Validation of FFQ-based assessment of dietary lignans compared with serum enterolactone in Swedish women

The validity of using FFQ to assess dietary lignans is uncertain. We aimed to validate the use of FFQ for the assessment of dietary intake of lignans compared to the serum biomarker enterolactone, the main product of dietary lignans' metabolism in human subjects. A random sample of women, aged 55-75 years, from the Swedish Mammography Cohort was selected. Information from two FFQ, the FFQ-87 (sixty-seven food items) and the FFQ-97 (ninety-three food items), and blood samples were collected. Dietary intake of lignans (secoisolariciresinol, matairesinol, lariciresinol, pinoresinol, medioresinol and syringaresinol) was assessed by the FFQ. Serum concentrations of enterolactone were analysed by time-resolved fluoroimmunoassay. The correlation coefficient between energy-adjusted lignan intake and serum enterolactone was estimated in crude and multivariable-adjusted models, taking into account the factors potentially influencing the serum enterolactone. Among the 135 participants aged 55-75 years, with a mean BMI of 26.7 kg/m(2), the average energy-adjusted intake of total lignans was 1616 (SD 424) and 1516 (SD 409) mu g/d according to the FFQ-87 (forty-five food items containing lignans) and the FFQ-97 (sixty-five food items containing lignans), respectively. The mean concentration of serum enterolactone was 23.2 (SD 15.4) nmol/l. The adjusted Pearson's correlation between dietary intake of lignans assessed by the FFQ-97 and serum enterolactone was statistically significant (r 0.22, P=0.01). No significant correlation was observed for the FFQ-87 (r 0.09, P=0.30). The present study indicates that the FFQ-97 might be better than the FFQ-87 for assessing dietary intake of lignans, although the correlation was low.