Objective: Develop three phytoestrogen exposure metrics based on a dietary questionnaire, 24-hour diet recall, and urinary biomarkers and determine the correlation of these metrics. Evaluate the relationship between phytoestrogen consumption and central adiposity in prepubescent girls in regression models with metrics developed from each of the three different phytoestrogen measurement tools.
Methods: A cross-sectional study design was used for this analysis, using participants in the Growing up Female study as the frame for the population. Girls with urinary biomarker measurements, questionnaire data, and diet recalls administered within 24 hours of the urine sample were eligible (n=46). Exposure metrics were developed and analyzed to determine correlations. Waist-to-hip and waist-to-height ratios were calculated and linear regression was used to examine the relationship between these ratios and the biomarker and diet recall data.
Results: No relationship was found between the questionnaire data and the urinary biomarkers. Linear regression analysis examined 3 different regression models with waist-to-height and waist-to-hip ratios and BMI percentile, weekly activity, age, and race. In all final models BMI percentile was significant with p<.0001. R squared values showed the strongest correlations between the waist-to-height ratios and variables in the full model of the analysis, specifically the model with daidzein urinary biomarker with R square= 0.777.
Conclusion: No statistical significance was found between any of the biomarker exposure metrics and either waist-to-hip or waist-to-height ratios. This was a well designed study that can be improved with a larger sample size and more descriptive questionnaire data. Future studies should focus on the temporality of the different exposure metrics to better determine their correlation with waist-to-hip and waist-to-height ratios.