Secondary Logo

Journal Logo

Amino Acid Infusion Induces Thermogenesis and Reduces Blood Loss During Hip Arthroplasty Under Spinal Anesthesia

Widman, Jan, MD*,; Hammarqvist, Folke, MD, PhD†, and; Selldén, Eva, MD, PhD

doi: 10.1097/00000539-200212000-00053

The thermic effect of amino acids is augmented under general anesthesia and counteracts hypothermia. Mild hypothermia may increase surgical bleeding. We studied whether amino acids also induce thermogenesis under spinal anesthesia and whether this endogenous heat production reduces bleeding during hip arthroplasty. Rectal temperature, oxygen uptake, and perioperative bleeding were measured in 22 patients receiving an IV amino acid mixture (Vamin 18®, 240 kJ/h) for 1 h before and then during spinal anesthesia and in 24 control patients receiving acetated Ringer’s solution. Blood loss was calculated after surgery by weighing the swabs and the content of the suction tubes after subtraction of the saline used. After surgery, the closed drains were weighed after 24 h. In the amino acid group, the preanesthesia temperature increased by 0.4°C ± 0.2°C (P < 0.01) and was unchanged in controls. At end of surgery, core temperature had decreased by 0.9°C ± 0.4°C in controls and by 0.4°C ± 0.3°C in the amino acid patients (P < 0.01). Oxygen uptake increased by 26 ± 7 mL/min, or 16% ± 5% (P < 0.05), from baseline in the amino acid patients, whereas it was unchanged in the controls. Blood loss during surgery was significantly larger in the control patients (702 ± 344 mL) than in the amino acid patients (516 ± 272 mL) (P < 0.05). After surgery, there were no significant differences in shed blood volume. In conclusion, amino acid infusion also induced a thermogenic response under spinal anesthesia. In addition, the prevention of temperature decrease during spinal anesthesia seemed to have a positive effect on intraoperative blood loss.

*Department of Orthopedics, St. Göran Hospital; †Gastrocentrum, Department of Surgery, Huddinge University Hospital; and ‡Department of Anesthesia and Intensive Care, Karolinska Hospital, Karolinska Institute, Stockholm, Sweden

Supported by grants from The Stockholm County Council, Public Health and Medical Science, Department of Research Development and Teaching.

August 7, 2002.

Address correspondence and reprint requests to Eva Selldén, MD, PhD, Department of Anesthesia and Intensive Care, Karolinska Hospital, S-171 76 Stockholm, Sweden. Address e-mail to

Anesthetics induce hypothermia during surgery, mainly because of decreased heat production and impairment of normal thermoregulatory control. Spinal and epidural anesthesia may cause hypothermia of the same magnitude as general anesthesia (1). Perioperative hypothermia is associated with potentially severe complications in several organ systems, one of them being impaired coagulation (2). Schmied et al. (3) showed that a decreased core temperature increased blood loss during total hip arthroplasty. Mild hypothermia reduces platelet function and depresses activation of the coagulation cascade (4,5). This can lead to increased bleeding.

To avoid hypothermia during anesthesia and surgery, different devices have been used. All are mainly based on the prevention of heat loss, e.g., a forced-air heater for the patient’s upper body combined with prewarming (37°C) of IV fluids. However, another way to prevent the development of hypothermia is to stimulate endogenous heat production. Previous studies have found that IV amino acid infusions exert enhanced thermogenic effects during general anesthesia (6). The mechanism behind this phenomenon is not fully understood, although nutrient intake stimulates energy expenditure, and hence heat production, in the awake state (7). The thermic effect of amino acid infusion has not been investigated during neuraxial anesthesia. In this study, we tested whether IV amino acid infusion in patients with spinal anesthesia induces thermogenesis and prevents hypothermia and, thereby, reduces blood loss in patients undergoing primary hip arthroplasty.

Back to Top | Article Outline


The study protocols were reviewed and approved by the institutional ethics committee. All patients were informed of the study and its purpose and possible risks before giving their consent to participate.

We studied 46 patients scheduled for primary hip arthroplasty at St. Göran’s Hospital, Stockholm, Sweden. They were otherwise healthy, with no history of bleeding disorders, and were receiving no medication but antiinflammatory and analgesic drugs for their hip problems (Table 1). Antiinflammatory drugs were withdrawn at least 1 wk before surgery. Low-molecular-weight heparin 5000 IU was given subcutaneously once daily at 8:00 pm, starting the day before surgery. In 22 patients, an IV amino acid infusion was started 1 h before anesthesia, continued throughout the operation, and completed at the end of surgery. Twenty-four control patients received corresponding volumes of acetated Ringer’s solution. Each group contained an equal number of men and women. The randomization procedure was composed of sealed envelopes, for 24 men and 24 women, marked for either amino acid or acetated Ringer’s infusion. Two patients in the amino acid group were excluded because of incomplete records. Each envelope was opened just before the study procedure. Neither the surgeon nor the patient was aware of which infusion the patient received. The infusion bag was covered with a small opaque sheet.

