Intravascular volume loading with crystalloid solution is often practiced to avoid hypotension during the induction of anesthesia. Given that infused crystalloid is usually equilibrated with the interstitial fluid within several tens of minutes, decreasing the fluid transfer from the intravascular space to the interstitial space may increase the effectiveness of this treatment. Volume kinetic studies demonstrated that factors that may decrease the shift of crystalloid fluid to the interstitial space include intravascular hypovolemia1 and a decrease of mean arterial blood pressure (MAP) during anesthesia.2,3
Despite the observations described above, the effect of preoperative dehydration on the interstitial distribution of crystalloid solution is unclear. Volume kinetic studies are usually performed for subjects after overnight fasting,1 – 4 and oral fluid intake is still restricted for 6 to 8 hours before surgery in some countries.5,6 Contrary to the older concept that plasma volume is maintained at the expense of the interstitium, simple dehydration (i.e., loss of water alone) leads to an approximately proportional reduction of interstitial fluid volume and plasma volume.7 The considerable tendency of dehydrated interstitium to absorb fluid may reduce the effectiveness of infused crystalloid solution to achieve plasma volume expansion.7
In the present study, we used bioelectrical impedance spectroscopy to investigate crystalloid redistribution into the interstitial space during general anesthesia in patients undergoing minor surgery. The extracellular fluid volume (ECF) estimated by bioelectrical impedance is relatively insensitive to plasma volume change, and thus a change in ECF mainly reflects interstitial fluid volume change.1,8 The aim of this study was to test the hypothesis that preoperative dehydration after overnight fasting affects crystalloid redistribution into the interstitial space and thus the magnitude of hypotension during general anesthesia.
The study was approved by the Ethics Review Board of Hyogo College of Medicine and written informed consent was obtained from each patient after explanation of the study. We studied consecutive ASA physical status I or II patients aged 20 to 65 years who were scheduled for tympanoplasty under general anesthesia from April 2009 to December 2009. Exclusion criteria included a history of cardiac, pulmonary, liver, or renal disease, hypertension, or medication with diuretics. Patients weighing >80 kg were also excluded because the required fluid infusion rate would have exceeded the maximum fluid infusion rate at which the fluid infusion pump could function (i.e., 1200 mL/h).
All patients fasted from midnight. According to the routine procedure used in our institution, patients who were scheduled for induction of anesthesia starting at 9:00 AM (i.e., the first case on the operative day) did not receive preoperative fluid loading. Patients scheduled for the second or later interventions on the operative day received fluid loading in the ward with acetated Ringer solution at a rate of approximately 1.5 mL · kg−1 · h−1 starting at 9:00 to 10:00 AM.
No patients were premedicated. The patients voided in the ward 30 minutes before entering the operating room. Antithromboembolic stockings were fitted to both lower legs in the ward. After the patient arrived at the operating room, a vein on one arm was cannulated for those patients who were the first case for the day, whereas the preexisting venous cannula was used for the subsequent patients who were to be operated on. Heart rate and noninvasive blood pressure were measured at 2.5-minute intervals on the ipsilateral side to the venous cannula. After 5 minutes of oxygen administration, the patients were given fentanyl 2 μg/kg IV over 2 minutes with no fluid administration other than that required to administer the anesthetic. The patients were then given propofol 2 mg/kg IV over 6 minutes (20 mg · kg−1 · h−1) using a syringe infusion pump to minimize cardiac depression induced by rapid propofol infusion.9 After completion of the propofol infusion, a laryngeal mask was inserted and anesthesia was maintained with sevoflurane (1.0%–1.5%) and oxygen (50%) in air with positive pressure ventilation in a circle system with a fresh gas flow of 3 L/min, together with a continuous infusion of remifentanil at a rate of 0.10 to 0.15 μg · kg−1 · min−1. A rewarming cover blanket system and fluid warmers were used to maintain the body temperature of patients above 36°C during surgery.
Coinciding with the start of propofol administration, 15 mL/kg acetated Ringer solution was infused IV over 60 minutes followed by 1 mL/kg acetated Ringer solution over the next 30 minutes. Thus, the 90-minute period encompassing the administration of propofol and acetated Ringer solution was defined as the study period. Hypotension, defined as a systolic blood pressure ≤75 mm Hg, was treated with 4-mg IV boluses of ephedrine. Bradycardia was defined as a heart rate <50 bpm and was treated with atropine 0.5 mg IV. Surgery was usually started 20 minutes after the end of induction of anesthesia.
