Journal of the American College of Nutrition, Vol. 26, No. 90005, 588S-591S (2007)
Published by the American College of Nutrition
Limitations of Data on Fluid Intake
Victor L. Fulgoni, III, PhD
Nutrition Impact, LLC, Battle Creek, Michigan
Address reprint requests to: Victor L. Fulgoni, III, PhD, Nutrition Impact, LLC, 9725 D Drive North, Battle Creek, MI 49014. E-mail: vic3rd{at}aol.com
 |
ABSTRACT
|
|---|
Background: The Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate, provided extensive tables of data on total water intake, water intake from beverages and water intake from foods; however, very little information was provided about the specific details of these analyses. Reported total water intake varied with gender and somewhat with age and suggested that further sources of variation in total water intake were due to differences in body size, physical activity, and climatic exposure.
Objective and Methods: In an attempt to better understand the variation in reported water intake, we examined water intake from foods and beverages and plain water intake in the National Health and Nutrition Examination Surveys (NHANES), 1999–2002.
Results: Non-Hispanic black children (4 to 18 years) had less water from food and beverages (
0.15 L/day) than non-Hispanic whites and Hispanics. Non-Hispanic black children had less (0.15 L/day) plain water than other ethnicities. Adult non-Hispanic whites had more water intake from food and beverages (0.4 L/day) and non-Hispanic blacks had less (0.2 L/day) than Hispanics and Other ethnic group. Non-Hispanic black adults had less (0.2 L/day) plain water than non-Hispanic whites and the Other ethnic group but had similar plain water consumption as Hispanics.
Conclusion: These data indicate further investigation into the sources of variation of water intake should include consideration of ethnic differences.
Key words: NHANES, water intake, Adequate Intake, limitations
|
INTRODUCTION
|
|---|
The Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate (DRI) [1], established an Adequate Intake (AI) for water based on median intake of water consumption and of water from foods and beverages. These intakes were estimated from 24-h recall data for food and beverages and certain food frequency questions (e.g., How much plain drinking water do you usually drink in a 24-h period? Include only plain tap or spring water) used in the National Health and Nutrition Examination Surveys (NHANES) 1988–1994, after adjustment (for most age groups) for intra-individual variation. The AI for water intake was set at 1.3–3.3 L/day for children and 2.7 to 3.7 L/day for adults. Water AI in men was set about 1 L/day higher than in women [1].
While the DRI report provides extensive data tables on total water intake and water intake from beverages and foods, it provides very little information on the specific details of these analyses. One striking aspect of the total water intake estimates is the wide range of intake reported. For example, in men and women ages 31 to 50 years, the difference between the 5th percentile and the 95th percentile was 4.2 and 3.7 L/day, with the range of intake exceeding the mean and median intake (3.9 and 3.6 L/day in men and 3.1 and 2.8 L/day in women, respectively).
In an attempt to quantify the variation in water intake estimates and compare this variation to that of other nutrients, a ratio of the 95th percentile to the 5th percentile was used. As a percent of mean intake, the range of total water intake was considerably less than the intake range of certain macronutrients, like protein and carbohydrate, but similar to the intake range of certain micronutrients, such as calcium (Table 1). Even with tremendous variation in reported water intake, serum osmolality (mmol/kg) was virtually unchanged across deciles of total water intake [1].
To help further understand sources of variation in total water intake, analyses were conducted using NHANES 1999–2000 data for water intake from food and beverages and from plain water.
|
MATERIALS AND METHODS
|
|---|
NHANES is an ongoing data collection initiative conducted by the National Center for Health Statistics of the Centers for Disease Control and Prevention to collect information about the health and diet of a nationally representative sample of a non-institutionalized US population. The NHANES design is a stratified, multi-stage probability sample. The NHANES 1999–2002 included oversampling in Mexican-Americans, non-Hispanic blacks, children, and adolescents. Additional information pertaining to the NHANES 1999–2002 survey design, survey methodology, and public use of the data is available from: www.cdc.gov/nchs/nhanes.htm [2].
Water intake from food and beverages was determined using available variables in the total dietary intake files. Plain water consumption was determined using response to the food frequency questionnaire (e.g., "Total plain water drank yesterday - including plain tap water, water from a drinking fountain, water from a water cooler, bottled water, and spring water"). Analyses were conducted for all children 4 to 18 years and for adults 19 years and older. Least square means and standard errors are presented for water from food and beverages and for plain water from models including gender, race/ethnicity, age, and body mass index status (normal, overweight, or obese). Various intake percentiles [i.e., 5th, 50th (median), and 95th] were also calculated. All analyses were weighted using the NHANES examination sample weights and adjusted for the complex sample NHANES design with the statistical package SUDAAN (SAS Version 9 from RTI International, Research Triangle Park, NC). Statistical hypotheses were tested at the 0.05 level of significance.
