During halothane anesthesia, infants fail to increase oxygen consumption in response to a cold stimulus in the form of an increase in temperature gradient between body and environment. Based on recent observations with isolated brown-fat cells, it seemed feasible that this inability to respond could be due to an inhibition of nonshivering thermogenesis during halothane anesthesia.
The rate of oxygen consumption was measured in cold-acclimated hamsters and rats. The rate evoked by norepinephrine injection in hamsters at an environmental temperature of approximately 24 degrees C was used as a measure of the capacity for nonshivering thermogenesis. Anesthesia was induced by 3% halothane and maintained by 1.5% halothane. One experimental series with spontaneously breathing hamsters and a second control series with spontaneously breathing rats and with rats whose lungs were mechanically ventilated were conducted.
Norepinephrine injection led to a fourfold increase in the rate of oxygen consumption in control hamsters; after this response had subsided, a second injection led to a similar effect. Halothane anesthesia caused an approximately 20% decrease in resting metabolic rate (P < 0.05) and a 70% inhibition of the thermogenic response to norepinephrine (P < 0.001). The halothane concentration yielding half-maximal inhibitory effect was estimated to be less than 1.0%. After the animals had recovered from halothane anesthesia, a completely restored thermogenic response to norepinephrine was observed. The inhibitory effect of halothane also was observed in hamsters maintained at normothermia and was therefore not secondary to the slight hypothermia that otherwise developed during anesthesia. In a series of control experiments, it was confirmed that rats also showed large thermogenic responses to norepinephrine injections, and it was found that, in spontaneously breathing halothane-anesthetized rats, the thermogenic response to norepinephrine was also much inhibited. Further, in halothane-anesthetized rats whose lungs were mechanically ventilated, and where blood gases were kept at virtually normal levels, the thermogenic response to norepinephrine was found to be similarly markedly inhibited.
A much diminished or abolished thermogenic response to injected norepinephrine was demonstrated in halothane-anesthetized animals. This implies that there would be a diminished ability to elicit nonshivering thermogenesis even when this process is physiologically induced. Such a diminished ability could in part explain the susceptibility of neonates and infants to hypothermia during halothane anesthesia.
Key words: Anesthesia: pediatric. Anesthetics, volatile: halothane. Metabolic rate. Heat production. Brown adipose tissue. Sympathetic nervous system, catecholamines. Norepinephrine. Oxygen consumption. Temperature regulation: nonshivering thermogenesis.
IN infants anesthetized with halothane, the cold stimulus resulting from an increase in the temperature gradient between the body and the environment does not lead to the increase in oxygen consumption  normally evoked in awake infants [2–4] to maintain normothermia. This lack of metabolic response during anesthesia may be part of the explanation for the hypothermia that is often encountered during anesthesia in infants. [5–9].
It is possible that the lack of response to a cold stimulus during halothane anesthesia may be due to a decreased ability of infants to peripherally activate non-shivering thermogenesis and not solely due to an anesthesia-induced change in the central thermoregulatory centers.  In support of this view, it has been shown that halothane inhibits norepinephrine-induced thermogenesis in isolated brown-fat cells. .
There is general agreement that facultative nonshivering thermogenesis emanates mainly from brown adipose tissue, both in newborn infants [4,8,9,12–15] and in newborns of most mammalian species,  as well as in cold-acclimated adult experimental animals. [17,18] The thermogenic activity of brown adipose tissue [19–22] is stimulated by norepinephrine, released from sympathetic nerves. This induces a cascade of events in the brown-fat cells,  leading to lipolysis of the triacylglycerols in the tissue and thus to an intracellular release of free fatty acids. The fatty acids are combusted in the well developed mitochondria of the tissue. The mitochondria are endowed with the tissue-specific uncoupling protein thermogenin (UCP)[24,25] that permits fatty acid combustion unhampered by the controlling effect of oxidative phosphorylation.