Patients underwent Foley catheterization at the end of induction of anesthesia, immediately after which urine was collected. A sample of collected urine was taken for the measurement of urine osmolality (pre-U osm, in mOsm/kg) and the remainder discarded. The osmolality and volume of urine produced during the study period (post-U osm in mOsm/kg and V U in mL/kg, respectively) were measured. Urine osmolality was measured by using a vapor pressure osmometer (Vapro® 5520; Wescor, Inc., Logan, UT).
Bioelectrical Impedance Measurements
Whole-body resistance and reactance were measured using a multifrequency bioimpedance analyzer (4200 Hydra; Xitron Technologies, San Diego, CA). Disposable electrodes (Red Dot 2330; 3M, St. Paul, MN) were used in the impedance analysis in which the skin had been previously swabbed with alcohol. Impedance measurements were performed using 20 frequencies ranging from 5 to 200 kHz. The impedance spectral data were transferred to a computer and analyzed using the software program provided by the manufacturer. The software also enabled the spectrum of reactance versus resistance to be plotted and for the resistance at zero frequency (corresponding to whole-body resistance for extracellular fluid [R e, in ohm]) to be determined by performing nonlinear curve-fitting and subsequent extrapolation.
Bioelectrical impedance measurements were conducted every 5 minutes during the study period on the side of the body contralateral to that of the venous cannula to avoid electrical interference between the venous cannula and electrodes for impedance measurements. Patients were laid on the operating table in the supine position, with legs and arms placed slightly apart for at least 15 minutes before the start of impedance measurement to establish stable fluid distribution within the body. Current source electrodes were placed at the dorsal surface of the third metacarpal bone, and the dorsal surface of the third metatarsal bone on the contralateral side of the body relative to the venous cannula. Voltage detection electrodes were placed on the dorsal surface of the wrist and the anterolateral surface of the patella on the contralateral side of the body relative to the venous cannula (Fig. 1A in the Appendix).
Patients whose pre-U osm was lower than the 25th percentile of pre-U osm for all patients were categorized in the “hydrated” group and those whose pre-U osm was higher than the 75th percentile of pre-U osm for all patients were categorized in the “dehydrated” group.
Values of R e at each measurement time point relative to baseline (i.e., at the start of propofol administration) and percent decreases of R e relative to baseline at the end of the study period (ΔR e) were obtained for each patient.
Whole-body ECF values for each patient, at both baseline and at the end of study period, were predicted from R e values in each patient using the equation derived from the dielectric properties of cell suspensions as described in the Appendix. 10,11 The change in whole-body ECF at the end of the study relative to baseline (ΔECF, in mL/kg) was obtained and compared with estimated net fluid balance (NFB, in mL/kg) during the study period. Values of NFB were determined from the infused fluid volume, urine output, and estimated insensible water loss (i.e., 0.6 mL · kg−1 · h−1)12 during the study period.
Data are expressed as mean (SD) or median (interquartile range) depending on the distribution. Continuous variables, including post-U osm and time to induction expressed in hours from 9:00 AM to the start time of induction of anesthesia, and ΔR e, were compared between the groups, with the Student t test or the rank-sum test used to test for significance. Because it was found that MAP was stable during the 30- to 90-minute period, MAP during this period of time relative to baseline was compared between the groups. The χ2 test was performed for analysis of categorical variables such as the number of patients who required ephedrine >2 times during the study period.
A second classification of some patients was made with respect to ΔR e values measured in the study, with groupings made according to small values of ΔR e (i.e., < the 25th percentile of ΔR e for all patients) and large values of ΔR e (i.e., > the 75th percentile of ΔR e for all patients). A range of variables, including V U and MAP during the 30- to 90-minute period relative to baseline, were compared between the groups.
For ΔECF, the data were analyzed using a linear regression model, with NFB as an independent variable. Additionally, comparison of the ΔECF and NFB methods was made using Bland-Altman plots, with NFB used as the reference method. ΔECF values were subtracted from NFB values, and the mean difference (bias) and 2 SDs of bias (limits of agreement) were calculated. The SigmaStat 3.0 software package (SPSS, Inc., Chicago, IL) was used for statistical analyses. A P value <0.05 was considered significant.
Sixty patients were enrolled for this study. The median, the 25th percentile, and the 75th percentile of pre-U osm values for all patients were 560, 378.5, and 759.5 mOsm/kg, respectively (Fig. 1, upper panel). Accordingly, patients with a pre-U osm <378.5 mOsm/kg were categorized in the hydrated group (n = 15), whereas those with a pre-U osm >759.5 mOsm/kg were categorized in the dehydrated group (n = 15).
