Topical Area: Methods and Protocols
Objectives : Investigators sought to: 1) evaluate the validity of a water frequency questionnaire (WFQ) to estimate mean daily water intake (WI) in adults through comparison to the gold standard, deuterium oxide (D2O) disappearance, and 2) evaluate reliability of WFQ to estimate WI.
Methods : Data collection occurred over three weeks, with validity of WFQ (vs. D2O) assessed during week one (W1) and reliability assessed between W1 and week three (W3). Healthy, free-living adults (n=103; 51% female; 41±14 y; BMI, 26.5±5.5 kg·m-2) consumed D2O (0.1 g·kg-1 lean mass) at the start of W1 and provided urine samples immediately before ingestion, the following day, and at the end of the week to calculate total body water turnover (WTO). Seven day beverage consumption during W1 and W3 was retrospectively estimated using the WFQ. The WFQ included 17 beverage types with specified volumes (e.g. water (8 oz); soft drink (12 oz)) and nine frequency options ranging from ‘Never or less than 1 per week’ to ‘7+ per day’. Investigators converted beverage volumes to mL and calculated WI for each week. Food frequency questionnaires were also completed at the end of both weeks. Diets were analyzed with Nutritional Data System for Research software to estimate water intake from solid foods and metabolic water production by macronutrient oxidation. Water from food and macronutrient oxidation were subtracted from WTO to estimate WI via D2O. WFQ reliability was assessed via intraclass correlation (ICC) and Cronbach α, and validity was assessed via Bland-Altman plots.
Results : The mean difference in WI between D2O (2902±1321 mL·day-1) and WFQ (2589±1560 mL·day-1) was -308±166mL (P=0.065). No systematic bias was observed between methods (R2=0.032, P=0.071) as 3% of WI variance was explained by differences between methods. WFQ WI was not different between W1 (2589±1560 mL·day-1) and W3 (2441±1334 mL·day-1, P=0.135). Cronbach α=0.8415 demonstrated high internal consistency in WI estimation. The ICC was 0.834 (95% CI: 0.754, 0.887), demonstrating good test-retest reliability of WFQ.
Conclusions : In conclusion, the WFQ produced reliable WI estimates that were comparable to the gold standard of D2O assessment. This questionnaire may be a practical and easy method for assessing water intake in adults.
Funding Sources :
The study was funded by Danone Research.
Graduate Teaching Associate
Hydration Science Lab, Arizona State University, Phoenix, Arizona
Human Integrated Physiology Laboratory, University of Wyoming, Laramie, Wyoming, USA
Department of Kinesiology, University of Montreal, Montreal, Quebec, Canada
University of Arkansas, Fayetteville, Arkansas, USA
Hispanoamerican University, San Jose, Costa Rica
Mayo Clinic, Division of Endocrinology, Rochester, Minnesota, USA
Danone Research, Palaisaeu, France
Hydration Physiology & Metabolic Health Group Manager
Agricultural Statistics Laboratory, University of Arkansas, Fayetteville, AR
Hydration Science Lab, Arizona State University, Phoenix, Arizona, USA