1    DETECTION OF EUHYDRATION IN HUMANS FROM THE 1  DIURESIS RESPONSE TO A WATER LOAD 2  (PRE-PRINT: This manuscript was submitted 3  to a peer-reviewed journal on July 29, 2011) 4  Authors: Catalina Capitán-Jiménez & Luis Fernando Aragón-Vargas 5  Corresponding author: 6  7  Catalina Capitán 8  75 m Sur Escuela La Pitahaya 9  Cartago, COSTA RICA 10  Phone: +506 2591-1889 11  Fax: +506 2225-0749 12  e-mail: ktaucr@gmail.com) 13   14  Affiliations: both authors are with the 15  Human Movement Science Laboratory 16  School of Physical Education and Sports 17  University of Costa Rica 18  11-501-2060 Montes de Oca 19  San José, Costa Rica 20  21  E-mail for Luis F. Aragon-Vargas: luis.aragon@ucr.ac.cr 22  23  24  2    ABSTRACT 1  Aim: to calculate the minimum amount of water to be ingested in order to find significant 2  differences in one-hour urine volume between euhydrated and dehydrated humans. 3  Methods: Five participants (22.6±2.9 years old, 63.70±13.18 kg; mean±standard deviation) 4  were evaluated following an overnight fast, on eight different non-consecutive days. For the 5  euhydration condition (EuA, EuB, EuC and EuD), they remained seated for 45 minutes. For the 6  dehydration condition (DeA, DeB, DeC and DeD), they exercised intermittently at 33±4° C, 7  65±6% relative humidity until they were dehydrated by 1% of body mass (BM). The order of 8  treatments was randomized. Next, they ingested a water volume equivalent to 0.5% BM, 0.72% 9  BM, 1.07% BM or 1.43% BM in 30 minutes, under both the euhydration and the dehydration 10  conditions. Urine volumes were collected 0, 30, and 60 minutes after water ingestion. 11  Results: baseline values were consistent among conditions (p>0.05), and there was no 12  difference in water intake volume between the euhydration and the dehydration conditions 13  (p>0.05). There was a clear association between the volume of water intake and urine volume 14  (R2= 0.64, R2a= 0.58; p= 0.001); in addition, this tendency was different between euhydration 15  and dehydration (interaction p=0.005). Finally, ingestion of water equivalent to 1.07% BM or 16  more resulted in a 95% CI for the urine volume difference between euhydration and dehydration 17  greater than 100 mL. 18  Conclusion: the minimum volume of water to be ingested to detect a 100 mL difference in one-19  hour urine volume between euhydrated and dehydrated humans is 1.07% BM. 20  Key words: water intake, urine, acute hydration status, euhydration, dehydration, fluid balance 21  3    Introduction. 1  Dehydration from sweat loss places athletes at a disadvantage against their competitors; as 2  little as 2% of body mass loss may impair mental focus and some motor skills, and will increase 3  heart rate and overall demand on the cardiovascular system which, in turn, may have a negative 4  impact on sports performance (Shirreffs, 2003; Grandjean and Campbell, 2004; Armstrong et al 5  2007). It is therefore desirable to avoid dehydration during training and competition. Because 6  previous, uncorrected dehydration can have a negative impact on performance as well 7  (Armstrong, Costill and Fink, 1985), it is also important to assess acute hydration status before 8  a training session or competition and to correct any existing hypohydration. 9  Normally, it is not easy to accurately and reliably measure acute hydration status, and it 10  becomes more difficult with athletes who have larger fluid losses and frequently train twice a 11  day (Cheuvront and Sawka, 2005). Some practical recommendations to assess acute hydration 12  status with athletes include using urine color and/or urine specific gravity; it is also common to 13  register body mass before and after exercise to estimate sweat loss (Cheuvront and Sawka, 14  2005; Oppliger and Bartok, 2002). The weight loss method can only give a relative amount of 15  dehydration occurring over a period of time; the urine values are subject to wide variations 16  under dynamic conditions, and the results can be misleading (Armstrong, 2007). 17  A recent study from Capitán and Aragón-Vargas (2009) proposed the evaluation of acute 18  hydration status from the urine response to a standard water load equivalent to 1.43% of body 19  mass (BM). Urine volume was monitored over 5 hours, in an attempt to verify whether urine 20  elimination was different depending on pre-existing, controlled dehydration levels. Results from 21  the study showed there is a clear difference in five-hour urine volume in response to the water 22  load, between euhydration (1236.8 ± 489.4 mL), and dehydration (375.3 ± 170.2 mL, 235.9 ± 23  66.0 mL, and 261.7 ± 51.8 mL at 1, 2, and 3% BM, respectively, p = 0.001). In addition, the 24  difference could be detected after only one hour of monitoring (Capitán Jiménez and Aragón 25  4    Vargas, 2009); based on these results, the authors proposed a test, which was later shown to 1  be both valid and reliable (Capitán-Jiménez and Aragón-Vargas, 2010). 