t�ddtdaa�t

RESEARCH

Research and Professional Briefs

Factors Associated with Low Drinking WaterIntake among Adolescents: The Florida YouthPhysical Activity and Nutrition Survey, 2007

SOHYUN PARK, PhD, MS; BETTYLOU SHERRY, PhD, RD; TERRENCE O’TOOLE, PhD; YOUJIE HUANG, MD, DrPH

stsastwintJ

abda(yudt

ihnbiai

(tvrsibiyibioct

ABSTRACTThere is limited information on which characteristics areassociated with water intake among adolescents. Thiscross-sectional study examined the association betweendemographic, dietary, and behavioral factors and low wa-ter intake as the outcome measure. Analyses were basedon the 2007 Florida Youth Physical Activity and Nutri-tion Survey using a representative sample of 4,292 stu-dents in grades six through eight in 86 Florida publicmiddle schools. Multivariable logistic regression wasused to calculate adjusted odds ratios (ORs) and 95%confidence intervals for factors associated with low waterintake (�3 glasses water per day). About 64% of studentshad low water intake. Factors significantly associatedwith low water intake were Hispanic ethnicity and non-Hispanic other (vs non-Hispanic white; ORs 0.79 and0.76, respectively), drinking no 100% juice, drinking it �1ime/day, and drinking it 1 to 2 times/day (vs drinking it3 times/day; ORs 1.83, 1.91, and 1.32, respectively),rinking no milk and drinking �2 glasses of milk/day (vsrinking �2 glasses/day; ORs 1.42 and 1.41, respec-ively), drinking �1 soda/day (vs drinking none; OR 1.40),rinking fruit-flavored drinks/sports drinks �1 time/daynd drinking it �1 time/day (vs drinking none; ORs 1.49nd 1.41, respectively), eating at a fast-food restaurant3 days/week (vs none; OR 1.38, respectively), not par-

icipating on team sports or participating on 1 to 2 team

S. Park is an epidemiologist and B. Sherry is a lead epi-demiologist, Division of Nutrition, Physical Activity,and Obesity, and T. O’Toole is a public health analyst,Division of Nutrition, Physical Activity, and Obesity,National Center for Chronic Disease Prevention andHealth Promotion, Centers for Disease Control and Pre-vention, Atlanta, GA. Y. Huang is the state chronic dis-ease epidemiologist, Bureau of Epidemiology, FloridaDepartment of Health, Tallahassee.

Address correspondence to: Sohyun Park, PhD, MS,Division of Nutrition, Physical Activity, and Obesity,National Center for Chronic Disease Prevention andHealth Promotion, Centers for Disease Control and Pre-vention, 4770 Buford Hwy, NE, Atlanta, GA 30341.E-mail: spark3@cdc.gov

Manuscript accepted: February 7, 2011.Published by Elsevier Inc. on behalf of the American

Dietetic Association.0002-8223/$0.00

tdoi: 10.1016/j.jada.2011.05.006

ports in previous 12 months (vs participating on �3eams; ORs 1.77 and 1.24, respectively), and consumingnack/soda while watching television/movies “sometimes”nd “most/every time” (vs never; ORs 1.65 and 2.20, re-pectively). The strongest factor associated with low wa-er intake was frequent consumption of snacks/sodashile watching television/movies. Although study find-

ngs should be corroborated in other states and in aationally representative sample, they may be useful inargeting adolescents for increased water consumption.Am Diet Assoc. 2011;111:1211-1217.

The prevalence of childhood obesity has increased sub-stantially in the United States (1,2) and is a majorpublic health concern because of morbidities associ-

ted with it (3-5). High consumption of sugar drinks haseen associated with obesity among youth (6-10). Sugarrinks are the largest source of added sugar (11) andccount for 7.1% of total energy in the average US diet12). Given that the average energy intake among USouth has increased during the past few decades, partic-larly from sugar drinks (13,14), the substitution ofrinking water for sugar drinks might help them to main-ain a healthier weight (10,15).

Adequate water consumption is important in prevent-ng dehydration, which has been associated with adverseealth outcomes, including certain cancers (16,17), coro-ary heart disease (16), kidney stones (16), and severeronchopulmonary disorders (18). Although total waterntake requirements can be met by consuming beveragesnd foods, drinking water is a zero-energy, thirst-quench-ng option (16) that obviously will not promote obesity.