Table 1

Table 1

All patients were prepared according to the standard preoperative routine after an overnight fast. At 1.5 h before anesthesia, they received oral premedication of 5 mg of diazepam. A thermometer probe was inserted 10–15 cm into the rectum, and two cubital veins were catheterized. One catheter was used for amino acid/saline infusion, and the other catheter was used for anesthetic drug administration. No warming device was used, and all IV fluids were at room temperature, except blood, which was warmed. The ambient temperature was recorded for each patient and was maintained at 21°C. During anesthesia and surgery, 500 mL/h of acetated Ringer’s solution was infused IV in all patients.

Before test infusions of amino acids or acetated Ringer’s solution were started at 1 h before anesthesia induction, baseline measurements of temperature, heart rate, pulse oximetry, and pulmonary gas exchange were made in all study patients. In addition, lean body mass was assessed by bioimpedance analysis (Tanita Body Fat Analyzer TBF 305; Tanita Corp., Tokyo, Japan). Rectal temperature was continuously measured throughout anesthesia and was recorded every 5 min by using a thermometer with two channels, which was also used for operating room temperature monitoring. The measuring accuracy was ±0.1°C, and the response time of the probes was 4.2 s. The temperature probes were calibrated against a precision thermometer with a sensitivity of ±0.01°C. Heart rate and pulse oximetry were measured with HP oximeter M304. Respiratory gas exchange was measured by indirect calorimetry for 6–15 min (mean, 11 min) by using a ventilated hood technique. The Deltatrac is periodically controlled in the laboratory for measurements of gas flow and concentrations and was calibrated before each study. During the periods of gas collection, the gas flow and concentrations were measured continuously, and gas exchange was automatically calculated and recorded for 1 min. Analysis of pulmonary gas exchange was repeated after the end of surgery. No supplementary oxygen was given to the patients during the measurements to avoid the well known possible error in gas exchange analysis.

After baseline measurements, a balanced mixture of 19 amino acids (Vamin® 18 g/L; Fresenius Kabi, Uppsala, Sweden) was given IV to 22 patients at a rate of 126 mL/h, which corresponds to 240 kJ of energy per hour. The 24 control patients received corresponding volumes of nutrient-free acetated Ringer’s solution. The test infusions were given by infusion pump. After 1 h of IV infused amino acids or acetated Ringer’s solution, spinal anesthesia was induced in all patients by using 17 ± 0.2 mg of isobaric bupivacaine. The dermatome level of blockade was recorded. During anesthesia, the usual monitoring was used. In 18 of the patients given the amino acid infusion and in 17 of the control patients, sedative drugs were given in small bolus doses as a complement to spinal anesthesia. The average dose was 100 mg of propofol. Intraoperative vasoactive drugs were not used. All patients received a primary, unilateral total hip arthroplasty because of osteoarthrosis. The seven participating surgeons were all experienced, and three surgeons performed 70% of the operations. The surgical procedure was standardized with the patient in supine position; the modified Hardinge technique was used with an anterolateral approach.

The target minimum hematocrit was prospectively determined to be 33%. No preoperative hemodilution was used. Throughout surgery, hematocrit was determined at 30-min intervals, together with intraoperative fluid balance, on the basis of aspirated suction volume and sponges. Estimated blood loss was replaced by Ringer-Dextran® (Braun, Bromma, Sweden) or, if hematocrit was <33%, by transfusion of packed red cells. Weighing all swabs continuously and, at the end of surgery, weighing the contents in the suction bottles, compensating for used irrigation, accurately measured the total intraoperative blood loss. One closed low-vacuum suction drainage was inserted: Bellovac® Ch 14. The postoperative blood loss was considered to be the collected shed blood after 24 h at drain removal. Autologous blood transfusion was given after surgery when the hemoglobin concentration was less than 80 g/L or if the patient developed unstable vital signs. Blood hemoglobin concentration, INR (international normalized ratio), activated partial thromboplastin time, and platelet counts were assessed the day before surgery, immediately after surgery, and the day after surgery. Bleeding time was not measured in this study, because the correlation between bleeding time and blood loss in individuals is poor, as previously reported (8). During recovery, postoperative pain relief was given as needed after postanesthesia repetition of baseline measurements.

