Amino Acid-Induced Thermogenesis Reduces Hypothermia During Anesthesia and Shortens Hospital Stay

Abstract

Amino acid infusion during general anesthesia induces thermogenesis and prevents postoperative hypothermia and shivering. We propose that amino acid prevention of hypothermia during anesthesia shortens the hospital stay. Core temperatures and pulmonary oxygen uptake were measured in 45 patients, receiving an IV amino acid mixture, 126 mL/h, before and/or during isoflurane anesthesia and 30 control patients receiving acetated Ringer’s solution. At awakening, mean core temperature was 36.5° ± 0.1°C in the amino acid group and 35.7° ± 0.1°C (P < 0.001) in the controls. Energy expenditure increased by 54% ± 9% from baseline in amino acid patients in whom shivering was uncommon, but only by 5% ± 4% (P < 0.001) in control patients, of whom the majority developed postoperative shivering. The estimated difference in hospital stay between the two groups was 2.7 days (CI 95%: 1.3–4.0). Multiple regression analysis showed that the variables best predicting hospitalization were duration of surgery, amino acid treatment, and awakening temperatures. Duration of surgery was similar in the two groups and core temperatures at awakening were a result of amino acid infusion, which indicates that amino acid infusion during anesthesia and surgery was the most important factor for the shorter hospitalization.

Implications 

Amino acid infusion during general anesthesia induces thermogenesis and prevents postoperative hypothermia and shivering. Multiple regression analysis indicated that this resulted in a shorter hospital stay.

During general anesthesia, there is a marked decrease in energy expenditure and heat generation (¹). In addition, central thermoregulatory function is impaired, delaying hypothermia defense mechanisms (²). The hypothermic effect is compounded during prolonged operations, especially those in which thoracic and abdominal cavities are exposed to operating room temperatures. Postoperatively, hypothermic patients are uncomfortable and at risk of developing several hypothermia-related complications, such as shivering, coagulation disturbances (³), ischemic cardiac events (⁴), and decreased resistance to wound infections (^5,6). The latter is known to prolong hospitalization (⁶).

Consequently, the prevention of intraoperative hypothermia is of interest, and much effort is focused on heat loss (^7,8). Furthermore, anesthesia-induced hypothermia and hypometabolism can be prevented by amino acid administration (^9–12). These findings are based on nutrient stimulation of energy expenditure and thermogenesis, especially from proteins (¹³) and amino acids (¹⁴). This extra heat production was not only present during general anesthesia with muscle relaxation, it was also increased five-fold (⁹), when compared with awake individuals (¹⁵). This resulted in patients that, at emergence from anesthesia, were normothermic and free from shivering (^9–12).

These data initiated the hypothesis that amino acid infusion during anesthesia and surgery shortened hospitalization. The aim of the present follow-up investigation was, therefore, to analyze if hypothermia-prevention by amino acid-induced thermogenesis during anesthesia and surgery influences the duration of hospital stay. Discharge data from patients treated with IV amino acid infusion perioperatively were compared with data from a control group of patients, treated with isovolumic infusions of nutrient-free acetated Ringer’s solution.

Methods

The 75 patients, 27 men and 48 women, who participated in the study were all scheduled for abdominal surgery. Anthropometric data are presented in Table 1. All patients were otherwise healthy and did not have any maintenance medical treatment, and thus classified as ASA physical status I. They have participated in previous reports (^9–12). These reports were not focused on the endpoint of the current study, namely duration of hospitalization. Forty-five of the patients received an IV amino acid infusion before and/or during anesthesia, and 30 control patients received corresponding volumes of IV nutrient-free acetated Ringer’s solution. The surgical procedures performed are listed in Table 2.

Table 1: Anthropometric and Clinical Data from Patients Receiving IV Amino Acids, 126 mL/h or Isovolumic Saline Solution Before/During Anaesthesia

Table 2: Surgical Procedures in the Two Study Groups

Every second patient who agreed to participate was assigned to receive amino acids during anesthesia and surgery. The remaining patients were controls. All patients had been informed of the purpose and risks of the study before giving their consent to participate. The study protocols were reviewed and approved by the institutional ethics committee. Patients, nurses, and surgeons were unaware of whether amino acids or acetated Ringer’s solution was given. Otherwise, the study was not blinded. Attending surgeons, unaware of the patient’s group assignment and core temperature, determined when to begin oral feeding after surgery, remove sutures, and discharge from the hospital. The timing of discharge was based on routine surgical considerations, including the return of bowel function, control of bleeding or any infections, and adequate healing of the wound.