The adequate intake (AI) level for water in any formfoods, beverages, and drinking water) was established byhe Institute of Medicine as a guide for preventing ad-erse effects of dehydration rather than for decreasingisks for chronic diseases (16). Specific AI levels vary byex and age. The AI levels for total water are 2.4 L/day,ncluding 1.8 L (�8 c) in the form of water or othereverages for boys 9 to 13 years of age and 3.3 L/day,ncluding 2.6 L (�11 c) as beverages for boys 14 to 18ears of age. The AI levels for total water are 2.1 L/day,ncluding 1.6 L (�7 c) in the form of water or othereverages for girls 9 to 13 years of age and 2.3 L/day,ncluding 1.8 L (�8 c) as beverages for girls 14 to 18 yearsf age. People who are physically active or living in hotlimates require even more water (16). However, based onhe 2005-2006 National Health and Nutrition Examina-

ion Survey (NHANES), the average total water intake

Journal of the AMERICAN DIETETIC ASSOCIATION 1211

oaatteyigicflpcclbobdc

MTYcisYkdDSd

itagccpuicc

csvfI

fvmsr

to

OTabvddvgvf

ESvyntDwwDvscpd�dBdi(t�whwmgw

SDssaaiuiiptapw

among US youth (aged 12 to 19 years) was 2.4 L/day (1.6L/day from food and beverages other than drinking waterand 0.8 L/day in plain water) (19).

Although many studies have examined the associationf sugar drink consumption with various demographicnd behavioral factors (14,20,21), few have examined thessociation between drinking water consumption andhese factors. Some studies have shown possible associa-ions between water intake and factors such as age, race/thnicity, education level, and physical activity amongoung children (22-24) or adults (25-27). However, theres limited information on the extent to which demo-raphic and behavioral factors are associated with waterntake among adolescents (19,26). Both societal norms/ulture/behaviors and individual knowledge/attitudes in-uence behaviors associated with weight status, so it isossible that low intake of drinking water may be asso-iated with other less-healthful behaviors, such as highonsumption of sugar drinks, excessive screen time, andow levels of physical activity. This is a potential linketween low water consumption and obesity. The purposef this cross-sectional study was to assess associationsetween low drinking water intake and demographic,ietary, and behavioral factors among Florida adoles-ents.

ETHODShis cross-sectional study was based on the Floridaouth Physical Activity and Nutrition Survey (YPANS),onducted in spring 2007. YPANS, developed by the Flor-da Department of Health, was a statewide, self-reported,chool-based survey of public middle school students.PANS monitored attitudes, health behaviors, andnowledge of physical activity and nutrition among stu-ents (28). YPANS used questions from the Centers forisease Control and Prevention Youth Risk Behaviorurveillance System, which had previously been vali-ated (29,30).The eligible population for the survey included all Flor-

da regular public middle school students in grades sixhrough eight. The population was stratified by regionnd sampled systematically proportional to enrollment inrades six through eight. Within schools, either alllasses meeting during a particular period of day or alllasses in a required subject were included in the sam-ling frame. Systematic equal probability sampling wassed to select classes from each school that participated

n the survey. All students in selected classes were in-luded in the sample. Every fourth student in selectedlasses was asked to complete the survey.

Consent forms from parents (passive consent in mostases) and assent from students were obtained before theurvey was administered. Because these analyses in-olved existing deidentified data, this study was exemptrom the Centers for Disease Control and Preventionnstitutional Review Board process.

A total of 4,669 students in grades six through eightrom 86 Florida middle schools completed the 2007 sur-ey. However, 377 respondents were excluded because ofissing data for the outcome variable. The final analytic

ample was 4,292 respondents. The estimated overall

esponse rate for the Florida YPANS was 70%, based on

1212 August 2011 Volume 111 Number 8

he number of schools that participated and the numberf eligible students.

utcome Variablehe outcome of interest was water intake. Students weresked, “During the past 7 days, how many glasses orottles of water did you drink?” For �2 tests, the outcomeariable was categorized into drinking �1 glass water peray, 1 to 2 glasses water per day, or �3 glasses water peray. For the logistic regression analyses, the outcomeariable, low water intake, was defined as drinking �3lasses water per day, based on the findings from a pre-ious study done in youth (19) and the data distributionrom our study.