|
RESULTS
|
|---|
Water intake data from food and beverages for children and adults are presented in Table 2. Children consumed 1.4 L/day with males consuming significantly more than females (1.5 vs. 1.2 L/day). Adult water consumption from food and beverages was 2.0 L/day; males consumed significantly more water from food and beverages than females (2.4 vs. 1.8 L/day). There was a large range in the water amount from foods and beverages in both children and adults; the difference between the 5th and 95th percentile was over 2 L/day in children and over 3 L/day in adults. Non-Hispanic black children consumed significantly less water (0.15 L/day) from foods and beverages than did non-Hispanic white and Hispanic children. Adult non-Hispanic whites consumed more water from food and beverages than non-Hispanic blacks (0.60 L/day) and Hispanics or Other ethnic groups (0.35 L/day). Age significantly affected water consumption from food and beverages in both children and adults; older children consumed more water from foods and beverages but older adults consumed less water from foods and beverages. BMI status had no effect on water consumption from foods and beverages (data not shown).
Water intake data from plain water for children and adults are presented in Table 3. Children consumed 0.8 L/day plain water while consumption in adults was 1.8 L/day. Both children and adult males consumed significantly more plain water than females. Non-Hispanic black children consumed significantly less plain water (0.15 to 0.20 L/day) than children of other ethnic groups. Adult non-Hispanic whites and adults in the Other ethnic group consumed more plain water than non-Hispanic blacks (0.15 to 0.30 L/day). Age significantly affected plain water consumption in both children and adults; older children consumed more plain water and older adults consumed less plain water. Normal weight adults consumed less plain water than did obese adults (data not shown).
|
DISCUSSION
|
|---|
There are certain limitations on current water intake estimates. Total water intake estimates in the DRI combined data from a food frequency questionnaire on tap water consumption and water from foods and beverages from the 24-h dietary recall. Errors from both methods of estimating water intake will be at least additive, though no information is available to assess errors in these measurements regarding water intake. "Usual intake" calculations from the NHANES 1988–1994 data were conducted to establish the DRIs for water. While typically two days of intake can be used to calculate population estimates of usual intake of most nutrients, no data were uncovered confirming that two days of intake measures would be adequate to assess usual water intake.
Data have been reported validating the US Department of Agriculture's 5-Step Multiple Pass dietary recall method currently being used for NHANES. In men, there were no significant differences in dietary recall versus actual intake; however, energy, protein, carbohydrate, and fat were overestimated by the recall method by 7.4 to 8.5% [3]. In women, dietary recall was significantly higher for energy and protein (8.3 and 0.7%, respectively) but carbohydrate and fat were similar versus actual intake [4]. However, no data were presented on validating the 24-h recall neither for water intake from food and beverages, nor for the food frequency question regarding plain water intake.
NHANES 1999–2002 data on a second 24-h recall has not been made publicly available and thus data presented in this report are based on a single intake measure. As such, the variability of water intakes will reflect both differences between and within individuals since on any given day, a particular individual might consume much more food, beverages, and plain water or much less than usual. While mean intake levels from a single recall and multiple days of recall will be quite similar, water intake distribution will always be much larger when using only a single recall.
While there are limitations in water intake estimates, NHANES data are collected with the best methods currently available. Future research efforts should include a better understanding of variation in water intake, possibly by confirming current intake methods to accurately measure actual intake. In the mean time, current data on water intake estimates should be used with the knowledge of data limitations.
|
FOOTNOTES
|
|---|
Conflict of Interest Disclosure: There are no conflicts of interest to declare in connection with this work.
Received July 16, 2007.
|
REFERENCES
|
|---|
- Institute of Medicine:"Dietary Reference Intakes for Water, Potassium, Sodium, Chloride, and Sulfate."
Washington, DC: National Academics Press,2004
. Available from: www.nap.edu/catalog/10925.html
- National Health and Nutrition Examination Survey (NHANES): 1999 to Current. Washington, DC: National Center for Health Statistics. Available from: www.cdc.gov/nchs/nhanes.htm
- Conway JM, Ingwersen LA, Moshfegh AJ: Accuracy of dietary recall using the USDA five-step multiple-pass method in men: an observational validation study.J Am Diet Assoc104
:595
–603,2004
.[Medline]
- Conway JM, Ingwersen LA, Vinyard BT, Moshfegh AJ: Effectiveness of the US Department of Agriculture 5-step multiple-pass method in assessing food intake in obese and nonobese women.Am J Clin Nutr77
:1171
–1178,2003
.[Abstract/Free Full Text]
- Moshfegh A, Goldman J, Cleveland L:"What We Eat in America, NHANES 2001–2002: Usual Nutrient Intakes from Foods Compared to Dietary Reference Intakes."
Washington, DC: US Department of Agriculture, Agricultural Research Service,2005
.
This article has been cited by other articles:
|
|
|
|
 
A. K Kant, B. I Graubard, and E. A Atchison
Intakes of plain water, moisture in foods and beverages, and total water in the adult US population--nutritional, meal pattern, and body weight correlates: National Health and Nutrition Examination Surveys 1999-2006
Am. J. Clinical Nutrition,
September 1, 2009;
90(3):
655 - 663.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