On the basis of preliminary data, a study group of at least 30 patients would give a statistical test power of 80% at an α level of 0.05 to detect a 300-mL hypothermia-induced increase in blood loss. For comparison of different observations within and between the groups, data were first analyzed by repeated-measures analysis of variance, and differences were then calculated by post hoc testing (Scheffé test). Data are given as means ± sd in the text and tables and as mean ± sem in the figures.

Back to Top | Article Outline


There was no difference in anthropometrical data in the two groups (Table 1). The duration of surgery was similar in the two study groups. The mean height of spinal anesthesia, assessed by cold spray sensitivity, was at dermatome level T7 ± 1 in the treatment group and at dermatome level T8 ± 1 in the control group. Hemodynamically, systolic blood pressure values were similar in both study groups to the spinal anesthesia and during surgery (Table 1).

The baseline temperature before amino acid or acetated Ringer’s infusion and spinal anesthesia was 36.6°C ± 0.3°C and 36.9°C ± 0.3°C in the amino acid and control groups, respectively. During 1 h of amino acid infusion before the induction of spinal anesthesia, the mean core temperature increased by 0.3°C ± 0.2°C from baseline values (P < 0.001), whereas it was unchanged in the controls receiving Ringer’s acetate (Fig. 1). Throughout surgery, the reduction in core temperature was more marked in the control group than in the amino acid group (Fig. 1). At the end of surgery, after 120 ± 5 min and 135 ± 7 min (not significant) of spinal anesthesia in the amino acid and control groups, respectively, the average decrease in core temperature from baseline was significantly larger in the controls (0.9°C ± 0.4°C) than in the amino acid patients (0.4°C ± 0.3°C) (P < 0.01) (Fig. 1). Core temperature exceeded 35.5°C in all patients throughout the study period.

Figure 1

Figure 1

Baseline oxygen uptake, measured before the start of infusion, did not differ between the groups (Table 1). However, after the end of surgery, the mean oxygen uptake was increased by 26 ± 22 mL/min, or 16% ± 15 percent, in amino acid patients (P < 0.01), whereas it was unchanged in controls (Fig. 2).

Figure 2

Figure 2

Intraoperative blood loss was significantly larger in control patients at the end of surgery (702 mL; range, 90–1220 mL) than in patients who received amino acids (516 mL; range, 130–1490 mL) (P < 0.05). There were no significant differences in shed blood volume or in the administered volume of allogenic blood between the two study groups during the 24-h study period (Table 2). Initial hemoglobin concentrations did not differ between the two study groups, and the decrease after surgery and on the first postoperative day was similar (Table 3).

Table 2

Table 2

Table 3

Table 3

Before surgery, activated partial thromboplastin time was normal in both groups, and it did not change during the study (Table 3). INR increased significantly (P < 0.05) in both groups, immediately after surgery, from baseline values of 1.01% ± 0.01% and 1.03% ± 0.01% in the amino acid and control groups, respectively. The next morning, INR remained increased in both groups (Table 3). Blood platelet counts showed normal preoperative values in all patients, and these decreased significantly (P < 0.05) in both patient groups immediately after surgery. The next morning, platelet count remained decreased in both the amino acid and control groups (Table 3). However, there were no differences in coagulation values between the groups at any time.

Back to Top | Article Outline


Our study confirmed that amino acid infusion induces thermogenesis during spinal anesthesia, as previously shown during general anesthesia (6). This was reflected by a smaller temperature decrease during surgery and by a stimulation of oxygen uptake. Hence, metabolic rate increased at end of surgery. In addition, this prevention of hypothermia also resulted in a reduced perioperative blood loss.

Amino acid-induced thermogenesis is less manifest during spinal anesthesia than previously demonstrated during general anesthesia (6,9) (Fig. 2). This seems logical, considering the suggested mechanism behind the augmented thermic effect of amino acids during general anesthesia. It is based on nutrient-induced thermogenesis; i.e., nutrient intake, especially proteins and amino acids, stimulates resting energy expenditure and heat production (7). The administration of proteins/amino acids in awake individuals results in an approximately 20% increase in whole-body heat content and a significant increase in body temperature (10,11). During anesthesia, this thermic effect is enlarged. However, because of the decreased metabolism caused by anesthetics and the subsequent decrease in core temperature, the amino acid-induced heat production can be described only as the difference in temperature reduction or oxygen uptake between the two study groups.