xposure Variablestudent characteristics were examined. Demographicariables were age (12 years or younger, 13 years, and 14ears or older), sex, race/ethnicity (non-Hispanic white,on-Hispanic black, Hispanic, and other). Other charac-eristics were self-reported academic grades (A, B, C, and/F) and intention to control weight (trying to loseeight, trying to gain weight, trying to stay the sameeight, and not trying to do anything about weight).ietary variables (reported consumption during the pre-ious 7 days) were frequency of 100% fruit juice con-umed per day (none, �1, 1 to 2, and �3), glasses milkonsumed per day (none, �2, and �2), sodas consumeder day (none, �1, and �1), frequency of fruit-flavoredrinks or sports drinks consumed per day (none, �1, and1), number of breakfasts eaten (none, 1 to 6, and 7), andays in which fast food was eaten (none, 1 to 2, and �3).ehavioral characteristics were team sports participationuring the past year (none, 1 to 2 teams, �3 teams),ndividual sports participation during the past yearnone, 1 to 2 sports, and �3 sports), time spent watchingelevision on an average school day (none, �2 hours, and3 hours), time spent watching television on an averageeekend day (none, �2 hours, 3 to 4 hours, and �4ours), and frequency of snack or soda consumption whileatching television or movies (never, sometimes, andost/every time). Unknown values or missing data re-

arding exposure variables ranged from 0.3% to 16% andere excluded from analysis when the variable was used.

tatistical Analysisata were weighted by sex, grade, and region to repre-

ent all Florida public middle school students in gradesix through eight. Descriptive statistics were expresseds proportions. �2 tests were used to examine differencescross categories. Odds ratios (ORs) and 95% confidencentervals for low drinking water intake were calculatedsing multivariable logistic regression. Because of signif-

cant correlation between team sport participation andndividual sport participation, we used only team sportarticipation in the logistic regression model. Of note,here were no confounding effects of the various bever-ges included in the study on low water intake. Age, sex,lus all variables that were significantly associated withater intake based on �2 tests were included in the initial

logistic regression model development. However, vari-

rt�ddtdaa�tt�it2

adcesFw1i

ables that were not significantly associated with low wa-ter intake were deleted from the final model. For themodeling analyses, our sample size was reduced to en-sure that only individuals with complete data on all of thevariables examined were included. Our analytic samplefor the modeling was 2,790 students. All statistical testswere two-sided and assumed significance at P�0.05. Allstatistical analyses were done using Statistical AnalysisSoftware (version 9.2, 2009, SAS Institute Inc, Cary, NC)and accounted for the two-stage cluster sample (studentsselected within classrooms within schools) and selectionprobability of the survey design.

RESULTS AND DISCUSSIONAbout 64% of respondents reported drinking �3 glasseswater per day during the previous 7 days. Based on �2

tests, drinking water intake varied significantly by race/ethnicity and self-reported academic grades (Table 1) aswell as dietary (juice, milk, soda, fruit-flavored drinks/sports drinks, breakfast, and fast-food consumption) andbehavioral factors (sports participation, television view-ing, and consumption of a snack/soda while watchingtelevision/movies) (Table 2).

Results of multivariable logistic regression showed thefollowing factors to be significantly associated with lowdrinking water intake: Hispanic ethnicity and non-His-

Table 1. Drinking water intake among public middle school studentsSurvey, 2007a

Student characteristic All respondents

n 4™™™™™™™™Total 4,292 100Age (y) 4,279�12 1,584 35.7�1.613 1,422 33.7�0.9�14 1,273 30.5�1.2Sex 4,274Boy 2,061 51.1�0.9Girl 2,213 48.9�0.9Race/ethnicity 4,114White, non-Hispanic 1,737 45.3�2.8Black, non-Hispanic 840 20.0�2.1Hispanic or Latino 1,097 23.5�2.2Other, non-Hispanic 440 11.1�0.6Self-reported academic grades 3,800Mostly A’s 1,266 33.5�1.5Mostly B’s 1,449 38.1�1.0Mostly C’s 840 22.0�1.1Mostly D’s and F’s 245 6.5�0.6Intention to control weight 3,882Trying to lose weight 1,790 45.4�0.8Trying to gain weight 541 14.3�0.6Trying to stay the same weight 954 24.5�0.6Not trying to do anything about weight 597 15.9�0.6

aBecause of rounding, weighted percentages may not add up to 100%.b�2 tests were used for examining differences across categories.