In experimental animals, the sympathetic stimulation that physiologically activates brown-fat thermogenesis in the cold can be mimicked by injections or infusions of relatively high doses of norepinephrine at thermoneutral conditions. Such treatments induce a large thermogenic response.  It has been demonstrated that the magnitude of this response is identical to the magnitude of cold-induced nonshivering thermogenesis.  Therefore, it is accepted that the capacity for nonshivering thermogenesis can be estimated by measuring the peak level of norepinephrine-induced oxygen consumption under thermoneutral conditions. However, under such conditions—in which no extra heat production is needed to remain normothermic—the extra heat production induced by norepinephrine injection is a thermal load to the animal. To remain normothermic, the animal will try to compensate for this extra heat production by, if possible, increasing heat loss to the surroundings.
Also in the newborn human infant, norepinephrine infusion leads to a large increase in the rate of the oxygen consumption (heat production); nearly a doubling of the metabolic rate is observed.  However, this response is diminished with increasing age,  in parallel with a decrease in the content of brown adipose tissue in the body. [12,30].
It may be deduced from the above that in studies of nonshivering thermogenesis, cold-acclimated hamsters or rats may be used as experimental models for the human infant. Both of these species have earlier been demonstrated to show a large thermogenic response to norepinephrine, [26,31,32] just like human newborns.  Further, it was with brown-fat cells isolated from hamster that the previous observations of an inhibitory effect of halothane on norepinephrine-induced thermogenesis were made. .
The aim of the current study was to investigate whether halothane had an inhibitory effect on the ability of the organism to elicit nonshivering thermogenesis, as would be predicted from the observations on isolated brown-fat cells.  The thermogenic response to an injection of norepinephrine was therefore used to assess the capacity of the animal for nonshivering thermogenesis during halothane anesthesia. An advantage with this method was that the effects of halothane on central thermoregulatory mechanisms or any effects of halothane on the transduction of signals in the sympathetic nervous system would be circumvented by the peripheral injection of norepinephrine.
Materials and Methods
These studies were approved by the Animal Experimentation Ethical Committee (Stockholm North and North Sweden Regions).
Hamsters. Adult (about 3 months old) Syrian hamsters (Mesocricetus auratus) of either sex, bred at Stockholm University, were used. The hamsters, kept in single cages, were acclimated to cold (5–6 degrees Celsius) and short photoperiod (6 h light/18 h darkness) for at least 2 months. They had free access to food (a diet consisting of rabbit/guinea-pig pellets (John Hansson), supplemented with sunflower seeds, dried carrots, wheat germ, and oats) and tap water. The body weight of the hamsters at the time of the experiments was 109 plus/minus 4 g.
Rats. Female Sprague-Dawley rats were obtained from a local supplier (Eklund). The rats, kept in single cages, were acclimated to cold (5–6 degrees Celsius) and short photoperiod (6 h light/18 h darkness) for at least 3 months. They had free access to food (a standard rat pellet diet, Beekay) and tap water. The body weight of the rats at the time of the experiments was 283 plus/minus 7 g.
All animals were allowed to eat and drink freely until they were placed in the metabolic chamber; no food or water was available in the chamber, and no fluids were administered during the experiments. No animal was used for more than one experiment.
Whole-body oxygen consumption was followed in an open-circuit system, modified from that earlier described. .
The animal was placed in a plastic metabolic chamber. This chamber was placed inside a thermostated incubator (LEEC Incubator model PFC 2CK) at 23–25 degrees Celsius (or as otherwise indicated). A constant air flow was drawn by a pump (Tecnomara type R 301) through the open-circuit system, i.e., through a halothane vaporizer (Fluotec Cyprane) into the metabolic chamber, and from the metabolic chamber, through a gas monitor for volatile anesthetics (Servo Gas Monitor 120, Siemens-Elema) and through a condenser (-3 degrees Celsius) and a drying tube containing CaCl2. From the pump, the air was led through a turbine flow meter (FTO Omniflo, Flow Technology), which, through a feed-back system, controlled the pump to ensure a constant air flow (0.59 l/min), into an oxygen analyzer (Servomex Paramagnetic Oxygen2Transducer Series 1100). The analyzer was calibrated daily with two known gas mixtures (15% and 25% oxygen in nitrogen; AGA). The signal from the oxygen analyzer was connected to a pen recorder that continuously recorded the oxygen concentration (routine setting 21–18.5% Oxygen2). Before the start of each experiment, the oxygen concentration in the incoming air was measured; it was 20.94%. The introduction of halothane into the system led, as expected, to a slight reduction in the oxygen concentration of the air; with a maintenance concentration of 1.5% halothane, the oxygen concentration in the incoming air was measured to be 20.63%, which was exactly the expected reduction (1.5% of 20.94% is 0.31%). The reduction in oxygen concentration due to the addition of halothane is observable in the recordings shown in Figure 1and Figure 4as a baseline shift, indicated on the figures. It was not considered likely that a reduction in oxygen concentration from 20.94% to 20.63% had any effect on the metabolic parameters observed.