2  The urine volume differences reported are consistent with a normal renal response: reducing 3  urine production during dehydration, but increasing it when a well hydrated person ingests water 4  (Sawka et al. 2007; Valtin and Schafer,1995). The aforementioned method is promising, but in 5  order for it to be practical, the volume of water ingested should be close to the urine volume 6  eliminated in a reasonable time (i.e., one hour) in the euhydrated condition. This would allow for 7  the person to initiate exercise without too much fluid excess in his/her body. 8  Therefore, the purpose of this study was to calculate the minimum water volume that must be 9  administered to a person to identify significant differences in one-hour urine volume between a 10  euhydrated and a dehydrated state. 11  Materials and methods 12  Participants. Four young females and a young male (22.6 ± 2.9 y.o.; mean ± standard 13  deviation) agreed to participate in this study and signed an informed consent. They were 14  healthy, physically active (exercised at least four times a week), had no known heart, renal or 15  endocrine problems, had never suffered heat illnesses, and at the time of the study were not 16  ingesting any diuretics. The study was approved by the University’s Ethics and Science 17  Committee. 18  Study design. This was an experimental study with two conditions: Euhydration (Eu) and 19  Dehydration (De). In addition, four water ingestion volumes (water loads) were pre-determined 20  (0.5% 0.72%, 1.07%, and 1.43% of body mass, equivalent to 350, 500, 750 and 1000 mL, 21  respectively, for a 70-kg individual); these volumes were the same for both conditions. Each 22  participant completed a total of eight treatments in a factorial block design with two conditions 23  by four water loads (2x4) (each person is one block) (see Fig. 1). The order of treatments was 24  randomized. 25  26  5    Figure 1. Factorial block design (2x4). 1   2   3   4  Procedures 5  Preparation. Each participant reported to the laboratory at 7 a.m. on eight different non-6  consecutive days, following an overnight fast (a minimum of 10 hours without solids or liquids). 7  Upon arrival, they provided a urine sample which was analyzed for urine specific gravity (USG) 8  with a manual refractometer (ATAGO®, model URC – Ne, d 1.000-1.050) to estimate initial 9  hydration status. This urine sample was discarded. 10  After completely emptying their bladders, participants were weighed nude to the nearest 10 11  grams on a calibrated scale (e-Accura®, model DSB291). This fasting body mass (FastBM) was 12  used to calculate the fluid volume to be ingested by each individual. Next, they ingested a 13  standardized 750-kcal breakfast (24.6% fat, 20.7% protein, and 54.7% carbohydrate, including 14  250 mL of liquid and 1500 mg sodium), and proceeded to rest for 30 minutes. 15  6    Exercise. Following the rest period all participants were weighed nude and dry (baseline body 1  mass, BBM); those whose protocol for the day called for dehydration started intermittent 2  exercise: 15 minutes of stationary cycling, 15 minutes of jogging on a treadmill, and again 15 3  minutes of stationary cycling, as long as necessary to achieve a dehydration equivalent to 1% 4  BBM; body mass was measured at the end of every 15 minutes with participants nude and dry. 5  Participants exercised in a controlled environment chamber kept at 33 ± 4°C dry bulb and 65 ± 6  6% relative humidity, at a moderate-to-high intensity (75% to 80% of maximum heart rate) 7  monitored with a Polar® heart rate monitor, model A1; maximum heart rate was estimated from 8  220 – age. 9  When their individual protocol did not require exercise (the four Euhydration sessions), 10  participants remained at rest outside the chamber for 45 minutes. 11  Once the exercise or prolonged resting period was over, all participants took a cold shower. 12  They were instructed not to drink any fluid, and to completely empty their bladders in a 750 mL 13  plastic container. This urine was weighed on a food scale (OHAUS® Compact Scales, model 14  CS2000), to the nearest 1 g. All participants were weighed again nude and dry at this point to 15  obtain post-exercise body mass (PEBM). 