panic other (vs non-Hispanic white; ORs 0.79 and 0.76, t

espectively), drinking no 100% juice, drinking it �1ime/day, and drinking it 1 to 2 times/day (vs drinking it3 times/day; ORs 1.83, 1.91, and 1.32, respectively),rinking no milk and drinking �2 glasses milk/day (vsrinking �2 glasses/day; ORs 1.42 and 1.41, respec-ively), drinking �1 soda/day (vs drinking none; OR 1.40),rinking fruit-flavored drinks/sports drinks �1 time/daynd drinking it �1 time/day (vs drinking none; ORs 1.49nd 1.41, respectively), eating at a fast-food restaurant3 days/week (vs none; OR 1.38, respectively), not par-

icipating on team sports or participating on one to twoeam sports in previous 12 months (vs participating on3 teams; ORs 1.77 and 1.24, respectively), and consum-

ng snack/soda while watching television/movies “some-imes” and “most/every time” (vs never; ORs 1.65 and.20, respectively) (Table 3).AI levels for beverages (including drinking water)

mong adolescents range from 1.6 L/day to 2.6 L/dayepending on sex and age, and those living in hotlimates should consume even more (16). However, av-rage water intake among Florida adolescents wasomewhat lower than that in other youth populations.or example, a mean water intake (tap and bottled)as 1.31 L/day among 138 Canadian adolescents (aged3 to 17 years) (31). In another study, mean plain waterntake was 0.80 L/day among 1,997 US youth (aged 12

tudents’ characteristics, Florida Youth Physical Activity and Nutrition

Drinking Water Intake During the Previous 7 d

1 glass/d 1-2 glasses/d >3 glasses/d P valueb

™™™™™™%�standard error ™™™™™™™™™™™™™™™™™38.4�0.9 25.2�0.7 36.3�0.8

0.758.1�1.3 25.2�1.1 36.7�1.29.3�1.4 24.2�1.2 36.6�1.57.7�1.6 26.5�1.4 35.8�1.3

0.188.5�1.4 24.0�1.0 37.5�1.28.3�1.1 26.6�1.1 35.1�1.0

�0.00017.5�1.2 27.7�1.2 34.8�1.05.2�1.8 21.8�1.7 33.0�1.67.8�1.6 22.9�1.1 39.3�1.52.6�2.6 26.1�2.3 41.3�2.3

�0.00012.9�1.6 29.5�1.3 37.6�1.57.8�1.5 24.7�1.2 37.5�1.43.5�1.5 21.9�1.5 34.6�1.66.4�3.9 19.4�2.6 34.1�4.1

0.087.2�1.4 23.8�1.1 39.0�1.32.3�2.3 25.1�2.0 32.6�2.47.9�1.7 27.3�1.6 34.9�1.56.4�2.0 26.9�1.9 36.7�1.9

by s

<

™™™™3

333

33

3433

3344

3433

o 19 years) (19).

August 2011 ● Journal of the AMERICAN DIETETIC ASSOCIATION 1213

Table 2. Drinking water intake among public middle school students by students’ dietary and behavioral factors, Florida Youth Physical Activityand Nutrition Survey, 2007a

Dietary or behavioral factor

Drinking Water Intake During the Previous 7 d

P valuebAll respondents <1 glass/d 1-2 glasses/d >3 glasses/d

n 4™™™™™™™™™™™™™™™™%�standard error ™™™™™™™™™™™™™™3100% fruit juice intake during the previous 7 d 3,741 �0.0001None 747 20.2�0.9 44.1�2.0 21.3�1.7 34.6�1.8�1 time/d 1,788 48.1�0.9 41.5�1.3 26.3�1.0 32.2�1.21-2 times/d 748 19.6�0.7 26.5�1.6 31.7�1.6 41.8�1.7�3 times/d 458 12.2�0.7 28.8�2.1 22.6�1.8 48.6�2.6Milk intake during the previous 7 d 4,107 �0.0001None 475 11.2�0.6 46.5�2.3 20.5�1.9 33.1�1.9�2 glasses/d 2,079 50.3�0.8 41.6�1.2 26.7�1.0 31.7�0.9�2 glasses/d 1,553 38.5�0.8 32.0�1.2 24.6�0.9 43.4�1.4Soda intake during the previous 7 d 4,110 �0.0001None 784 18.9�0.7 32.0�2.1 22.8�1.6 45.2�2.0�1 soda/d 2,056 50.1�1.0 42.5�1.2 25.3�0.9 32.1�1.0�1 soda/d 1,270 31.0�1.0 35.4�1.5 27.1�1.4 37.4�1.4Fruit-flavored drink or sports drink intake during

