European Journal of Clinical Nutrition (2015), 1–6

© 2015 Macmillan Publishers Limited All rights reserved 0954-3007/15

www.nature.com/ejcn

ORIGINAL ARTICLE
Fluid intake and hydration status in obese vs normal weight
children
C Maffeis1, M Tommasi1, F Tomasselli1, J Spinelli1, E Fornari1, N Scattolo2, M Marigliano1 and A Morandi1

BACKGROUND/OBJECTIVES: Little is known on the relationship between obesity and hydration levels in children. This study
assessed whether and by which mechanisms hydration status differs between obese and non-obese children.
SUBJECTS/METHODS: Hydration levels of 86 obese and 89 normal weight children (age: 7–11 years) were compared. Hydration
was measured as the average free water reserve (FWR = urine output/24 h minus the obligatory urine output [total 24 h excreted
solutes/97th percentile of urine osmolality of children with adequate water intake, that is, 830 mOsm/kg]) over 2 days. Three days of
weighed dietary and fluid intakes were recorded. Non-parametric tests were used to compare variables that were skewed and to
assess which variables correlated with hydration. Variables mediating the different hydration levels of obese and normal weight
children were assessed by co-variance analysis.
RESULTS: Obese children were less hydrated than normal weight peers [FWR = median (IQR): 0.80 (−0.80–2.80) hg/day vs 2.10
(0.10–4.45) hg/day, P o 0.02; 32% of obese children vs 20% of non-obese peers had negative FWR, P o 0.001]. Body mass index
(BMI) z-score (z-BMI) and water intake from fluids correlated with FWR (ρ = − 0.18 and 0.45, respectively, both Po 0.05). Water intake
from fluids completely explained the different hydration between obese and normal weight children [FWR adjusted for water from
fluids and z-BMI = 2.44 (0.44) hg vs 2.10 (0.50) hg, P = NS; B coefficient of co-variation between FWR (hg/day) and water intake from
fluids (hg/day) = 0.47, P o 0.001].
CONCLUSIONS: Obese children were less hydrated than normal weight ones because, taking into account their z-BMI, they drank
less. Future prospective studies are needed to explore possible causal relationships between hydration and obesity.
European Journal of Clinical Nutrition advance online publication, 14 October 2015; doi:10.1038/ejcn.2015.170

INTRODUCTION Therefore, the aim of this study was to assess whether fluid
Obesity is the most common nutritional disorder in children from intake and hydration status are associated with body size in a
western or westernized countries, with a relevant impact on public sample of obese and normal weight children.
health. Obese children have a high chance (40–80%) of becoming
obese adults, may suffer from several metabolic and non-metabolic
complications, have a higher risk of developing cardio-metabolic MATERIALS AND METHODS
diseases later in life and show lower life expectancy than individuals Subjects
who were not obese in childhood.1–3 Two hundred pre-pubertal subjects from all socio-economic strata were
Food intake together with inadequate physical activity is a crucial recruited from a single large primary school district in the main residential
factor that promotes excess fat mass gain in obesity prone area of Verona (Italy), with permission from the local school authorities.
individuals.4 Dietary food energy density is positively associated Sample size was decided on to ensure that at least 85 obese children and
with body weight in adults and children.5 Water content in food as 85 normal weight children would complete the study, in order to have 90%
well as fluid intake may have a role in body weight homeostasis by statistical power to detect a minimal difference of 150 g of free water
reserve (FWR) (see study protocol) between the two groups, assuming a
affecting overall diet energy density. Recent data have shown that
5% probability of alpha error and a standard deviation of 100 in both
hydration status was associated with both energy density of the diet groups.
and dietary profile: children with higher hydration levels were those Inclusion criteria were as follows: age range 7–11 years; Caucasian
with a dietary profile closer to the Recommended Dietary Intake.6 ethnicity; body mass index (BMI) higher than the age and sex BMI cutoff for
Several beverages contain macronutrients as well as minerals obesity (obese group); and BMI lower than the age and sex BMI cutoff for
and vitamins that contribute to total macro- and micro-nutrient overweight (normal weight group). International Obesity Task Force cutoff
intake. Therefore, type and quantity of liquid intake contribute to references were used.7 Exclusion criteria were as follows: congenital or
total energy and nutrient intake, as well as to the energy density chronic diseases, malformations, chronic use of drugs and puberty.
Children of parents who sent back a positive reply to our written invitation
of the diet. By implication, fluid intake may reasonably affect both
were given a physical examination and recruited if they qualified based on
body composition and hydration status. the inclusion/exclusion criteria.
To the best of our knowledge, no studies have explored the Both obese and normal weight children were recruited and participated
relationship between fluid intake, hydration status and body in the study lasting an entire year, with similar recruiting rates across the
composition in children. seasons.