16  Water load. After showering and weighing out, each participant ingested a volume of water 17  equivalent to 0.5%, 0.72%, 1.07% or 1.43% of FastBM, according to his or her particular 18  protocol for the day. This water load was divided into three equal volumes for ingestion, with a 19  10-minute break after the first and second aliquot. 20  Urine collection. Participants were instructed to completely empty their bladders into labeled 21  plastic containers immediately upon completion of water ingestion (time 0) and after 30 and 60 22  minutes; they remained at rest for this urine collection period. The containers were weighed to 23  the nearest 1 g, and the volume was recorded assuming 1 g is equivalent to 1 mL. 24  7    Statistical analysis. Descriptive statistics (mean and standard deviation) were calculated for 1  age, body mass, and height in order to characterize the participants. These and the other 2  variables were checked for normality. 3  Inferential statistics were calculated using JMP version 7. To verify that baseline values were 4  the same, a one-way, repeated measures analysis of variance (ANOVA) was performed for 5  each of two reference variables: USG and FastBM. In addition, a two-way, two-factor repeated-6  measures ANOVA (2 conditions X 4 water loads) was performed on the actual volume of water 7  ingested for each water load. 8  A multiple regression analysis was performed in order to understand the main effects of the 9  condition (acute hydration status: euhydration or dehydration), as well as acute hydration status 10  interacting with water loads (interaction), on urine volume. For this regression, one-hour urine 11  volume was used as the dependent variable; predictor variables were water load and acute 12  hydration status (0 for dehydration and 1 for euhydration). In this model, fasting body mass 13  (FastBM) and initial USG were used as covariates since they are pre-existing conditions which 14  may influence urine volume. Individuals were included as blocks to minimize the effect of the 15  error due to natural biological variability among people. This is the conceptually correct, full 16  model which must be evaluated. 17  Another multiple regression analysis was performed but this time without including the 18  individuals’ blocks as a predictor variable (the simplified model). This was done in order to 19  obtain operational results, that is, to obtain a specific water load (as a %BM) where differences 20  could be identified between euhydration and dehydration. With the regression equation obtained 21  from this analysis, 95% confidence intervals (95%CI) were calculated for urine volumes for both 22  the euhydrated and the dehydrated conditions. 23  To establish the minimum difference in one-hour urine volume between conditions, the upper 24  limit of the 95%CI for dehydration was subtracted from the lower limit of the 95%CI for 25  euhydration, according to the formula: 26  8    [Eq. 1] UpperLimitCIDeLowerLimitCIEuDifferenceMinimumExpected %95%95 −= 1  Finally, a basic multiple regression analysis was performed using only water load, condition, and 2  their interaction, for comparison purposes (the basic model). 3  Results 4  All participants completed the study. Their age = 22.6 ± 2.9 y.o.; height = 1.63 ± 0.06 m; body 5  weight = 63.7 ± 13.18 kg (mean ± standard deviation). 6  Baseline conditions. No differences in body mass or urine specific gravity were found among 7  treatments (see Table 1). Actual water intake volumes for each of the water loads were not 8  different when comparing euhydration and dehydration (see Table 2). 9   10  Table 1. Baseline values. 11   12   13   14   15   16   17   18   19   20   21  FastBM: fasting body mass. USG: urine specific gravity.  22  * p = 0.98 among treatments 23  + p = 0.99 among treatments 24   25  Treatment FastBM (kg) * ( X ± SD) USG+ ( X ± SD) EuA (0.5%) 63.62 ± 14.60 1.02 ± 0.01 EuB (0.72%) 63.65 ± 14.46 1.02 ± 0.01 EuC (1.07%) 63.59 ± 14.51 1.02 ± 0.01 EuD (1.43%) 63.78 ± 14.68 1.02 ± 0.01 DeA (0.5%) 63.73 ± 14.47 1.02 ± 0.01 DeB (0.72%) 63.62 ± 14.57 1.02 ± 0.01 DeC (1.07%) 63.97 ± 14.77 1.02 ± 0.01 DeD (1.43%) 63.65 ± 14.39 1.02 ± 0.01 9    Table 2. Actual water intake for each treatment. 1  2  3  4  5  6  For all water loads, p > 0.05 euhydration vs. dehydration. 7  8  Multiple regression analysis. Table 3 shows the results of the multiple regression models at 9  each one of three stages of complexity. There is a significant interaction between the two 10  predictor variables (water load and condition) regardless of the model. There is also a significant 11  main effect of water load on urine volume. 