the previous 7 d3,875 0.001

None 580 14.8�0.6 36.5�2.5 21.0�1.8 42.6�2.2�1 time/d 1,794 47.1�0.8 39.8�1.1 26.9�1.0 33.3�1.1�1 time/d 1,501 38.0�0.9 37.0�1.4 26.3�1.1 36.7�1.2Eating breakfast during the previous 7 d 4,188 �0.0001None 591 14.0�0.8 46.6�2.4 21.1�1.9 32.3�2.31-6 mornings/wk 1,667 39.9�0.8 41.7�1.3 24.7�1.1 33.6�1.2Every morning 1,930 46.1�1.2 33.2�1.3 27.0�1.1 39.9�1.3Fast-food consumption during the previous 7 d 4,107 0.003None 1,066 26.2�0.9 36.7�1.9 22.3�1.6 41.0�1.81-2 days/wk 1,978 48.1�0.8 37.5�1.2 26.8�1.0 35.7�1.0�3 days/wk 1,063 25.7�1.0 41.6�1.4 25.9�1.4 32.6�1.5Team sports played during the past year 3,953 �0.0001None 1,549 38.9�1.1 44.6�1.3 23.4�1.1 32.0�1.11-2 teams 1,605 40.8�0.8 36.2�1.5 26.6�1.3 37.3�1.3�3 teams 799 20.3�0.9 30.3�1.7 24.8�1.5 44.9�1.8Individual sports played during the past year 3,934 �0.0001None 732 18.5�0.8 49.3�2.0 23.0�1.6 27.7�1.71-2 sports 2,090 53.1�0.9 39.3�1.5 26.9�1.2 33.8�1.1�3 sports 1,112 28.5�0.7 29.3�1.6 22.9�1.1 47.8�1.6Television viewing—weekday 3,672None 288 8.0�0.5 38.0�3.3 19.9�2.4 42.1�3.2�2 h/d 1,890 51.8�1.3 34.8�1.3 27.4�1.1 37.8�1.1�3 h/d 1,494 40.2�1.3 41.5�1.4 23.9�1.1 34.6�1.3Television viewing—weekend 3,599 �0.0001None 224 6.2�0.4 43.9�3.4 15.9�2.6 40.3�3.8�2 h/d 1,279 36.0�1.2 32.7�1.7 29.1�1.4 38.2�1.53-4 h/d 979 27.2�0.8 37.7�1.6 25.8�1.4 36.5�1.6�4 h/d 1,117 30.5�1.3 43.2�1.5 22.6�1.5 34.2�1.4Eating a snack or drinking a soda while watching

television or movies3,926 �0.0001

Never 250 6.6�0.4 30.0�3.1 19.9�2.5 50.1�3.0Sometimes 2,241 57.0�1.0 35.6�1.2 26.1�1.0 38.3�1.0Most or every time 1,435 36.4�1.0 43.5�1.4 25.0�1.2 31.5�1.1

aBecause of rounding, weighted percentages may not add up to 100%.

b�2 tests were used for examining differences across categories.

1214 August 2011 Volume 111 Number 8

included a sample of 2,790 students.

The finding that Hispanics were significantly less likelyto have a low drinking water intake than non-Hispanicwhites contrasted with results of a study of 3,867 USchildren and youth (aged 2 to 19 years) by Kant andGraubard (19), who found no significant difference inmean plain water intake levels among non-Hispanicwhites (0.63 L/day), Mexican Americans (0.51 L/day), andnon-Hispanic blacks (0.48 L/day) (19). There was no as-sociation between water intake and sex. Findings fromprevious studies of the association between water intakeand sex have been inconsistent. Based on 2005-2006NHANES data, mean plain water intake was signifi-cantly higher among boys than girls (aged 2 to 19 years)(19) and also higher among men than women (aged �19years) (26); results of a study based on 2005-2006NHANES data showed no significant different by sexamong adults (aged �20 years) (27); and results of aCanadian study involving subjects aged 0 to 60 yearsshowed a significantly higher mean water intake amongwomen (31). In a study among US adults, water intakeamong those who tried to lose weight during the previousyear was not significantly different from that amongthose who did not try to lose weight (27). Results of otherstudies addressing whether water intake might facilitateweight management have been mixed, and the conceptremains controversial (32).