Post-U osm values were lower than pre-U osm values in both the hydrated (55 [39–95] mOsm/kg vs 224 [160–339] mOsm/kg, P < 0.0001) and dehydrated (181 [152–267] mOsm/kg vs 867 [807–976] mOsm/kg, P < 0.0001) groups (Fig. 1, lower panel). The dehydrated group showed higher post-U osm values than the hydrated group (181 [152–267] mOsm/kg vs 55 [39–95] mOsm/kg, P = 0.001).
Patient Characteristics and Measurement Variables
There were no significant differences between the hydrated and dehydrated groups with respect to gender, weight, body mass index (BMI), ASA physical status, and V U (Table 1). Average age was younger in the dehydrated group compared with the hydrated group (44 vs 52 years), and average time to induction was shorter in the dehydrated group compared with the hydrated group (1.1 vs 5.3 hours). The number of patients with time to induction = 0 (i.e., starting at 9:00 AM) was larger for the dehydrated group compared with the hydrated group (7 vs 1, P = 0.013). Blood loss during surgery was negligible for both groups.
Changes of MAP and R e
MAP decreased during the induction of anesthesia and remained almost constant at approximately 60% of that at baseline during the 30- to 90-minute period for both groups (Fig. 2, upper panel). There were no significant differences between the hydrated and dehydrated groups with respect to MAP at baseline (88  vs 91  mm Hg, P = 0.55) with 95% confidence interval (CI) for the difference of means (−13.2 to 7.2 mm Hg), mean MAP during the 30- to 90-minute period (58  vs 60  mm Hg, P = 0.54) with 95% CI for the difference of means (−5.8 to 3.1 mm Hg), mean MAP during the 30- to 90-minute period relative to baseline (0.67 [0.10] vs 0.67 [0.10], P = 0.85) with 95% CI for the difference of means (−0.070 to 0.084), and number of patients who were administered ephedrine >2 times during the study period (7 vs 3, P = 0.12).
The lower panel in Figure 2 shows the time course of R e relative to baseline during general anesthesia for the hydrated and dehydrated groups. Relative R e during general anesthesia could be described by a pronounced decrease during the period with a high fluid infusion rate, followed by a gradual decrease during maintenance fluid infusion. The mean values of relative R e at each measurement time point were similar for the 2 groups and there was no difference with respect to ΔR e value between the hydrated and dehydrated groups (5.6% [2.0%] vs 6.0% [1.9%], P = 0.58) with 95% CI for the difference of means (−1.85% to 1.06%).
Comparison of Variables by ΔR e
No significant differences were found between groups of patients classified according to a small ΔR e (<4.40%, n = 15) and a large ΔR e (>7.45%, n = 15) (see Methods) with respect to gender, age, ASA physical status, time to induction, pre-U osm, post-U osm, MAP at baseline, and mean MAP during the 30- to 90-minute period. Weight and BMI for the group with a large ΔR e were larger compared with the group with a small ΔR e (65 [55–75] kg vs 53 [52–58] kg, P = 0.009) and (22.8 [21.8–25.9] kg/m2 vs 21.8 [20.3–22.4] kg/m2, P = 0.026), respectively.
As shown in Figure 3, the group with a large ΔR e had a smaller V U value compared with the group with a small ΔR e (1.82 [1.53–2.71] mL/kg vs 4.15 [2.79–6.73] mL/kg, P = 0.002), whereas no difference was found between the groups with small and large ΔR e with respect to MAP during the 30- to 90-minute period relative to baseline (0.61 [0.53–0.70] vs 0.63 [0.61–0.73], P = 0.48) with 95% CI for the difference of means (−0.100 to 0.048).
Relationship Between ΔECF and NFB
There was a positive correlation between ΔECF and NFB (r 2 = 0.16, P = 0.0014, upper panel in Fig. 4). The bias between ΔECF and NFB was 0.37 mL/kg and the limits of agreement were −7.32 to 8.07 mL/kg (Fig. 4, lower panel). A negative correlation was found between average ΔECF and the difference in ΔECF (r 2 = 0.20, P = 0.0003).
The major finding of this study was that the hydrated and dehydrated groups categorized according to pre-U osm did not show significant differences regarding crystalloid redistribution into the interstitial space or in relation to the magnitude of hypotension during general anesthesia.