Experimental Procedure for Spontaneously Breathing Hamsters. The routine procedure and the definition of experimental values for the experiments on spontaneously breathing hamsters were as follows. A cold-acclimated hamster was placed in the metabolic chamber (volume 1.3 l) and left undisturbed for a chamber-adaptation period of about 2 h. The initial resting metabolic rate (RMR-1) was defined as the lowest rate of oxygen consumption maintained for at least 5 min during the second hour in the chamber (i.e., we have eliminated the metabolic effect of exploratory behavior from the estimate of resting metabolic rate). In the experimental group, about 3% halothane (Fluothane, ICI) was then mixed with the incoming air; after the induction period, the concentration of halothane was reduced to 1.5%(or as otherwise indicated, cf. Figure 3). Control animals were not anesthetized. A second level of resting metabolic rate (RMR-2) was defined as the lowest rate maintained for 5 min during the third hour in the chamber, which thus corresponded to the resting level during anesthesia in the experimental group. The hamsters were then injected intraperitoneally with norepinephrine (1 mg/kg body weight: 1 mg (-)-Arterenol bitartrate salt (Sigma)/ml 0.9% NaCl). The norepinephrine-induced metabolic rate (NE-1) was defined as the highest rate maintained for 5 min after this norepinephrine injection. After the injection, halothane administration was continued for about 60 min and then stopped, and the hamster was allowed to recover from anesthesia for about 1 h. RMR-3 was defined as the lowest metabolic rate maintained for 5 min during this period. The hamster was then injected with norepinephrine as above; the value for NE-2 was defined as the highest metabolic rate maintained for 5 min after this second norepinephrine injection. The recording was ceased about 1 h after the second norepinephrine injection.
To compensate for the effect of body weight on metabolic rate (the reduction of specific metabolic rate with increased body weight),  the rate of oxygen consumption was expressed in ml Oxygen2*symbol* min1*symbol* kg0.75.
In some anesthetized hamsters, a thermometer (Cole-Parmer Instrument) was placed rectally for continuous measurement of body temperature.
Experimental Procedure for Spontaneously Breathing Rats and for Rats Whose Lungs Were Mechanically Ventilated. A series of control experiments were performed with rats. The experiments consisted of two parts. In the first part, the rat was placed in a Plexiglas metabolic chamber with a volume of 5.1. A constant air flow of 0.59 l/min was drawn through the metabolic chamber and through the oxygen analyzing system described above. Oxygen consumption was measured at a room temperature of 24–25 degrees Celsius over a 2-h period. The rat was then injected with norepinephrine, as above, and the oxygen consumption was registered for an additional hour.
For the second part of the control experiments, three experimental protocols were performed with at least two rats being examined in each.
In the first protocol, each rat remained unanesthetized in the metabolic chamber and was injected a second time with norepinephrine.
In the second protocol, each rat was kept in the metabolic chamber during the entire experiment, and the second injection of norepinephrine was administered during 1.5% halothane anesthesia.
In the third protocol, each rat was anesthetized in the metabolic chamber with 3% halothane in air. After the induction period, the rat was placed on an operating table and anesthesia was maintained with halothane via a face mask. The trachea was surgically cannulated with a tube having an ID of 1.67 mm and an OD of 2.42 mm. An airtight condition was established with a banding around the trachea. The tube was connected to a ventilator (Servo Ventilator 900C, Siemens). Anesthesia was maintained with 1.5% halothane in air.