A few studies examining possible associations betweenwater intake and dietary or behavioral factors have beenconducted among children (19) and adults (25,27). Con-sistent with previous findings, our study showed thatstudents with low water intake were less likely to be100% juice and milk consumers and more likely to be sodaand fruit-flavored drink/sports drink consumers. Ourfindings suggest that low water intake might be associ-ated with a higher consumption of a cluster of less health-ful beverage intake behaviors. In a study of 4,755 USadults (aged �18 years), regular soft drinks were con-sumed daily by 45% of water consumers and 63% ofnon–water consumers. Furthermore, after controlling fordemographics, water consumers were 25 times less likelyto drink sugar drinks than non–water consumers. In con-trast, milk was consumed daily by 49% of water consum-ers and 38% of nonconsumers; and fruit juices were con-sumed daily by 29% of water consumers and 17% ofnonconsumers (25).

The finding that Florida adolescents who participatedin team or individual sports during the previous yearwere likely to drink more water than those who did notwas similar to findings of Kant and Graubard (19) thatparticipation in physical activity was associated withhigher water intake among 3,867 US children and youth(aged 2 to 19 years) (19). Television viewing was signifi-cantly associated with low water intake in our study,whereas Kant and Graubard (19) found no significantdifference in plain water intake by television and com-puter use. Furthermore, consumption of a snack/sodawhile watching television/movies was associated with lowwater intake. Our findings may be attributable in part toextensive television advertisements for high-energy andlow-nutrient beverages and foods that are targeted toadolescents (33). Television viewing has been associatedwith sugar drink consumption (34). Another explanation

Table 3. Factors associated with low water intake (drinking �3glasses water/d) among middle school students, Florida YouthPhysical Activity and Nutrition Survey, 2007 (N�2,790)a

CharacteristicsAdjustedodds ratio

95%confidenceinterval

Age (y)�12 Referent group —13 0.87 0.71-1.08�14 1.04 0.86-1.24SexGirl Referent groupBoy 0.97 0.82-1.15Race/ethnicityWhite, non-Hispanic Referent group —Black, non-Hispanic 0.97 0.79-1.20Hispanic or Latino 0.79 0.65-0.97Other, non-Hispanic 0.76 0.59-0.98100% fruit juice intake

during the previous 7 dNone 1.83 1.39-2.42�1 time/d 1.91 1.46-2.491-2 times/d 1.32 1.01-1.72�3 times/d Referent group —Milk intake during the

previous 7 dNone 1.42 1.10-1.83�2 glasses/d 1.41 1.20-1.66�2 glasses/d Referent group —Soda intake during the

previous 7 dNone Referent group —�1 soda/d 1.40 1.14-1.71�1 soda/d 1.07 0.86-1.33Fruit-flavored drink or sports

drink intake during theprevious 7 d

None Referent group —�1 time/d 1.49 1.17-1.91�1 time/d 1.41 1.10-1.80Fast-food consumption

during the previous 7 dNone Referent group —1-2 d/wk 1.23 1.00-1.50�3 d/wk 1.38 1.04-1.83Team sports played during

the previous 12 moNone 1.77 1.45-2.161-2 teams 1.24 1.02-1.52�3 teams Referent group —Frequency of snack or soda

consumption whilewatching television ormovies

Never Referent group —Sometimes 1.65 1.21-2.25Most or every time 2.20 1.54-3.14

aTo have a complete data set without any missing data, this logistic regression model

might be an association between television viewing and

August 2011 ● Journal of the AMERICAN DIETETIC ASSOCIATION 1215

physical inactivity (35,36), which would lead to a de-creased need for fluid intake. Leisure-time physical activ-ity and higher average physical activity have been asso-ciated with higher water intakes among youth (aged 2 to19 years) (19) and adults (aged �20 years) (27).