All patients were prepared according to standard preoperative routines, after an overnight fast, with oral premedication, 1–3 mg of lorazepam, except for 3 patients subjected to laparoscopic cholecystectomy, who were given cetobemidon, 5 mg IM. Temperatures in the operating rooms were kept between 20° and 23°C. Except for warmed blood, four units in a control patient and one unit in an amino acid-treated patient, no warmed infusions or warming devices were used. Blood loss was <600 mL in the remaining patients (see Table 1). IV fluids given to the patients were at room temperature. In all patients, nutrient-free acetated Ringer’s solution was infused at 500 mL/h during surgery.

A balanced mixture of 19 amino acids (Vamin 18gN/l^®; Pharmacia & Upjohn, Stockholm, Sweden) was infused IV at a rate of 126 mL/h (corresponding to 240 kJ/h). These infusions were given 1 h before the induction and continuing 1 h (n = 8) or 1.5 h (n = 8) into anesthesia and surgery, 2 h before the induction of anesthesia (n = 8), or for 2.5 h during anesthesia and surgery (n = 21). Postoperative measurements were performed at awakening, 15–20 min after extubation. Control patients received equal volumes of nutrient-free acetated Ringer’s solution (Ringeracetat^®; Pharmacia & Upjohn). After emergence from anesthesia, 1000–1500 mL of a 5% glucose solution was administered in all patients until the next morning. On the first postoperative day, the patients were on a liquid diet, or if that was not possible, they received an equal amount of calories via an IV glucose solution.

In 13 amino acid-treated and in 14 control patients, blood temperature was continuously recorded at a sampling frequency of 1 Hz via thermistor catheters inserted into the pulmonary artery, as described by Brundin et al. (¹⁶). The measuring accuracy of the precision blood thermometer was ±0.001°C, and the sensitivity was 0.0003°C. In the other patients, rectal thermometry (MC8700 thermometer; Exagon, Roskilde, Denmark) was used, with the probe inserted 10–15 cm, to assure the detection of core temperature. The measuring accuracy of the rectal thermometer was ± 0.1°C. All thermistors were calibrated against the same precision thermometer (¹⁶), in order to avoid artifacts caused by different temperature probes. Temperature recording was started immediately before the onset of amino acid/acetated Ringer’s solution infusion and continued throughout the anesthesia.

Respiratory gas exchange was measured in the awake state before anesthesia, and within 15–20 min after extubation at emergence in all patients, by indirect calorimetry with a ventilated hood technique (Deltatrac; Datex, Helsinki, Finland). The Deltatrac was periodically controlled in the laboratory for measurements of gas flow and gas concentrations and was calibrated before each study. During periods of gas collection, gas flow and concentrations were measured continuously, and gas exchange was automatically calculated and recorded for 1 min.

The anesthetic protocol was standardized. Anesthesia was induced with thiopental, 5 mg/kg, and maintained with either isoflurane or enflurane in 40% oxygen/60% nitrous oxide. Standard monitoring was used. The trachea was intubated after an IV bolus dose of atracurium (0.5 mg/kg) followed by a continuous infusion of 0.5 mg · kg⁻¹ · h⁻¹. Before surgery, fentanyl 3 μg/kg was given. The atracurium infusion was terminated 0.5 h before the end of the operation, and muscle relaxation was antagonized by using 2.5 mg of neostigmine with 0.5 mg of glycopyrrolate or atropine at the end of surgery. During recovery, all patients were observed by the same observer for the presence or absence of clinically apparent shivering.

Standard statistical methods were used. For comparison between the groups, data were first analyzed by repeated measurements analysis of variance. The unpaired t-test was used where appropriate. Correlations between different studied variables that potentially contributed to hospital stay were performed using simple and multiple linear forward stepwise regression. Data in the text and tables were given as mean ± SEM, except for Table 3, in which mean ± SD was used.

Table 3: Statistical Analysis with T-Test and Univariate Correlation (r) for Hospital Stay

Results

There were no differences in age, height, weight, or duration of surgery between the two study groups (Table 1). Arterial oxygen saturation was above 95% in all patients throughout the anesthesia.

