M12 - Use of urinary sugars biomarker to assess measurement error structure of self-reported intake in NPAAS

Investigator Names and Contact Information

Victor Kipnis (vk3b@nih.gov)

Introduction/Intent

The association between added sugars in diet and chronic diseases, such as cancer, cardiovascular diseases and diabetes, has long been postulated. However, adverse health effects have been difficult to study in an epidemiological context, due to unreliability of self-reported sugars intake. Measurement error in self-reported dietary intake, a major source of bias in nutritional epidemiology, can attenuate estimates of disease risk and reduce statistical power, so that some potentially important associations may be obscured. Recently, 24-hour urinary sucrose and fructose were developed as dietary biomarkers for total sugars intake in two highly controlled feeding studies, and it has been suggested as a reference instrument to assess validity of self-reported sugars intake.

The aim of this project is to use the newly developed sugars biomarker to investigate the measurement error structure in self-reported sugars intake in the NPAAS. We propose to analyze sucrose and fructose in 538 urine collections from the NPAAS participants. Using data from the previous feeding study in which this biomarker was developed, we will derive the parameters of the biomarker measurement error model that will be used to estimate total sugar intake from urinary sucrose and fructose in the NPAAS participants. A multivariate regression model will be used to investigate whether participants’ characteristics, such as age, body mass index, race, education, physical activity, smoking or use of dietary supplements are associated with measurement error in self-reported sugars intake. The findings from this work will enable adjustment for measurement error in self-reported total sugars intake and correction of the risk estimates for the effect of sugars on disease outcomes in future analysis in the WHI investigating the association between dietary sugars and disease.

Aims

1. To estimate total sugars intake in the NPAAS participants based on the biomarker measurement (Table 1).
2. To investigate the measurement error structure in total sugars intake in the NPAAS participants as assessed by FFQ, 4-day food record (4DFR) and 24HDR.
3. To investigate whether the measurement error model depends on participants’ characteristics such as age, sex, BMI, race, education, physical activity, smoking or use of dietary supplements.