1
Unit of Pediatric Diabetes and Metabolic Diseases, Department of Life and Reproduction Sciences, University of Verona, Verona, Italy and 2Chemo-Clinical Analysis Laboratory,
Frà Castoro Hospital, San Bonifacio, Verona, Italy. Correspondence: Professor C Maffeis, Unit of Pediatric Diabetes and Metabolic Diseases, Department of Life and Reproduction
Sciences, University City Hospital, 1P.le Stefani, Verona 37126, Italy.
E-mail: [email protected]
Received 3 June 2015; revised 31 July 2015; accepted 12 August 2015
 Obesity and hydration status in children
C Maffeis et al
2
Informed consent was obtained from children and their parents. calculated as the daily water intake from metabolism/overall water
The protocol was approved by the Institutional Ethics Committee of intake × 100.6 For each participant, we assessed whether the child was
Verona (Italy). very probably misreporting nutritional intakes, by verifying whether their
reported EI/estimated basal metabolic rate ratio (EI/BMR) fell outside the
95% confidence interval of EI/BMR identified by Goldberg’s equation,
Study protocol
which corresponds, at the individual level and using a 3-day food record
After physical examination, the children and parents were provided with and assuming a physical activity level of 1.55, to a 1–2.40 interval.16 No
two urine containers for urine collection and a digital scale for weighing participant fell outside this interval.
food. Moreover, they were instructed by a dietician to fill out a weighed
dietary record for 3 days, including one weekend day. During the second
Hydration status assessment. Twenty-four hour urine samples were
and third day of diet recording, 24-h urine was collected. At the end of
collected on two independent days. The children and their parents were
each day of urine collection, urine containers were delivered to the
carefully instructed on how to collect a 24-h urine sample, from about 0700
laboratory for volume measure, and samples were frozen and stored for
biochemical analysis. to 0700 h the next day. The timing of voiding was recorded. Urine samples
were considered acceptable if the urine creatinine excretion value related
Physical characteristics. At recruitment, physical examination was to body weight was above the fifth percentile for the corresponding age
performed for anthropometric measurements (height, weight, waist and sex groups.17 As a marker of hydration status, average urine osmolality
circumference), analysis of body composition and clinical assessment of was calculated as the average urine osmolality (2 × Na+[Urea]/2.8)/urine
puberty. Weight was measured to the nearest 0.5 kg on standard weight) over the two 24 h collections. Moreover, to estimate the ‘water
physician’s beam scales, with the child wearing only underwear and no reserve’, the FWR was used.17 FWR was calculated as measured urine
shoes. Height was measured to the nearest 0.5 cm on a portable amount (kg/24 h) minus the obligatory urine volume; the obligatory urine
stadiometer without shoes, with the child’s heels, buttocks, shoulders and volume (kg/24 h) results from the excreted 24- h solutes (mOsm/24 h)/the
head against the vertical wall with line of sight aligned horizontally. BMI 97th percentile of urine osmolality of healthy children with adequate water
was calculated as weight (in kilograms) divided by height (in meters) intake, that is, 830 mOsm/kg (physiologic criterion of the upper limit
squared. BMI values were standardized (BMI z-scores) using age (to the of euhydration adopted for children with a typical western diet).18–20 The
nearest 6th month) and sex-specific median, standard deviation and power mean of the FWR measured from the 2 days of urine samples was used.
of the Box-Cox transformation (least mean square method) based on Positive values of FWR indicate euhydration, and negative values indicate
national norms.8 Waist circumference was measured to the nearest 0.5 cm the risk of hypohydration.20
while the subjects were standing, after gently exhaling, as the minimal
circumference measurable on the horizontal plane between the lowest
portion of the rib cage and the iliac crest. The same investigator measured Statistical analysis
height, weight and waist circumference. Waist-to-height ratio was Data were shown as median and interquartile range. All analyses were
calculated and used as an index of body fat distribution, as previously performed in a pooled group of boys and girls because preliminary
described.9,10 Pubertal status was assessed using Tanner stages for pubic analyses showed that, both in normal weight and in obese children, FWR,
hair, breast configuration and genital status as references.11 Body fat mass as well as simple correlations between FWR and anthropometric and
