQ and FR for reported whole grain intake and isoflavonoids, as well as for intake of macronutrients, except protein. The correlation coefficient between the two methods was on average r 0·30, lowest for lignans (r -0·11) and highest for alcohol (r 0·65). Reported energy intake was lower in the DQ compared with FR (8523 v. 9249 kJ (2037 v. 2211 kcal), respectively; P = 0·014). Bland-Altman plots showed an acceptable agreement; most cases were within the limits (95 % CI) of agreement on reported energy intake, as well as intake of macronutrients, except protein (which was underestimated in the DQ compared with the FR). The correlation of alkylresorcinol with WG intake was statistically significant in the DQ (r 0·31, P = 0·015), but not in the FR (r 0·18, P = 0·12) and the weighted κ was 0·29 and 0·11, respectively. In conclusion, the results showed that the DQ have a reasonable validity for measuring WG intake and most nutrients, and, after some adjustments regarding protein intake assessment have been made, the DQ will be a promising tool.