Baseline temperature and pulmonary oxygen uptake before amino acid infusion and anesthesia did not differ between the two groups (Table 1). Immediately after the induction of anesthesia, the core temperature started to decrease in both groups. However, during the entire period of anesthesia and surgery, the decline in core temperature was significantly more in control patients, 0.8° ± 0.1°C/h with a maximal decrease of 1.2 ± 0.1°C, than in amino acid treated patients, 0.5° ± 0.1°C/h, maximal decrease of 0.8 ± 0.1°C (P < 0.001, Table 1).

At awakening, the average increase in oxygen uptake, from baseline, was 111 ± 17 mL/min in all patients receiving amino acids, indicating a rise in oxidative heat production by 37 ± 6 W (P < 0.001). This restored core temperature to, on average, 36.5 ± 0.1°C, which was not different from baseline. In the control group, however, the increase in oxygen uptake at emergence was only 10 ± 10 mL/min or 4 ± 3 W above baseline (not significant), and core temperature remained low, 35.7° ± 0.1°C (P < 0.001, Table 1).

All control patients had shivering at awakening, whereas shivering was noted in only three amino acid-treated patients. Hypothermia prevention and awakening temperatures were similar whether amino acids were infused before or during anesthesia and surgery.

The duration of hospitalization was 6.4 ± 0.3 days in amino acid treated patients (n = 45) and 8.2 ± 0.7 days in control patients (n = 30) (P = 0.012) (Fig. 1) (Table 3). Thus, the estimated difference in mean hospital stay between amino acid and control patients was 2.7 days, with a 1.3 to 4.0 day 95% confidence interval. No significant differences in hospitalization were found between patients subjected to general surgery and those to hysterectomy (Table 3).

Figure 1: Days of hospitalization after abdominal surgery in 45 patients receiving IV amino acids, 126 mL/h, and 30 control patients. Dotted lines indicate mean values in each group.

A stepwise multiple linear regression analysis was performed in all 75 patients by using hospital stay as the dependent variable. The independent variables analyzed were age, height, weight, sex, baseline temperature, awakening temperature, duration of surgery in minutes, and whether amino acids were given. First, a univariate analysis was performed to evaluate the correlation of each independent variable with regard to length of hospital stay. This revealed three significant variables: amino acid treatment, duration of surgery, and baseline temperature. They all were linearly related to the duration of hospital stay (Table 3). The multivariate analysis then showed the strongest relationship between hospital stay and duration of surgery (P < 0.001), to be followed by whether amino acids were given (P < 0.001), whereas baseline temperature was shown not to be related to the length of hospital stay. However, the multivariate analysis revealed awakening temperature to be a significant variable for the duration of hospitalization (P < 0.01, Table 4). Using these three variables (Table 4), the best predicting estimate for hospital stay was found to be 44% (adjusted r² = 0.441, P < 0.001).

Table 4: Multivariate Linear Forward Stepwise Regression Analysis with Respect to Days of Hospitalization

Discussion

The most important finding in this investigation was that amino acid infusions given before and/or during anesthesia and surgery reduced the incidence of hypothermia and shivering in patients on emergence and shortened hospital stay.

Duration of hospitalization after surgery is a multifactorial and complex end point with several confounding factors. First, the investigator bias will have to be scrutinized. Our team knew what kind of infusion was used. Patients, surgeons, and ward staff did not, and it was they who set the discharge criteria. It is therefore not likely that the investigators influenced the time for hospital discharge. Second, postoperative complications could severely jeopardize the time for discharge. As a result, fairly straightforward, standardized surgical procedures were investigated, and postoperative complications were few. Only one postoperative wound infection occurred in each group, and blood loss in all patients but one in each group was below 600 mL. There was, however, one patient in the control group who was found to have a malignant disease which prolonged his hospital stay (discharged on Day 22). Most likely, this patient did not influence the conclusions, because the statistical difference between groups regarding hospital stay remained when analyses were performed without his data. This is important because such a confounding factor is not eliminated by the linear stepwise multiple regression analysis used.

The ideal method for assessment of body temperature in connection with anesthesia and surgery is a matter of debate, as all available methods have potential limitations. In our study, body temperature was measured by using blood and rectal thermometry. A rectal thermometer, when correctly applied, reflects core temperature and remains a reproducible method to show changes from baseline in patients. It is convenient to use in both awake and anesthetized states, although it has a relatively long response time. It is true that an open abdomen is a source of error, which, however, can be at least partly controlled by warmed irrigation fluids and good surgical technique. This artifact was, in the current study, similar in the two groups. It is therefore not likely that it influenced results and conclusions.