Air from a central gas cylinder was connected to a rotameter (adjusted to a fresh gas-flow of 2 l/min) via the halothane vaporizer and the gas monitor for volatile anesthetics, to the low-pressure inlet of the ventilator. Ventilation was volume-controlled, with expiratory tidal volumes of 6 ml, a respiratory frequency of 90 breaths/min, 20% inspiratory time, and 0% pause time. These adjustments gave a peak airway pressure of 8 cm H2O. The expired gas was led from the ventilator through an adapter with a volume of 0.25 l. From the adapter, the gas was drawn (flow 0.59 l/min) through the oxygen analyzing system described above, back to the adapter, and then to a scavenging system. One carotid artery was catheterized with a polyethylene catheter (ID 0.580 mm and OD of 0.965 mm). Throughout the period, the animal's lungs were mechanically ventilated and blood samples of 0.5 ml were analyzed for blood gases (ABL2 MT A/S, Radiometer)(2–6 samples per animal). Body temperature was continuously monitored with a thermometer (Yellow Springs Thermivolt 741–1) inserted 3 cm into the rectum and was maintained between 36 degrees Celsius and 37 degrees Celsius with a halogen lamp (70 W, Halogen H3) placed above the animal.
When the animal was in steady-state regarding oxygen consumption, blood gases, and body temperature, it was injected with norepinephrine as described above.
Data Analysis and Statistics
The values presented are mean plus/minus SEM from the indicated number of experiments (animals). Statistical significance of difference was tested with Student's paired or unpaired t test, as indicated. (The analyses include comparisons within groups with an identified control; this thus means that ANOVA with subsequent Dunnett's test is an adequate statistical testing method; but as the number of treatments k here is only two (control or halothane anesthetized), the number of comparisons is only k - 1 = 1; as pointed out elsewhere,  Dunnett's test and Student's test are then identical.)
Experiments on Hamsters
(Figure 1) shows representative traces of whole-body oxygen consumption of an unanesthetized (control) hamster (A) and of a hamster that, during a part of the experiment, was anesthetized with halothane (B). In the control hamster, the expected large response to norepinephrine was seen, whereas during halothane anesthesia, the response was much attenuated.
In Figure 2, data from a series of experiments performed as illustrated in Figure 1are summarized. In control animals (n = 7)(Figure 2(A)), a stable resting metabolic rate of 11 plus/minus 1 ml Oxygen2*symbol* min1*symbol* kg0.75 was measured. After injection of norepinephrine, the metabolic rate increased to 43 plus/minus 3 ml Oxygen2*symbol* min1*symbol* kg0.75 (NE-1), after which the animals returned to a resting metabolic rate (RMR-3) of 12 plus/minus 1 ml Oxygen2*symbol* min1*symbol* kg0.75, which was not significantly different from the initial rates (RMR-1 and RMR-2). The metabolic rate observed after a second norepinephrine injection (NE-2) was similar (49 plus/minus 5 ml Oxygen sub 2 *symbol* min1*symbol* kg0.75) to that observed after the first (NE-1) injection. Even if analyzed as the norepinephrine-induced increase in metabolic rate (Delta NE-1 = NE-1 - RMR-2) and (Delta NE-2 = NE-2 - RMR-3), no difference between the responses to the two injections was found: Delta NE-1 was 33 plus/minus 3 ml Oxygen2*symbol* min1*symbol* kg0.75 and Delta NE-2 was 37 plus/minus 5 ml Oxygen2*symbol* min1*symbol* kg0.75.
The initial resting metabolic rate (RMR-1) of the hamsters in the experimental group (n = 5;Figure 2(B)) was not different from that in the control group. During anesthesia, the metabolic rate (RMR-2) was 9 plus/minus 1 ml Oxygen2*symbol* min1*symbol* kg0.75; i.e., about 20% lower than both the initial resting rate (RMR-1) in these hamsters (P < 0.05; paired t test) and than the rate during the same time interval in the control hamsters (P < 0.02; unpaired t test).