12  The regression equation from Table 3 (simplified model), which is able to explain 58% of the 13  variance (R2 a= 0.58), is as follows: 14  [Eq. 2] )*51.0()*66.60()*99.233(31.8659 FastBMcondWLVol +−+= 15  )](*)(*)88.155[()*26.8559( condWLUSGi +− 16  where 17  Vol is the one‐hour urine volume; 18  WL, the water load, is the water intake volume expressed as a percentage of body mass 19  (%FastBM); 20  cond is the numerical value for each condition: use 0 for dehydration or 1 for euhydration; 21  FastBM is the fasting body mass; and 22  USGi is the urine specific gravity for the initial sample obtained upon arrival. 23  Water load (%FastBM) Euhydration (mL) X ± SD Dehydration (mL) X ± SD 0.50% 318.11 ± 73.11 318.63 ± 72.33 0.72% 458.29 ± 104.09 458.08 ± 104.88 1.07% 1.43% 680.37 ± 155.27 912.05 ± 209.90 684.00 ± 158.03 910.20 ± 205.80 10     1  Table 3. Secuence of regression models. 2  Model 1 (Basic) Model 2 (Simplified) (with covariates) Model 3 (Complete) Parameters Estimate Probability Estimate Probability Estimate Probability Intercept -24.47 0.6702 8659.31 0.0111 -1531.49 0.7843 Individual - - - - 1588.54 0.1130 Water load 226,82 0.0003 233.99 0.0001 222.09 0.0001 Condition(0) FastBM Initial USG -60.84 - - 0.2927 - - -60.66 0.51 -8559.27 0.2428 0.7868 0.0124 -51.62 104.63 -5057.64 0.2941 0.0350 0.2826 Water load * Condition 153,29 0.0111 155.88 0.0046 149.63 0.0043 11    Both USG and FastBM were included as covariates in this regression equation in order to 1  reduce error, considering that these variables were not controlled for experimentally. 2  Figure 2 represents the simplified regression and shows how the slopes corresponding to the 3  euhydration and dehydration conditions are different from each other (urine volume changes in 4  response to different water loads are different). 5   6  Figure 2. Water load by condition interaction 7  8  Table 4 shows one-hour urine volume 95%CIs for the euhydration and dehydration conditions, 9  corresponding to various arbitrary water loads, as calculated from the simplified multiple 10  regression model (Eq. 2). Expected minimum differences are also shown, as estimated using 11  Equation 1. According to Table 4, the water load should be at least 1.07%FastBM to obtain a 12  minimum difference of 100 mL between a euhydrated and a dehydrated condition. 13  14  12    Table 4. Ninety-five percent confidence intervals for one-hour urine volume. 1   2  Dehydration Euhydration Urine volume 95%CI (mL) Urine volume 95%CI (mL) Lower limit Upper limit Lower limit Upper limit Water load (%FastBM) Minimum expected difference (mL) 0.50% 0.85% 0.00 42.75 78.07 149.41 93.76 187.27 246.50 295.89 15.69 37.86 0.93% 50.09 154.23 218.81 322.95 64.58 1.05% 56.48 166.26 261.91 371.79 95.65 1.07% 57.15 169.35 270.03 382.37 100.68 1.20% 57.65 189.93 312.52 445.04 122.59 1.43% 51.13 230.81 374.45 554.51 143.64  3  Discussion 4  The purpose of this study was to determine the minimum amount of water to be ingested by an 5  individual in order to find significant differences between euhydration and dehydration in the 6  urine volume excreted in one hour. With an R2= 0.64 and R2a = 0.58 (p = 0.001), a strong 7  association has been identified between urine volume and predictor variables (acute hydration 8  13    status and water load). In addition, if a water load equivalent to 1.07% of body mass is 1  administered to an individual, a difference of at least 100 mL will be found with 95% certainty. 2  A previous study by Capitán Jiménez and Aragón Vargas (2009) showed a clear difference in 3  urine volume between individuals in a euhydrated and a dehydrated state, using a water load 4  equivalent to 1.43%BM. Nevertheless, they stated the need to identify a smaller water load for 5  the test, in order to reduce the amount of extra fluid remaining in the body after one hour in the 6  case of euhydrated individuals. By reducing the water load to 1.07%BM, the remaining fluid 7  would be substantially reduced: for instance, a 70-kg individual would ingest 750 mL (as 8  opposed to 1 liter), of which approximately 438 mL would be eliminated in one hour. 9  To estimate the lowest possible water load in this study, a multiple regression analysis was 10  used. The complete model from Table 3 included all possible variables which could influence 11  one-hour urine volume in response to a water load. For this analysis, individuals were included 12  as blocks, expecting significant differences among participants. This, however, did not happen, 13  as the global effect of individuals was not significant (Table 3). A second, simplified model was 14  tested where individuals were not included as blocks, a model which enables the authors to 15  generalize the present results. 