and body fat-free mass were estimated by the BIA (bioelectrical impedance dietary variables, did not differ by gender. Proportions between obese
analysis) scale (Bioelectrical Impedance, BIA: Tanita BC 420 MA, Tanita and normal weight children were analyzed by the χ2 test. Physical
Corporation, Tokyo, Japan), using the manufacturer’s equations.12 characteristics, dietary data and hydration status of obese and normal
weight children were compared by the Mann–Whitney U-test, due to
Dietary intake of macronutrients and water. A 3-day weighed dietary non-normal distribution of variables. Spearman’s correlation data were
record of food and fluid and the amount consumed was kept by the used to assess correlations between hydration status, body composition
parents.13,14 The electronic scale provided by the study personnel at and dietary variables. On the basis of the correlation analysis results,
recruitment was used by the child’s parents, and in a few instances by the we built a co-variance analysis (ANCOVA) model with FWR as the
grandmothers, to weigh food. Parents reported the total food intake of dependent variable, obesity as the fixed factor (yes = 1, no = 0) and
their children, both at meals and between meals, and the children were BMI z-score, energy density of fluids and fluid intake as covariates, to
encouraged to report all the foods, including snacks, consumed outside search for variables explaining the different FWR between normal
the home. Kitchen supervisory staff recorded the food consumed at school. weight and obese children. Data were analyzed using SPSS version 20.0
Each family was provided with a logbook for recording foods and software (SPSS, Chicago, IL, USA). A P-value of o 0.05 was considered as
beverages consumed. Written instructions with examples of completed statistically significant.
forms were provided. A complete description of how the food was
prepared and recipes for composite dishes were also requested. A dietician
checked logbooks with each family for completeness and accuracy. As an
aid in determining the amount of food and drinks consumed outside the
RESULTS
home, pictures of different items were shown, along with cups, glasses, Twenty-five subjects, eleven males and fourteen females, did not
spoons and food shapes of different portion sizes. Food and drink energy complete dietary recording and/or urine collection or had a urine
values, as well as their water content, were calculated from tables of food creatinine excretion value related to body weight below the fifth
composition set out by the National Institute of Nutrition, with the use percentile for the corresponding age and sex groups and were
of a computerized database and analysis program (Metadieta, Meteda,
therefore excluded from data analysis. The results reported refer
S.Benedetto del Tronto, Italy).15 Diet energy density was calculated as a
ratio between energy intake (EI) (kcal/day) and weight of foods and fluids to 175 children (89 males and 86 females). Age and BMI of the
consumed (g/day). excluded children were not significantly different from those
The post absorption hydration index (PAI) was used to describe the included in the data analysis.
amount of water available after consuming 100 g of food and beverages.6
As the amount of water available to prevent dehydration immediately after
Physical characteristics
consumption is not equivalent to the amount of water consumed, the PAI
considers the amount of carbohydrates and sodium of foods and Age and gender distribution were not significantly different in the
beverages affecting the availability of water. The formula of the PAI obese and normal weight children (Table 1). Physical character-
(g functional water/100 g food or beverage) is as follows: x g water/100 g – istics (height, weight, BMI, BMI z-score, waist circumference, waist/
(y g carbohydrates/100 g × 3) – (z mg sodium/100 g × 0.3). Correspond- height ratio) as well as fat mass and fat-free mass were
ingly, the percent water-supplying of foods refers to the amount of significantly higher in obese than in normal weight children.
functionally available water per 100 g of food intake (g/day) in foods such
as fruits and vegetables, butter, cheese and oil, classified according to the
When expressed in kilograms, total body water was significantly
PAI40 g. The water intake from food metabolism was calculated as fat higher in obese than in normal weight children, whereas it was
intake (g/day) × 1.07 + carbohydrate intake (g/day) × 0.55 + protein intake significantly lower when expressed as a percentage of body
(g/day) × 0.41, and the percent water intake from metabolism was weight.

