Testing the Use of Urinary Biomarkers to Assess Added Sugar Intake in Obese Postmenopausal Women

Abstract

RationaleAssessment of dietary intake is most frequently done by self‐reported measures such as 24‐hour recalls and food frequency questionnaires. Although these measures are useful, there is error associated with self‐report. This is found to be particularly true in overweight/obese individuals and for specific foods perceived as having negative health outcomes (e.g. added sugars) which are often under‐reported. Therefore objective measurements are needed to more accurately assess dietary intake of certain nutrients. Urinary sucrose and fructose excretion measured in urine samples collected over 24 hours has been validated as a biomarker of added sugar intake in controlled feeding studies. However, whether a single spot urine sample can be used to assess self‐reported added sugar intake is less clear, especially in overweight/obese individuals.ObjectiveTest correlations between urinary biomarkers from a single spot urine sample and self‐reported dietary intake of added sugars in obese postmenopausal women.MethodsFasting urine samples were collected at the baseline visit from 8 postmenopausal women with a BMI >30, who were enrolled in a dietary intervention study aimed at lowering added sugars and increasing fiber and fatty fish. Urine samples were transferred to cryovials and frozen at −20 C until assayed in batches. Urinary sucrose and fructose excretion was measured using a modified enzymatic assay (Rosche BioPharm). Participants completed 24 hour recalls at the baseline visit with the guidance of a Registered Dietitian using the USDA 5‐step multiple‐pass method designed to enhance accuracy of dietary data collection. Self‐reported added sugar intake (g/d) was estimated using Nutritional Data System for Research (NDSR). Pearson correlations were used to examine associations of added sugar intake with urinary sucrose and fructose excretion.ResultsPreliminary baseline data from 24‐hour recalls revealed a mean added sugar intake of 40.1 g/d (SD 25.2), with a range of 0 – 97 g/d. Mean urinary sucrose excretion was 0.024 g/L (SD 0.004), range 0.000 – 0.108 g/L. Fructose in the urine was undetectable. A statistically significant correlation (r=0.74, p=0.04) was observed between baseline urinary sucrose excretion and added sugar intake from 24 hour recalls. There was no significant correlation (r= 0.15, p=0.81) between baseline urinary sucrose excretion and self‐reported sucrose intake, which includes sucrose naturally found in plant foods.ConclusionPreliminary cross‐sectional data from a small sample of postmenopausal obese females suggest that urinary sucrose excretion can be used as an objective biomarker of added sugar intake. This warrants further testing in a larger sample.Support or Funding InformationInternal Ohio State University Funds