Given that urine osmolality for euhydration varies widely (587–766 mOsm/kg for a 75-kg healthy man),13 we compared extreme cases of hydration status in which pre-U osm values were outside the interquartile range of pre-U osm. The critical values of pre-U osm for categorization in our study (378.5 and 759.5 mOsm/kg) corresponded to the “extremely hyperhydrated” and “slightly or very dehydrated” status in a healthy man, respectively.13 A high post-U osm in the dehydrated group compared with the hydrated group without any difference in V U lends support to the assumption that the dehydrated group was preoperatively more dehydrated than the hydrated group. A high metabolic rate in the younger patients may in part explain the younger age of the dehydrated group. Time to when induction occurred for the dehydrated group was short, with almost half of the patients in the dehydrated group undergoing the induction of anesthesia at 9:00 AM, suggesting that overnight fasting may result in considerable dehydration.
A small bias between ΔECF estimated by bioelectrical impedance and NFB (0.37 mL/kg) is consistent with previous studies showing that bioelectrical impedance measurements may be useful for the evaluation of tissue edema11,14,15 and fluid depletion.16,17 Indeed, we have no direct evidence showing that ΔECF estimated by bioelectrical impedance mainly reflects fluid volume changes in the interstitial space, but a previous finding that bioelectrical impedance changes occurred 5 to 10 minutes delayed after the start of IV crystalloid infusion supports this assumption.8 A tendency that as ΔECF increased (e.g., >15 mL/kg), ΔECF became larger than NFB (i.e., NFB − ΔECF <0, Fig. 4) may be in part attributed to the methodology used for the derivation of ΔECF. Given that ΔECF was theoretically derived from body segment volume, a large weight and BMI for the group with a large ΔR e (i.e., a large ΔECF) may cause errors in the estimation of body segment volume arising from differences in anthropometric variables for these patients. A large V U in the group of patients with a small ΔR e (Fig. 3, upper panel) suggests that as crystalloid redistribution into the interstitial space is decreased (i.e., a small ΔR e), more crystalloid remains in the intravascular space and thereafter is lost via the urine (i.e., a large V U). This scenario does not contradict the finding that MAP during the 30- to 90-minute period relative to baseline was not significantly different between the groups with small and large ΔR e (Fig. 3, lower panel) given that MAP is determined not only by plasma volume but also by homeostatic mechanisms including vascular tone and compensatory venous capacitance.
Contrary to our hypothesis, such fluid redistribution between the intravascular and interstitial spaces during crystalloid infusion was not affected by preoperative hydration status. Given that crystalloid redistribution into the interstitial space is determined by the force exerted by the interstitium to extract fluid from the intravascular space, our result suggests that this force (i.e., hydrostatic and colloid osmotic pressure gradients across the capillary wall) may not change significantly with preoperative dehydration. This scenario is reasonable if we assume that plasma volume and interstitial fluid volume are decreased proportionally during overnight fasting. Moreover, a decrease of MAP induced by anesthetics may cancel out the force exerted by the dehydrated interstitium to extract fluid from the intravascular space. However, significant intravascular hypovolemia after overnight fasting was unlikely to be seen in our study given that a study of gynecological patients showed that blood volume after a 10-hour preoperative fast was not significantly different from calculated normal values in comparable nonfasted patients.18 This finding may provide a reasonable explanation for similar changes of MAP during general anesthesia between the hydrated and dehydrated groups. Additionally, consistent with our result, a recent study demonstrated that there was no significant relationship between fasting time and decrease of MAP during propofol induction (40 mg · kg−1 · h−1) in young patients (aged 18–65 years).19 However, we cannot deny the possibility that bolus injection of propofol or propofol induction in older patients may enhance hypotension at the induction of anesthesia because of preoperative dehydration.
Limitations to our study include the assessment method of hydration status. Indeed, there may be no simple single measurement of hydration status,20 but urine osmolality is a useful index for acute and longitudinal changes in hydration status in healthy men.13 Urine osmolality lags behind rapid changes in hydration status during rehydration after exercise-induced dehydration,21 but this time lag is unlikely to be the cause of any anomaly in our study given that changes in preoperative hydration status progress slowly over several hours. Another limitation is that this was not a randomized controlled study and thus the categorization of hydration status was performed post hoc by pre-U osm. As a consequence, there were significant differences between the groups regarding age and time to induction. Accordingly, we cannot exclude the possibility that good cardiovascular stability in younger patients in the dehydrated group may have masked severe hypotension during general anesthesia arising from preoperative dehydration.