Our study is one of the first to examine associationsbetween low water intake and demographic, dietary, andbehavioral factors among adolescents. However, there area few limitations. First, YPANS is self-reported and issubject to potential reporting bias. Second, because theterms “glass” and “bottle” were not defined in YPANS andwater intake from water fountains was not included inestimates of water intake, the actual amount of waterconsumed by respondents is uncertain. Third, due to thecross-sectional nature of the study design, no causal in-ferences can be made in these analyses. Our findingsrepresent only factors associated with the outcome of lowwater consumption. Longitudinal studies are needed toidentify predictors for low water intake and causal rela-tionships and mechanisms would have to be examined inexperimental laboratory studies. Finally, the responserate was 70%, and only 68% of survey participants wereincluded in the logistic regression, which further limitedthe generalizability of findings.

CONCLUSIONSNearly two thirds of respondents reported drinking �3glasses water daily in the previous week, and the per-centage who did so varied by demographic, dietary, andbehavioral characteristics. The factor most strongly asso-ciated with low drinking water intake was frequent con-sumption of a snack/soda while watching television/mov-ies. Other factors significantly associated with low waterintake were non-Hispanic ethnicity, low intake of 100%fruit juice and milk, high soda and fruit-flavored drink/sports drink intake, frequent consumption of fast foods,and nonparticipation in team sports. These findings sug-gest that efforts to increase water consumption amongadolescents may need to consider low daily water intakeas part of a cluster of other less-healthful behaviors andshould focus on those with these associated factors; how-ever, findings should be corroborated in other states andin a nationally representative sample.

STATEMENT OF POTENTIAL CONFLICT OF INTEREST:No potential conflict of interest was reported by the au-thors.

ACKNOWLEDGEMENTS: The authors thank theFlorida Department of Health for providing the YouthPhysical Activity and Nutrition Survey data.

The findings and conclusions in this article are thoseof the authors and do not necessarily represent theofficial position of the Centers for Disease Control andPrevention.

References1. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal

KM. Prevalence of overweight and obesity in the United States, 1999-2004. JAMA. 2006;295:1549-1555.

2. Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM. Prevalenceof high body mass index in US children and adolescents, 2007-2008.JAMA. 2010;303:242-249.

3. Dietz WH. Health consequences of obesity in youth: Childhood pre-dictors of adult disease. Pediatrics. 1998;101:518-525.

1216 August 2011 Volume 111 Number 8

4. Deckelbaum RJ, Williams CL. Childhood obesity: The health issue.Obes Res. 2001;9(suppl 4):239S-243S.

5. Li Y, Yang X, Zhai F. Childhood obesity and its health consequence inChina. Obes Rev. 2008;9(suppl 1):82-86.

6. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consump-tion of sugar-sweetened drinks and childhood obesity: A prospective,observational analysis. Lancet. 2001;357:505-508.

7. Ebbeling CB, Feldman HA, Osganian SK, Chomitz VR, EllenbogenSJ, Ludwig DS. Effects of decreasing sugar-sweetened beverage con-sumption on body weight in adolescents: A randomized, controlledpilot study. Pediatrics. 2006;117:673-680.

8. Malik VS, Schulze MB, Hu FB. Intake of sugar-sweetened beveragesand weight gain: A systematic review. Am J Clin Nutr. 2006;84:274-288.

9. Fiorito LM, Marini M, Francis LA, Smiciklas-Wright H, Birch LL.Beverage intake of girls at age 5 y predicts adiposity and weightstatus in childhood and adolescence. Am J Clin Nutr. 2009;90:935-942.

10. Muckelbauer R, Libuda L, Clausen K, Toschke AM, Reinehr T, Ker-sting M. Promotion and provision of drinking water in schools foroverweight prevention: Randomized, controlled cluster trial. Pediat-rics. 2009;123:E661-E667.

11. Guthrie JF, Morton JF. Food sources of added sweeteners in the dietsof Americans. J Am Diet Assoc. 2000;100:43-51.

12. Block G. Foods contributing to energy intake in the US: Data fromNHANES III and NHANES 1999-2000. J Food Comp Anal. 2004;17:439-447.

13. Duffey KJ, Popkin BM. Shifts in patterns and consumption of bever-ages between 1965 and 2002. Obesity. 2007;15:2739-2747.

14. Wang YC, Bleich SN, Gortmaker SL. Increasing caloric contributionfrom sugar-sweetened beverages and 100% fruit juices among USchildren and adolescents, 1988-2004. Pediatrics. 2008;121:e1604-e1614.

15. Wang YC, Ludwig DS, Sonneville K, Gortmaker SL. Impact of changein sweetened caloric beverage consumption on energy intake amongchildren and adolescents. Arch pediatr Adolesc Med. 2009;163:336-343.