16  It was with this second multiple regression analysis that a water load could be estimated to 17  result in a minimum expected difference of 100 mL of urine between euhydration and 18  dehydration, as calculated from the 95% confidence intervals. This 100 mL difference between 19  conditions had been defined a priori taking into consideration the fact that normal urine 20  production in one hour is approximately 60 mL; if a water load is ingested, this volume would be 21  expected to increase (Valtin and Schafer, 1995). 22  The simplified regression equation (Eq. 2) dictates mathematically that for each additional 23  percentage point that the water load is increased, urine volume would increase by (233.99 + 24  155.88 - 60.66 mL) = 329.21 mL for a euhydrated individual, or 233.99 mL for a dehydrated 25  one. In practice, however, the present results should not be extrapolated beyond the extreme 26  14    water loads tested here, namely, 0.50% and 1.43% BM, as urine volume behavior outside these 1  limits is currently unknown. 2  The results from this study give quantitative support to the intuitive statement that urine volume 3  is significantly different between individuals in a euhydrated and a dehydrated state, as already 4  confirmed in a previous study (Capitán and Aragón-Vargas, 2009); furthermore, the minimum 5  water load to be ingested to detect those differences with collection of urine volume for one hour 6  is identified. 7  In conclusion, the minimum water load that must be ingested by an individual to detect a 100 8  mL difference in one-hour urine volume between euhydration and dehydration is the equivalent 9  of 1.07% of body mass (750 mL for a 70-kg person). 10  Acknowledgments 11  The authors wish to thank María Isabel González Lutz, M.S., Professor, School of Statistics, 12  University of Costa Rica, for her valuable assistance with study design and statistical analysis, 13  and Walter Salazar, Professor, School of Physical Education, University of Costa Rica, for his 14  valuable input on the manuscript. 15  This study was supported by the Gatorade Sports Science Institute® through research projects 16  VI-245-A4-303 and VI-245-B0-315 at the University of Costa Rica. 17  18  References. 19  Armstrong, L. 2007. Assesing hydration status: The elusive gold standard. Journal of 20  the American College of Nutrition, 26 (5), 575S – 584S 21  Armstrong, L., Casa, D., Millard-Stafford, M., Moran, D., Pyne, S.and Roberts, W. 2007. 22  American College of Sports Medicine Position Stand: Exertional Heat Illness during 23  training and competition. Medicine & Science in Sports & Exercise, 39(3), 556 – 572. 24  15    Armstrong, L., Costill, D. and Fink, W. 1985. Influence of diuretic-induced dehydration 1  on competitive running performance. Medicine & Science in Sports & Exercise 17(4): 2  456 – 461. 3  Capitán, C. & Aragón-Vargas, LF. 2009. A practical method to assess euhydration. 4  Medicine & Science in Sports & Exercise 41 (5): 11s. 5  Capitán Jiménez, C. and Aragón Vargas, L. 2009. Método sencillo para comprobar 6  euhidratación. Pensar en Movimiento: Revista de Ciencias del Ejercicio y la Salud 7  7(1):23-31. 8  Capitán-Jiménez, C. and Aragón-Vargas, L. 2010. Elimination of urine in response to 9  water intake is consistent in well-hydrated individuals. MH Salud 7(2): 1-9. Available 10  from:http://www.una.ac.cr/mhsalud/index.php?option=com_booklibrary&task=view&id=511  0&catid=105&Itemid=41[ Accesed 28 june, 2011] 12  Cheuvront, S. and Sawka, M. 2005. Hydration Assessment of Athletes. Sports Science 13  Exchange # 97, 18 (2). 14  Grandjean, A. and Campbell, S. 2004. Hydration: Fluids for Life. ILSI North America. 15  Oppliger, R. and Bartok, C. 2002. Hydration testing for athletes. Sports Medicine, 32, 16  959- 971. 17  Sawka, M., Burke, L., Eichner, E., Maughan, R., Montain, S., and Stachenfeld, N. 2007. 18  American College of Sports Medicine Position Stand: Exercise and fluid replacement. 19  Medicine & Science in Sport & Exercise 39(2), 377 – 390. 20  Shirreffs, S. 2003. Markers of hydration status. European Journal of Clinical Nutrition, 21  57 (Suppl 2): S6- S9. 22  16    Valtin, H. and Schafer, J. 1995. Renal Function. Little, Brown, and Co. Boston, MA, 1  USA. 2