European Journal of Clinical Nutrition (2015) 1 – 6 © 2015 Macmillan Publishers Limited

 Obesity and hydration status in children
C Maffeis et al
3
Dietary intake different in obese and normal weight children. Sources of fluid
Table 2 shows the energy and nutrient intakes of obese and intake as percentage of total did not differ significantly in obese
normal weight children. Protein intake expressed as a percent of and normal weight children, and water intake accounted for about
total EI was significantly higher in obese than in non-obese 70% of total fluid intake in both groups (Figure 1).
children but not when expressed as grams per day. Simple Total water intake, as well as water from metabolism, food and
carbohydrate intake was significantly lower in obese than in non- beverages, was not significantly different in obese and normal
obese children. No significant differences in energy, total weight children (Table 3).
carbohydrate, starch, lipid and fiber intake were found between
the two groups. The energy density of the diet was significantly Hydration
higher in obese than in normal weight children.
The FWR was negative in 34.2% of obese and in 20.0% of normal
weight children (χ2 test = 4.63, P o 0.05). Urine osmolality was
Fluid intake significantly higher (P o 0.05), and FWR was significantly lower
Fluid and water intake from different sources is reported in (P o 0.02) in obese than in normal weight children (Table 4).
Table 3. More than 50% of obese and normal weight children did
not consume fruit juice or herbal infusions at all (median = 0).
Intake of water, fruit juice, soft drinks, milk, tea and infusions was Correlation analyses
not significantly different in obese and normal weight children. Adiposity (fat mass percent of body weight: ρ = − 0.17, P o0.05)
Moreover, EI as well as the energy density from different and some indexes of body size (BMI z-score: ρ = − 0.18, P o0.05)
beverages were not significantly different in the two groups. and body fat distribution (W/H ratio: ρ = − 0.17, P o 0.05) were
Finally, the daily EI covered by beverages was not significantly significantly associated with the hydration level of children (FWR).

Table 1. Physical characteristics of obese and normal weight children

Obese (n = 86) Normal weight (n = 89) P

Age (years) 9.5 (8.5–10.5) 9.5 (8.3–10.5) NS

Sex (M/F) 45/41 40/49 NS
Weight (kg) 45.9 (37.6–55.6) 30.5 (24.8–39.0) o0.001
Height (cm) 141.5 (131.5–149.0) 136.0 (126.0–147.0) o0.03
BMI (kg/m2) 23.0 (21.0–25.1) 16.6 (15.5–18.1) o0.001
BMI z-score 2.94 (2.1–3.5) − 0.22 (−1.2–0.4) o0.001
Waist (cm) 77.0 (70.5–82.5) 58.0 (55.0–63.0) o0.001
Waist-to-height 0.5 (0.5–0.5) 0.4 (0.4–0.4) o0.001
Fat mass (% weight) 30.8 (28.0–35.8) 16.8 (14.6–20.6) o0.001
Fat mass (kg) 13.7 (10.4–18.5) 4.9 (3.9–7.2) o0.001
Fat-free mass (kg) 32.0 (26.1–36.6) 24.3 (20.2–30.5) o0.001
Body H2O (kg) 24.0 (19.3–26.7) 18.4 (14.9–22.9) o0.001
Body H2O (% of body weight ) 51.0 (46.9–53.1) 60.6 (58.0–62.4) o0.001
Abbreviation: BMI, body mass index. Data are shown as median (interquartile range).

Table 2. Energy and nutrient intake and diet energy density of obese and normal weight children

Obese (n = 86) Normal weight (n = 89) P

Energy intake (kcal/day) 1.764 (1.638–2.017) 1.881 (1.680–2.127) NS

Protein (g) 63.7 (54.2–74.8) 61.1 (51.3–69.6) NS
Protein (% energy intake) 14.0 (12.5–15.7) 13.1 (11.6–14.3) o0.01
An/Veg protein 2.2 (1.5–2.8) 1.8 (1.4–2.6) NS
Fat (g) 65.9 (55.6–75.6) 69.0 (56.6–77.8) NS
Fat (% energy intake) 32.4 (28.5–36.8) 33.2 (30.6–36.5) NS
Saturated fat (g) 23.5 (18.8–28.0) 25.1 (19.6–29.1) NS
Saturated fat (% fat) 11.4 (9.7–13.7) 12.1 (9.8–13.6) NS
Monounsaturated fat (g) 31.8 (27.1–36.6) 30.8 (26.6–35.7) NS
Monounsaturated fat (% fat) 15.9 (13.5–18.1) 15.0 (13.0–16.9) NS
Polyunsaturated fat (g) 10.4 (8.0–12.7) 11.1 (8.5–14.1) NS
Polyunsaturated fat (% fat) 4.9 (4.2–5.9) 5.2 (4.3–6.2) NS
Cholesterol (mg) 200.4 (147.0–275.7) 191.7 (126.8–239.5) NS
Carbohydrate (g) 252.3 (216.3–297.4) 273.0 (234.9–315.3) NS
Carbohydrate (% energy intake) 53.2 (48.0–57.2) 53.5 (50.5–57.3) NS
Starch (g) 165.6 (138.0–199.5) 168.0 (140.8–209.4) NS
Simple carbohydrates (g) 86.5 (64.7–102.7) 93.3 (76.6–113.6) o0.03
Simple carbohydrates (% CHO) 17.5 (14.6–21.2) 20.0 (16.3–22.4) o0.03
Total fiber (g) 15.2 (12.0–19.6) 15.3 (12.4–19.0) NS
Energy density [kcal/(g/day)] 1.45 (0.95–1.90) 1.07 (0.87–1.50) o0.01
Data are shown as median (interquartile range).