16. Institute of Medicine, Food and Nutrition Board. Dietary ReferenceIntakes for Water, Potassium, Sodium, Chloride, and Sulfate. Insti-tute of Medicine Panel on Dietary Reference Intakes for Electrolytesand Water, Standing Committee on the Scientific Evaluation of Di-etary Reference Intakes. Washington, DC: National Academies Press;2005.

17. Jiang XJ, Castelao JE, Groshen S, Cortessis VK, Shibata DK, ContiDV, Gago-Dominguez M. Water intake and bladder cancer risk in LosAngeles County. Int J Cancer. 2008;123:1649-1656.

18. Manz F. Hydration and disease. J Am Coll Nutr. 2007;26(suppl 5):535s-541s.

19. Kant AK, Graubard BI. Contributors of water intake in US childrenand adolescents: Associations with dietary and meal characteristics—National Health and Nutrition Examination Survey 2005-2006. Am JClin Nutr. 2010;92:887-896.

20. Grimm GC, Harnack L, Story M. Factors associated with soft drinkconsumption in school-aged children. J Am Diet Assoc. 2004;104:1244-1249.

21. Storey ML, Forshee RA, Anderson PA. Beverage consumption in theUS population. J Am Diet Assoc. 2006;106:1992-2000.

22. Heller KE, Sohn W, Burt BA, Feigal RJ. Water consumption andnursing characteristics of infants by race and ethnicity. J PublicHealth Dent. 2000;60:140-146.

23. Sohn W, Heller KE, Burt BA. Fluid consumption related to climateamong children in the United States. J Public Health Dent. 2001;61:99-106.

24. Campos R, Montenegro-Bethancourt G, Vossenaar M, Doak C, Solo-mons N. Volume, frequency and participation in plain drinking waterconsumption by third and fourth-grade schoolchildren in Quet-zaltenango, Guatemala. Asia Pac J Clin Nutr. 2009;18:164-170.

25. Popkin BM, Barclay DV, Nielsen SJ. Water and food consumptionpatterns of U.S. adults from 1999 to 2001. Obes Res. 2005;13:2146-2152.

26. Fulgoni VL. Limitations of data on fluid intake. J Am Coll Nutr.2007;26(suppl 5):588s-591s.

27. Kant AK, Graubard BI, Atchison EA. Intakes of plain water, moisturein foods and beverages, and total water in the adult US population-nutritional, meal pattern, and body weight correlates: NationalHealth and Nutrition Examination Surveys 1999-2006. Am J Clin

Nutr. 2009;90:655-663.

28. Zapata LB, Bryant CA, McDermott RJ, Hefelfinger JA. Dietary and

3

physical activity behaviors of middle school youth: The YouthPhysical Activity and Nutrition Survey. J Sch Health. 2008;78:9-18.

29. Brener ND, Collins JL, Kann L, Warren CW, Williams BI. Reliabilityof the Youth Risk Behavior Survey questionnaire. Am J Epidemiol.1995;141:575-580.

30. Brener ND, Kann L, McManus T, Kinchen SA, Sundberg EC, Ross JG.Reliability of the 1999 Youth Risk Behavior Survey questionnaire. JAdolesc Health. 2002;31:336-342.

31. Pintar KDM, Waltner-Toews D, Charron D. Water consumption habitsof a south-western Ontario community. J Water Health. 2009;7:276-292.

32. Dennis EA, Flack KD, Davy BM. Beverage consumption and adultweight management: A review. Eat Behav. 2009;10:237-246.

3. Institute of Medicine. Food Marketing to Children and Youth: Threator Opportunity? Washington, DC: National Academies Press; 2006.

34. Barr-Anderson D, Larson N, Nelson M, Neumark-Sztainer D, Story M.Does television viewing predict dietary intake five years later in highschool students and young adults? Int J Behav Nutr Phys Act. 2009;6:1-8.

35. Lowry R, Wechsler H, Galuska D, Fulton J, Kann L. Televisionviewing and its associations with overweight, sedentary lifestyle, andinsufficient consumption of fruits and vegetables among US highschool students: differences by race, ethnicity, and gender. J SchHealth. 2002;72:413-421.

36. Hager RL. Television viewing and physical activity in children. JAdolesc Health. 2006;39:656-661.

August 2011 ● Journal of the AMERICAN DIETETIC ASSOCIATION 1217