Because only low risk patients were enrolled in this study, we cannot directly extrapolate our findings to high risk patients with cardiopulmonary disorders or prolonged fasting. However, our finding that dehydration after overnight fasting did not increase crystalloid redistribution into the interstitial space or the magnitude of hypotension during general anesthesia suggests that preoperative dehydration after overnight fasting per se does not provide a rationale for intravascular volume loading with crystalloid solution, at least in low risk patients. Such volume loading (i.e., 15 mL/kg over an hour) may be beneficial in patients undergoing minor surgery of short duration by reducing postoperative drowsiness and dizziness,22 but this procedure should not be aimed to prevent hypotension during the induction of general anesthesia. Moreover, given that patients are allowed to drink clear fluids up to 2 hours before general anesthesia, preoperative fluid infusion may be an unnecessary practice because it does not decrease the magnitude of hypotension during general anesthesia.
Name: Toshihiro Osugi, MD.
Contribution: Data collection, data analysis, and manuscript preparation.
Name: Tsuneo Tatara, MD.
Contribution: Study design, conduct of study, data analysis, and manuscript preparation.
Name: Sachiko Yada, MD.
Contribution: Data collection and data analysis.
Name: Chikara Tashiro, MD.
Contribution: Data analysis and manuscript preparation.
Derivation of Theoretical Equation for the Prediction of Extracellular Fluid Volume
Our preliminary study using segmental bioelectrical impedance analysis showed that increases of extracellular fluid volume (ECF) in both lower legs during fluid infusion were considerably smaller compared with other body segments given that the compression of tissue by the fitting of antithromboembolic stockings increases interstitial fluid pressure and thereby decreases the shift of crystalloid fluid from the intravascular space to the interstitial space. Therefore, we have measured impedance of 4 cylinders of body segments in series (i.e., forearm, upper arm, trunk, thigh) assuming that ECF changes during the study period occur in these body segments (Fig. 1A).
The resistivities of body segments at zero frequency (ρ, in ohm · cm) can be obtained from R e (ohm), which is determined by performing nonlinear curve fitting and subsequent extrapolation of impedance values at 5 to 200 kHz10:
where li (cm) and Vi (cm3) are the length and volume of body segment i (i = 1 for forearm, 2 = upper arm, 3 = trunk, 4 = thigh), respectively.
The values of li and Vi can be approximated as follows:
where ai is the ratio of the length of segment i to body height (H, in cm).
where bi and Di are the ratio of the weight of segment i to body weight (W, in kg) and the density of segment i in kg/cm3, respectively. Here, we use for ai, bi, and Di several values obtained from the literature (listed at the end of Appendix).23,24
The equation of the dielectric properties of a cell suspension was used to calculate whole-body ECF excluding both lower legs (liters) based on the assumption that the electrical properties of body tissues are similar to those of a concentrated suspension of spherical particles (cells) in the outer conducting medium (extracellular fluid).10
where V is total body volume excluding both lower legs (cm3) and ρecf is the resistivity of extracellular fluid (= 50 ohm · cm).
By substituting into Equation (4) the value for ρ obtained from Equation (1), we obtain the whole-body ECF excluding both lower legs.
Values of Constants
a 1: 0.17 (male), 0.17 (female); a 2: 0.15 (male), 0.14 (female); a 3: 0.39 (male), 0.39 (female); a 4: 0.22 (male), 0.23 (female).
b 1: 0.026 (male), 0.028 (female); b 2: 0.014 (male), 0.013 (female); b 3: 0.47 (male), 0.45 (female); b 4: 0.13 (male), 0.14 (female).
D 1 (×10−3, kg/cm3): 0.94 (male), 1.19 (female); D 2 (×10−3, kg/cm3): 1.01 (male), 1.09 (female); D 3 (×10−3, kg/cm3): 0.79 (male), 0.87 (female); D 4 (×10−3, kg/cm3): 1.39 (male), 1.28 (female).