© 2015 Macmillan Publishers Limited European Journal of Clinical Nutrition (2015) 1 – 6

 Obesity and hydration status in children
C Maffeis et al
4
Table 3.Total fluid and fluid sources intakes, total water intake and water from different sources, energy intake and energy density from fluids in
obese and normal weight children

Obese (n = 86) Normal weight (n = 89) P

Total fluid intake (ml/day) 1180 (859–1550) 1200 (833–1424) NS

Drunk water (ml/day) 833 (592–1125) 717 (500–1022) NS
Fruit juice (ml/day) 0 (0–66.7) 0 (0–118.0) NS
Soft drink (ml/day) 20.0 (0–110.0) 43.3 (0–108.3) NS
Milk (ml/day) 110.0 (18.3–200.0) 120.0 (0–200.0) NS
Tea (ml/day) 0 (0–131.6) 50.0 (0–133.3) NS
Infusion (herb tea) (ml/day) 0 (0–1.3) 0 (0–0.8) NS
Total water intake (fluids+solid foods) (ml/day) 1952 (1505–2275) 1683 (1431–2137) NS
Water from metabolism (ml/day) 230 (212–266) 251 (219–279) NS
Water from metabolism (% of total water intake) 12.9 (10.2–16.6) 13.9 (11.6–16.8) NS
Water from food (ml/day) 418 (304–577) 400 (246–533) NS
Water from food (% of total water intake) 22.7 (16.2–31.4) 23.1 (14.7–29.8) NS
Water from fluids (ml/day) 1,160 (863–1,522) 1,100 (845–1,427) NS
Water from fluids (% of total water intake) 63.3 (54.4–71.9) 64.8 (51.8–74.1) NS
Energy from fluids (kcal/day) 109.8 (75.5–165.7) 129.5 (80.6–183.8) NS
Energy from fluids (% of total daily energy) 6.0 (4.1–8.4) 7.1 (4.5–9.9) NS
Energy density of ingested fluids (kcal/100 ml/day) 0.1 (0–0.1) 0.1 (0–0.1) NS
Juice (kcal/day) 0.0 (0.0–28.3) 0.7 (0.0–50.1) NS
Soft drink (kcal/day) 7.4 (0.0–40.7) 16.0 (0.0–40.0) NS
Milk (kcal/day) 50.6 (8.4–92.2) 55.2 (0.0–92.0) NS
Tea (kcal/day) 0.0 (0–29.6) 11.2 (0–29.9) NS
Infusion (herb tea) (kcal/day) 0 (0-0) 0 (0-0) NS
Data are shown as median (interquartile range).

No significant correlations were found between FWR and

dietary variables, except with diet energy density (ρ = − 0.28,
Po 0.001).
FWR was associated with the energy density of beverages
(ρ = − 0.19, P o0.05) but not with the portion of daily EI covered
by fluids (ρ = 0.07, P = NS) or the variety (number of different
beverages ingested per day) of fluid intakes (ρ = 0.14, P = NS).

ANCOVA model explaining FWR variation

The ANCOVA model did not show a significantly different FWR for
obese and normal weight children, when adjusting for BMI z-score,
energy density of beverages and water intake from fluids
(covariates): obese vs normal weight children had 2.44 (0.44) hg
vs 2.10 (0.50) hg of FWR, P = NS, whereas water intake from fluids
was the only covariate significantly associated with FWR
(B coefficient of co-variation between FWR (hg/day) and water
intake from fluids (hg/day) = 0.47, P o 0.001), showing that the
Figure 1. Sources of fluid intake as percentage of total in obese and
hydration difference between obese and normal weight children
normal weight children.
was mediated by different fluid intakes in the two groups.