There were three strong statistical correlations with hospital stay with using multivariate analysis. These were duration of surgery, whether amino acids were given, and body temperature at awakening. The strongest was related to operating time, which is consistent with well recognized clinical experience (¹⁷). Although it was not a goal for our study to investigate this, it was highly correlated when both groups were analyzed together. There was, however, no difference in surgical time between patients who received amino acids and controls. Hence, another of the three factors must have played a major role for the shorter hospital stay. Was it caused by the awakening temperature, i.e., normothermia at emergence, or was it an effect of amino acids per se with concomitantly increased oxidative metabolism during anesthesia? Body temperatures at awakening are, in fact, secondary to amino acid infusion (^9–12). Hence, amino acid infusion dominates as the most logical factor concerning the shorter hospital stay.

Protein balance was previously shown to be better maintained postoperatively when amino acids were infused during surgery (¹²). This may reduce protein-wasting from skeletal muscle for the synthesis of proteins that are needed in the healing process at the site of trauma, thereby preserving skeletal muscle function in the early postoperative phase. The amount of amino acids given is, however, small and therefore most probably of limited value as an explanatory mechanism for better recovery postoperatively. In addition, Kurz et al. (⁶) recently showed that higher body temperatures after colorectal surgery, achieved by external heating, not only reduced the incidence of wound infection, it also resulted in a shorter hospitalization period. The discharge criteria were identical in our study. Because there is a conformity with regard to the influence on hospitalization between the use of external heating by Kurz et al. (⁶), and our study using “internal heating” from enhanced oxidative metabolism of infused amino acids, it seems as if amino acid treatment per se is not the key factor for the shorter hospital stay. Higher awakening temperatures seem to be the important factor, satisfying both the findings of Kurz et al. (⁶) and those we presented.

Another hitherto, not well studied mechanism that could explain both Kurz et al.’s (⁶) results and ours, is the beneficial effect of a higher awakening body temperature on postoperative shivering. Shivering is almost eliminated by normothermia at awakening irrespective of the method used for maintaining temperature balance, be it from external heating as used by Kurz et al. (⁶) or from endogenous thermogenesis in response to amino acids, as in our study. Based on this reasoning, it may be suggested that shivering at emergence from anesthesia could influence postoperative recovery. Further investigations are needed to evaluate if postoperative shivering can have such an important influence on outcome after anesthesia and surgery.

It is evident that there is no block of amino acid use caused by anesthesia and surgery. In fact, the thermogenic effect of amino acids on energy expenditure is enhanced five-fold as compared with the awake state (⁹). The basic mechanism for this is, however, not known, in contrast to Plattner et al. (¹⁸), who found that nonshivering thermogenesis from brown adipose tissue in neonates is most probably inhibited during anesthesia, a clinical result that supports earlier clinical observations (¹⁹) and previous experimental findings on brown fat cells (²⁰). Because nonshivering thermogenesis in brown fat cells is well described as a mitochondrial phenomenon (²¹), it is not likely that the enhanced nonshivering thermogenesis in response to amino acids during anesthesia is achieved via a similar mitochondrial mechanism as in brown fat. In addition, the present investigation is performed in adults, hitherto not demonstrated to have significant heat production from brown adipose tissue (^22,23). The mechanisms behind thermogenesis under anesthesia are important issues in future research. Based on the present findings, it suffices here to state that enhanced thermogenesis, most probably caused by stimulated cellular metabolism in response to amino acids during anesthesia and surgery, caused normothermia at awakening. This metabolic stimulation occurs as the surgical trauma has its peak effect. It therefore challenges the speculation that if “internal heating” from amino acids during surgery not only, like external heating, protects from postoperative hypothermia and shivering, it also is superior to external heating with regard to other postoperative complications, such as infections and disturbances of the coagulation system. The beneficial effects of postoperative hypothermia are unquestionable. Large epidemiologic studies are now needed to elucidate which temperature balancing method to recommend in the future. Until then, prudence is necessary when interpreting the current data that are not based on a stratified, randomized, double-blinded design.

In conclusion, the present investigation showed that amino acid infusions before and/or during anesthesia and surgery restored core body temperature at awakening, almost eliminated postoperative shivering, and resulted in 2.7 fewer days of hospital stay. The effects of a relatively harmless amino acid infusion, if patients with hepatic or renal insufficiencies and severe metabolic disorders are excluded, on complex mechanisms of temperature balance are dramatic. They are likely to enhance the quality of patient care and improve cost effectiveness.