1. Drobin D, Hahn RG. Volume kinetics of Ringer's solution in hypovolemic volunteers. Anesthesiology 1999;90:81–91
2. Ewaldsson CA, Hahn RG. Volume kinetics of Ringer's solution during induction of spinal and general anaesthesia. Br J Anaesth 2001;87:406–14
3. Li Y, Zhu S, Hahn RG. The kinetics of Ringer's solution in young and elderly patients during induction of general anesthesia with propofol and epidural anesthesia with ropivacaine. Acta Anaesthesiol Scand 2007;51:880–7
4. Lobo DN, Stanga Z, Aloysius MM, Wicks C, Nunes QM, Ingram KL, Risch L, Allison SP. Effect of volume loading with 1 liter intravenous infusions of 0.9% saline, 4% succinylated gelatine (Gelofusine) and 6% hydroxyethyl starch (Voluven) on blood volume and endocrine responses: a randomized, three-way crossover study in healthy volunteers. Crit Care Med 2010;38:464–70
5. Shime N, Ono A, Chihara E, Tanaka Y. Current practice of preoperative fasting: a nationwide survey in Japanese anesthesia-teaching hospitals. J Anesth 2005;19:187–92
6. Breuer JP, Bosse G, Seifert S, Prochnow L, Martin J, Schleppers A, Geldner G, Soreide E, Spies C. Pre-operative fasting: a nationwide survey of German anaesthesia departments. Acta Anaesthesiol Scand 2010;54:313–20
7. Aukland K, Nicolaysen G. Interstitial fluid volume: local regulatory mechanisms. Physiol Rev 1981;61:556–643
8. Tedner BT, Jacobson HS, Linnarsson D, Lins LE. Impedance fluid volume monitoring during intravenous infusion in healthy subjects. Acute Care 1983–;1984;10:200–6
9. Kazama T, Ikeda K, Morita K, Kikura M, Ikeda T, Kurita T, Sato S. Investigation of effective anesthesia induction doses using a wide range of infusion rates with undiluted and diluted propofol. Anesthesiology 2000;92:1017–28
10. De Lorenzo A, Andreoli A, Matthie J, Withers P. Predicting body cell mass with bioimpedance by using theoretical methods: a technological review. J Appl Physiol 1997;82:1542–58
11. Tatara T, Tsuzaki K. Segmental bioelectrical impedance analysis improves the prediction for extracellular water volume changes during abdominal surgery. Crit Care Med 1998;26:470–6
12. Tatara T, Tashiro C. Quantitative analysis of fluid balance during abdominal surgery. Anesth Analg 2007;104:347–54
13. Armstrong LE, Pumerantz AC, Fiala KA, Roti MW, Kavouras SA, Casa DJ, Maresh CM. Human hydration indices: acute and longitudinal reference values. Int J Sport Nutr Exerc Metab 2010;20:145–53
14. Bracco D, Berger MM, Revelly JP, Schütz Y, Frascarolo P, Chioléro R. Segmental bioelectrical impedance analysis to assess perioperative fluid changes. Crit Care Med 2000;28:2390–6
15. Itobi E, Stroud M, Elia M. Impact of oedema on recovery after major abdominal surgery and potential value of multifrequency bioimpedance measurements. Br J Surg 2006;93:354–61
16. Ackland GL, Singh-Ranger D, Fox S, McClaskey B, Down JF, Farrar D, Sivaloganathan M, Mythen MG. Assessment of preoperative fluid depletion using bioimpedance analysis. Br J Anaesth 2004;92:134–6
17. Ackland GL, Harrington J, Downie P, Holding JW, Singh-Ranger D, Griva K, Mythen MG, Newman SP. Dehydration induced by bowel preparation in older adults does not result in cognitive dysfunction. Anesth Analg 2008;106:924–9
18. Jacob M, Chappell D, Conzen P, Finsterer U, Rehm M. Blood volume is normal after pre-operative overnight fasting. Acta Anaesthesiol Scand 2008;52:522–9
19. Morley AP, Nalla BP, Vamadevan S, Strandvik G, Natarajan A, Prevost AT, Lewis CM. The influence of duration of fluid abstinence on hypotension during propofol induction. Anesth Analg 2010;111:1373–8
20. Kavouras SA. Assessing hydration status. Curr Opin Clin Nutr Metab Care 2002;5:519–24
21. Popowski LA, Oppliger RA, Lambert GP, Johnson RF, Johnson AK, Gisolfi CV. Blood and urinary measures of hydration status during progressive acute dehydration. Med Sci Sports Exerc 2001;33:747–53
22. Holte K, Kehlet H. Compensatory fluid administration for preoperative dehydration: does it improve outcome? Acta Anaesthesiol Scand 2002;46:1089–93
23. Clarys JP, Marfell-Jones MJ. Anatomical segmentation in humans and the prediction of segmental masses from intra-segmental anthropometry. Hum Biol 1986;58:771–82
24. Organ LW, Bradham GB, Gore DT, Lozier SL. Segmental bioelectrical impedance analysis: theory and application of a new technique. J Appl Physiol 1994;77:98–112