DISCUSSION
Table 4. Hydration of obese and normal weight children
Hydration has been scarcely explored in children, especially obese
Obese Normal weight P children. To the best of our knowledge, this is the first study
(n = 86) (n = 89) designed to assess the hydration status in obese children. Our
results show that median urine osmolality is high in both normal
Urine osmolality (mOsm/kg) 741 (589–908) 645 (491–836) o0.05 weight and obese children (75th percentile above the theoretical
FWR (hg/day) 0.8 (0.8–2.8) 2.1 (0.1–4.4) o0.02 upper limit of euhydration in both groups), similar to that
Ratio of children with 29/86 (34) 18/89 (20) o0.05 previously observed in a German population-based cohort of
negative FWR (%) healthy children from the DONALD study, which points to an
Abbreviation: FWR, free water reserve. inadequately low water intake.20 However, the study provides first
evidence of significantly higher osmolality in obese than in normal
weight children, with about one in three obese children vs one in
five normal weight children having a negative FWR. The mean
FWR was associated with total water intake (ρ = 0.46, P o 0.001) level of FWR was 38% lower in obese than in normal weight
and water intake from fluids (ρ = 0.45, P o 0.001) but not with children (Table 4). Consistently, a significant relationship between
water intake from food (ρ = 0.10, P = NS) or metabolic processes adiposity expressed as fat mass (% of body weight) as well as
(ρ = 0.09, P = NS). adiposity indexes (BMI z-score and waist/height ratio) and

European Journal of Clinical Nutrition (2015) 1 – 6 © 2015 Macmillan Publishers Limited

 Obesity and hydration status in children
C Maffeis et al
5
hydration was also found. The lower FWR associated with obesity with countries where sweetened beverages have been associated
was explained by lower z-BMI adjusted water intake from fluids in with obesity. In fact, although in our cohort plain water covered
obese than in normal weight children. In fact, even if absolute 65–70% of total fluid intake, similarly to that reported in other
water intake from fluids did not differ in the two groups (Table 3), European countries,30 it covered only 29% of fluid intake in US
BMI z-score adjusted water intake from fluids explained the children of NHANES.28
different FWR in obese and non-obese children in ANCOVA, which It could be argued that the fluid intake of obese children is
means that obese children are less hydrated because they have a underreported. In fact, screening for children falling outside the
lower water intake than their peers, when BMI is taken into 1–2.40 range of EI/BMR based on Goldberg’s method is far from
account. The higher chance of having hypo-hydration shown by being sufficient to exclude all misreporting subjects, and
obese children exposes them to potential clinical consequences. nutritional data of the present study are to be interpreted with
Children with negative FWR inevitably have a urine osmolality caution. However, some important arguments suggest that at
over the 830 mosm/kg cutoff used to calculate the obligatory least fluids should not be underreported. In fact, underreporting of
urine volume; hence, they avoid negative water balance and fluids by obese children would be in contrast with the fact that
consequent water depletion at the expense of a very high level of FWR was lower and urine osmolality higher in obese than in
urine concentration, and they do not have any ‘free’ water normal weight children and that lower fluid intake explained the
available to the body, besides that needed for osmoles excretion. lower FWR in obese children. Moreover, we assessed the
This may have negative consequences for urinary tract health.21–22 β-coefficient of correlation between BMI-adjusted fluid intake
Moreover, hydration status does not have a negligible impact on and urine volume, urine osmolality and FWR in obese and normal
mental performance and cognition.23 As obese children are weight children separately, by general linear models, obtaining
reported to have lower visuospatial organization and general similar values for the two groups (0.75 vs 0.75, − 0.32 vs − 0.54,
mental ability than normal weight children, an adequate hydration 0.56 vs 0.62, respectively), which rules out the above-mentioned
level should be pursued in these children.24 Assessment of potential bias.
hydration status in obese children should be part of family The strength of the present study is given by the following:
counseling for fluid intake. Of note, a simple urine color scale for (i) the 48-h urine collection design, which is a demanding task,
the assessment of urine osmolality by children has recently been especially for children; (ii) the measure of hydration and diet
validated and could help promote a higher water intake.25 performed with suitable and reasonably accurate methods for the