Although the mechanism is still incompletely understood, a central metabolic inhibitory pathway may be involved (12,13). During general anesthesia, the hypothalamic thermoregulation is depressed, and this inhibitory pathway is silenced. Hence, the thermic response to amino acid administration is exaggerated. During regional anesthesia, the impairment of central thermoregulation is less marked (14), at least during shorter procedures (15), as in this study. Thus, an attenuated thermic response to nutrients, as compared with during general anesthesia, might be expected. Nevertheless, in this study, amino acid infusion starting one hour before surgery in fact ameliorated hypothermia development under spinal anesthesia, although the increase in oxygen uptake at the end of surgery was more discrete than under general anesthesia. The small doses of propofol used for sedation might theoretically influence thermoregulation, but they were used similarly in both study groups and most probably did not affect the results. In this context, it might, however, be necessary to make a statement that the two studies—general anesthesia (9) versus this study—may not be directly comparable, because the patients were not randomized to anesthetic regimen.

Other known predictors of hypothermia during spinal anesthesia are a high spinal block level, 0.15°C for each incremental increase in dermatome level, and advanced age (16). In this study, there were no differences in age distribution, yet the amino acid group had a mean level block to T8 and the control patients to T7. The differences in core temperature decrease between the two study groups were, however, at no time less than in the range 0.27°C–0.49°C. Therefore, spinal block level could contribute only in part to the temperature differences.

Hypothermia is a well known trigger of coagulation disorders via several documented mechanisms, including reversibly inhibited platelet function, mainly related to reduced release of thromboxane A2(4,5) and depressed enzymatic reactions in the coagulation cascade (2). However, fibrinolysis is reported to remain normal during mild hypothermia (17). In conclusion, blood loss may be reduced if the patient is kept normothermic during surgery. Normally, this is achieved by external warming, e.g., by forced-air blanket, as in the study by Schmied et al. (3), who showed less perioperative blood loss and reduced transfusion requirements during hip arthroplasty in normothermia. In this study, hypothermia was ameliorated by endogenous thermogenesis by using amino acid infusion, with similar results—reduced intraoperative bleeding. The fact that the thermogenic response to amino acid infusion was weaker during spinal anesthesia than previously described during general anesthesia raises the question of whether bleeding might have been even more reduced had general anesthesia been used. Against this speculation is the well documented attenuation of bleeding with use of spinal/epidural anesthesia compared with general anesthesia (18,19). This observation was confirmed in our patients because the intraoperative blood loss was smaller in both groups compared with in the study of Schmied et al. (3).

We found no difference in postoperative blood loss. This might reflect the fact that the thermogenic effect of amino acid infusion was mainly exerted during anesthesia and surgery. A clinically interesting question—whether continuing amino acid infusion after surgery would also result in reduced blood loss after surgery—remains to be answered in future studies. Unfortunately, we were not able to measure postoperative temperatures in this study, thus making conclusions in this regard difficult. In addition, the complexity in determining postoperative blood loss may hide differences in bleeding. The postoperative internal blood amount in the wound hematoma may differ considerably. Recently, the blood volume of the postoperative hematoma was evaluated by using a scintigraphic method (20), and a huge variation was found. Moreover, volume and spread showed no correlation whether drains were used or not. These findings suggest that the volume of shed blood in postoperative drains may not accurately reflect the postoperative blood loss.

Adequate estimation of blood loss during and after surgery remains a complex dilemma. In contrast to our findings, one study showed no increased blood loss in hypothermic patients during hip arthroplasty with spinal anesthesia (21). Three different methods to calculate blood loss were used. The intraoperative loss was estimated visually, and, in addition, determinations of lost hemoglobin and total body hemoglobin balance were used. However, the results among the three methods varied widely (21). In this study, we chose to weigh all external losses in swabs and suction tubes immediately to avoid evaporation errors and to obtain as exact data as possible. The surgical drapes were not weighed, but those losses were considered to be similar in the two patient groups. Further studies are required to establish adequate methods of blood loss assessment, still a clinically important issue, especially after hip arthroplasty.

Platelet number decreased during surgery; however, no differences between hypothermic and normothermic patients were detected. Because bleeding was increased in the hypothermic patients, this finding supports impaired platelet function, as previously reported in vitro(5). Together with the increase in postoperative INR, similar in both study groups, the laboratory values in our patients are also in agreement with the findings in the study of Schmied et al. (3).

In this study, the amount of blood transfusions was similar in the two groups during the postoperative hospital stay and did not reflect the increased intraoperative blood loss in the control group. Possibly, postoperative transfusion requirements could have been reduced if our protocol had included hypotensive anesthesia or hemodilution.