Interestingly, the median of total fluid intake as well as fluid study purposes and the age of the participants.13,18
intake through beverages for both obese and normal weight Potential limitations of this study are as follows: (i) the cross-
children is ≈15–20% lower than EFSA (European Food Standards sectional design that does not allow for exploring the cause–effect
Agency) adequate intakes (girls: 1.9 liters/day total fluids and relationship between variables but just their degree of
1.3–1.5 liters/day fluids from beverages; boys: 2.1 liters/day total association; (ii) the recruitment area, limited to Italy, that does
fluids and 1.5–1.7 fluids from beverages).26 These data are not allow to extend the results obtained to other populations;
consistent with those reported by Stahl et al.6 who found for (iii) the narrow age range, limited to pre-pubertal children, that
717 German children from the DONALD study a total water intake does not allow extending the results obtained to younger children
median of 1647 ml/day in boys and 1483 ml/day in girls, those or adolescents; (iv) potential misreporting of nutritional data, as
reported by Bellisle et al.27 who found an average fluid intake of discussed above, despite screening for surely misreporting
1046 ml/day in a sample of 6- to 11-year-old French children, and participants.
those reported by Drewnowski et al.28 who found that only 15% of In conclusion, the results of this study showed that water intake
children met the recommended intake of water in the NHANES was the main determinant of hydration status in children and that
sample.28 In our study, water intake from fluids was the main obese children were less hydrated than normal weight children.
contributor to total water intake in both obese and normal weight Hydration status was not satisfactory in 34% of obese and 20% of
children (about 65%), whereas water intake from food and normal weight children. On the basis of these findings, it would be
metabolic processes did not have a relevant impact on hydration. interesting to test whether an intervention to increase water
In contrast, in the population-based sample of German children intake and hydration status could contribute to a positive change
from the DONALD study, both water from fluids and water from in children’s body composition.
foods contributed to the variance of total water intake and,
consequently, of the hydration level, probably because of a higher
contribution of water from foods to total water intake in this CONFLICT OF INTEREST
population compared with our sample (about 50 vs 23%, The authors declare no conflict of interest.
respectively).6
Among the different beverages, water was the major
contributor to total fluid intake in obese and normal weight ACKNOWLEDGEMENTS
children, making upon average two-thirds of the total amount, This work was funded by a grant of the Department of Life and Reproduction
similarly to what previously reported.25 No significant differences Sciences of the University of Verona and partly by a grant from the European
were found between the two groups of children in the proportion Hydration Institute (EHI). The EHI did not directly contribute to the study design, the
of fluid intake covered by fruit juice, soft drinks, milk, tea and collection, analysis and interpretation of data or in the writing of the manuscript. We
herbal infusions (Figure 1). Variety of beverage intake did not thank Ron Maughan for his helpful comments and suggestions.
affect total water intake from fluids and, by implication, hydration
status. Finally, the energy density of beverages did not have an
independent effect on hydration when total beverage intake and REFERENCES
BMI z-scores were taken into account in the analysis. In fact, the 1 Franks PW, Hanson RL, Knowler WC, Sievers ML, Bennett PH, Looker HC. Child-
amount of calories ingested with beverages was not statistically hood obesity, other cardiovascular risk factors, and premature death. N Engl J Med
2010; 362: 485–493.
different in the two groups. In other words, this sample of obese
2 Juonala M, Magnussen CG, Berenson GS, Venn A, Burns TL, Sabin MA et al.
children did not ingest more calories with beverages than lean Childhood adiposity, adult adiposity, and cardiovascular risk factors. N Engl J Med
children. This is in contrast with previous reports of association 2011; 365: 1876–1885.
between sweetened beverages and childhood obesity in western 3 Maffeis C, Moghetti P, Grezzani A, Clementi M, Gaudino R, Tatò L. Insulin resis-
countries29 and could be explained by the low average contribu- tance and the persistence of obesity from childhood into adulthood. J Clin
tion of soft drinks to total fluid intake in our sample compared Endocrinol Metab 2002; 87: 71–76.

© 2015 Macmillan Publishers Limited European Journal of Clinical Nutrition (2015) 1 – 6

 Obesity and hydration status in children
C Maffeis et al
6
4 Ambrosini GL, Emmett PM, Northstone K, Howe LD, Tilling K, Jebb SA. 17 Remer T, Neubert A, Maser-Gluth C. Anthropometry-based reference values for
Identification of a dietary pattern prospectively associated with increased 24- h urinary creatinine excretion during growth and their use in endocrine and
adiposity during childhood and adolescence. Int J Obes 2012; 36: 1299–1305. nutritional research. Am J Clin Nutr 2002; 75: 561–569.
5 Pérez-Escamilla R, Obbagy JE, Altman JM, Essery EV, McGrane MM, Wong YP et al. 18 Winberg J. Determination of renal concentration capacity in infants and children
Dietary energy density and body weight in adults and children: a systematic without renal disease. Acta Pediatr 1959; 48: 318–328.
review. J Acad Nutr Diet 2012; 112: 671–684. 19 Edelmann CM, Barnett HL, Stark H, Boichis H, Rodriguez-Soriano J. A standardized
6 Stahl A, Kroke A, Bolzenius K, Manz F. Relation between hydration status in test of renal concentrating capacity in children. Arch Pediatr Adolesc Med 1967;
children and their dietary profile – results from the DONALD study. Eur J Clin Nutr 114: 639–644.
2007; 61: 1386–1392. 20 Manz F, Wentz A, Sichert-Hellert W. The most essential nutrient: defining the
7 Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for adequate intake of water. J Pediatr 2002; 141: 587–592.
child overweight and obesity worldwide: international survey. BMJ 2000; 320: 21 Sontrop JM, Dixon SN, Garg AX, Buendia-Jimenez I, Dohein O, Huang SH,
1240–1243. Clark WF. Association between water intake, chronic kidney disease, and cardio-
8 Zoppi G, Bressan F, Luciano A. Height and weight reference charts for children vascular disease: a cross-sectional analysis of NHANES data. Am J Nephrol 2013;
aged 2-18 years from Verona, Italy. Eur J Clin Nutr 1996; 50: 462–468. 37: 434–442.
9 Maffeis C, Banzato C, Talamini G. Obesity Study Group of the Italian Society 22 Michaud DS, Spiegelman D, Clinton SK, Rimm EB, Curhan GC, Willett WC et al.
of Pediatric Endocrinology and Diabetology. Waist-to-height ratio, a useful index Fluid intake and the risk of bladder cancer in men. N Eng J Med 1999; 340:
to identify high metabolic risk in overweight children. J Pediatr 2008; 152: 1390–1397.
207–213. 23 Masento NA, Golightly M, Field DT, Butler LT, van Reekum CM. Effects of hydration
10 Katzmarzyk PT, Shen W, Baxter-Jones A, Bell JD, Butte NF, Demerath EW et al. status on cognitive performance and mood. Br J Nutr 2014; 111: 1841–1852.
Adiposity in children and adolescents: correlates and clinical consequences of fat 24 Li Y, Dai Q, Jackson JC, Zhang J. Overweight is associated with decreased
stored in specific body depots. Pediatr Obes 2012; 7: e42–e61. cognitive functioning among school-age children and adolescents. Obesity (Silver
11 Tanner JM, Whitehouse RH, Takaishi M. Standards for birth to maturity for height, Spring) 2008; 16: 1809–1815.
weight, height and weight velocity: British children 1965. Arch Dis Child 1966; 41: 25 Kavouras SA, Johnson EC, Bougatsas D, Arnaoutis G, Panagiotakos DB, Perrier E
454–495. et al. Validation of a urine color scale for assessment of urine osmolality in healthy
12 Barreira TV, Staiano AE, Katzmarzyk PT. Validity assessment of a portable children. Eur J Nutr; epub ahead of print 24 April 2015; PMID: 25905541.
bioimpedance scale to estimate body fat percentage in white and African 26 EFSA (European Food Standards Agency). Scientific opinion on dietary reference
American children and adolescents. Pediatr Obes 2013; 8: e29–e32. values for water. EFSA J 2010; 8: 1459.
13 Buzzard M. 24-Hour Dietary Recall and Food Record Method. Nutritional 27 Bellisle F, Thornton SN, Hébel P, Denizeau M, Tahiri M. A study of fluid intake
Epidemiology. Walter Willet: Oxford, UK, 1998. from beverages in a sample of healthy French children, adolescents and adults.
14 Livingstone MB, Prentice AM, Coward WA, Strain JJ, Black AE, Davies PS et al. Eur J Clin Nutr 2010; 64: 350–355.
Validation of estimates of energy intake by weighed dietary record and diet 28 Drewnowski A, Rehm CD, Constant F. Water and beverage consumption
history in children and adolescents. Am J Clin Nutr 1992; 56: 29–35. among children 4-13y in the United States: analysis of 2005-2010 NHANES data.
15 National Institute for the Research in Food and Nutrition. Food Composition Tables. Nutr J 2013; 12: 85.
EDRA Publ: Milano, Italy, 2007. 29 Morgan RE. Does consumption of high-fructose corn syrup beverages cause
16 Black AE. Critical evaluation of energy intake using the Goldberg cut-off for obesity in children? Pediatr Obes 2013; 8: 249–254.
energy intake: basal metabolic rate. a practical guide to its calculation, use and 30 Ozen AE, Bibiloni MD, Pons A, Tur JA. Fluid intake from beverages across age
limitations. Int J Obes 2000; 24: 1119–1130. groups: a systematic review. J Hum Nutr Diet 28: 417–442.

European Journal of Clinical Nutrition (2015) 1 – 6 © 2015 Macmillan Publishers Limited