Amino acids infused as a balanced mixture are usually well tolerated and routinely used in intensive care patients. Clinically, these findings, together with our previously reported attenuation of postoperative nitrogen excretion (22) and shorter hospital stay (9), imply that it may be beneficial for the surgical patient to receive amino acid infusion during anesthesia. Amino acids may be needed in the perioperative phase for wound healing, immunocompetence, and vital organ function. Nevertheless, prudence is necessary in patients with severe renal, hepatic, or metabolic illness.

In conclusion, amino acid infusions during spinal anesthesia exerted a thermogenic effect, although this was less marked than during general anesthesia. In addition, this prevention of hypothermia by internal thermogenesis resulted in decreased surgical bleeding. These findings add to recent evidence of fewer complications if hypothermia is prevented in patients during anesthesia and surgery.

The authors wish to thank Dr Britt Blomqvist and Solveig Eriksson for excellent technical assistance.

Back to Top | Article Outline


1. Frank S, Shi R, Raja S, et al. Core hypothermia and skin-surface temperature gradients: epidural vs. general anesthesia and the effects of age. Anesthesiology 1992; 77: 252–7.
2. Rohrer M, Natale A. Effect of hypothermia on the coagulation cascade. Crit Care Med 1992; 20: 1402–5.
3. Schmied H, Kurz A, Sessler D, et al. Mild hypothermia increases blood loss and transfusion requirements during total hip arthroplasty. Lancet 1996; 347: 289–92.
4. Valeri R, Cassidy G, Khuri S, et al. Hypothermia-induced reversible platelet dysfunction. Ann Surg 1987; 205: 175–81.
5. Michelson A, MacGregor H, Barnard M, et al. Reversible inhibition of human platelet activation by hypothermia in vivo and in vitro. Thromb Haemost 1994; 71: 633–40.
6. Selldén E, Brundin T, Wahren J. Augmented thermic effect of amino acids under general anaesthesia: a mechanism useful for prevention of anaesthesia-induced hypothermia. Clin Sci 1994; 86: 611–8.
7. Jéquier E. The influence of nutrient administration on energy expenditure in man. Clin Nutr 1986; 5: 181–6.
8. Rodgers RRPC, Levin J. A critical appraisal of the bleeding time. Semin Thromb Hemost 1990; 16: 1–20.
9. Selldén E, Lindahl SGE. Amino acid-induced thermogenesis reduces hypothermia during anaesthesia and shortens hospital stay. Anesth Analg 1999; 89: 916–22.
10. Brundin T, Wahren J. Influence of protein ingestion on human splanchnic and whole-body oxygen consumption, blood flow and blood temperature. Metabolism 1994; 43: 626–32.
11. Brundin T, Wahren J. Changes in splanchnic and whole-body oxygen consumption, blood flow and blood temperature in response to intravenous administration of amino acids. Am J Physiol 1994; 266: E396–402.
12. Downey JA, Mottram RF, Pickering GW. The location by regional cooling of central temperature receptors in the conscious rabbit. Am J Physiol 1964; 170: 415–41.
13. Hayward J, Baker M. Role of cerebral arterial blood in the regulation of brain temperature in the monkey. Am J Physiol 1968; 215: 389–403.
14. Matsukawa T, Sessler DI, Christensen R, et al. Heat flow and distribution during epidural anesthesia. Anesthesiology 1995; 83: 961–7.
15. Sessler DI. Perioperative heat balance. Anesthesiology 2000; 92: 578–96.
16. Frank S, El-Rahmani H, Cattaneo C, Barnes R. Predictors of hypothermia during spinal anesthesia. Anesthesiology 2000; 92: 1330–4.
17. Watts DD, Trask A, Soeken K, et al. Hypothermic coagulopathy in trauma: effect of varying levels of hypothermia on enzyme speed, platelet function, and fibrinolytic activity. J Trauma 1998; 44: 846–54.
18. Modig J. Beneficial effects on intraoperative and postoperative blood loss in total hip replacement when performed under lumbar epidural anesthesia. Acta Chir Scand Suppl 1988; 550: 95–103.
19. Shir Y, Raja S, Frank S, et al. Intraoperative blood loss during radical retropubic prostatectomy: epidural versus general anesthesia. Urology 1995; 45: 993–9.
20. Widman J, Jacobsson H, Larsson S, Isacsson L. No effect of drains on the postoperative hematoma volume in hip replacement surgery. Acta Orthop Scand. In press.
21. Johansson T, Lisander B, Ivarsson I. Mild hypothermia does not increase blood loss during total hip arthroplasty. Acta Anaesthesiol Scand 1999; 43: 1005–10.
22. Selldén E, Lindahl SGE. Postoperative nitrogen excretion after amino acid-induced thermogenesis under anesthesia. Anesth Analg 1998; 87: 637–40.
© 2002 International Anesthesia Research Society