In the anesthetized hamsters, the metabolic rate observed after norepinephrine injection (NE-1) was 17 plus/minus 3 ml Oxygen2*symbol* min1*symbol* kg0.75, which was much less than that in the control hamsters. The norepinephrine-induced increase in metabolic rate (Delta NE-1) during anesthesia was only 9 plus/minus 2 ml Oxygen2*symbol* min1*symbol* kg0.75, compared to 33 plus/minus 3 ml Oxygen2*symbol* min1*symbol* kg0.75 observed in the control hamsters (P < 0.001, unpaired t test); i.e., it was reduced by as much as 73%, compared to that seen in control hamsters. After recovery from anesthesia, the resting metabolic rate (RMR-3; 14 plus/minus 2 ml Oxygen2*symbol* min1*symbol* kg sup 0.75) was slightly increased compared to that before anesthesia. The metabolic response to norepinephrine (NE-2) was restored after recovery: the norepinephrine-induced increase (Delta NE-2) was 36 plus/minus 5 ml Oxygen2*symbol* min1*symbol* kg0.75, a value that was significantly higher than that observed in the same hamsters during anesthesia (9 plus/minus 2 ml Oxygen2*symbol* min1*symbol* kg0.75; P < 0.001; paired t test) but not different from that observed after the second norepinephrine injection in control hamsters (37 plus/minus 5 ml Oxygen2*symbol* min1*symbol* kg0.75; unpaired t test).
To obtain a dose-response curve for the inhibitory effect of halothane on norepinephrine-induced non-shivering thermogenesis, a series of experiments was performed, principally as that illustrated in Figure 1but with a different halothane maintenance concentrations being used in each experiment. A compilation of the results of these experiments is found in Figure 3. It is seen that both the resting metabolic rate (RMR-2) and especially the rate after norepinephrine injection (NE-1) were decreased in a dose-dependent manner by higher halothane concentrations (Figure 3). At concentrations greater or equal to 2% halothane, practically no effect of norepinephrine was discernable; half-maximal inhibition was observed with less than 1.0% halothane. The response to norepinephrine after recovery (NE-2) was unaffected by the previous exposure to halothane.
It was observed that, during halothane anesthesia at an ambient temperature of 24 degrees Celsius, the body temperature of the hamsters declined steadily, by about 3 degrees Celsius per hour (not shown). To ensure that the decreased response to norepinephrine observed during halothane anesthesia was not secondary to the lowered body temperature (which had thus decreased by 2–3 degrees Celsius at the time of norepinephrine injection), the chamber temperature in three experiments was increased to 30 degrees Celsius. Under these conditions, the body temperature of the hamster remained at 37–38 degrees Celsius but the response to norepinephrine was still severely diminished, being 6 plus/minus 3 ml Oxygen2*symbol* min1*symbol* kg0.75, i.e., only [nearly equal] 20% of the [nearly equal] 35 ml Oxygen2*symbol* min1*symbol* kg0.75 seen in the control situation (not shown, but see also the rat experiments below where a normothermic body temperature was maintained by external heating).
Experiments on Rats
To verify that the effect of halothane observed above was not secondary to alterations in blood gases or to a decreased body temperature, a series of control experiments were performed in rats (Figure 4). It was first confirmed that, in unanesthetized cold-acclimated rats, norepinephrine injection led to a large increase in the rate of oxygen consumption (Figure 4(A)), here from a resting metabolic rate of 14 ml Oxygen2*symbol* min1*symbol* kg0.75 to a peak rate of 41 ml Oxygen2*symbol* min1*symbol* kg sup 0.75, i.e., an increase of 27 ml Oxygen2*symbol* min1*symbol* kg0.75. As in the hamster experiments, there was no difference between the response to the first and the second norepinephrine injections (an increase of 29 ml Oxygen2*symbol* min1*symbol* kg0.75).
Also in agreement with the results described above in hamsters, halothane anesthesia led to an inhibition of the response to norepinephrine in spontaneously breathing rats (Figure 4(B)), here from an increase of 25 ml Oxygen2*symbol* min1*symbol* kg sup 0.75 in the unanesthetized period to 4 ml Oxygen2*symbol* min sup 1 *symbol* kg0.75 during anesthesia.
In rats whose lungs were mechanically ventilated, it was possible to keep the blood gases and pH stable and within normal ranges during anesthesia with 1.5% halothane in air (Table 1). In these rats, no increase in oxygen consumption was elicited by norepinephrine injection (Figure 4(C)), although a pronounced increase was found before anesthesia, in the same animal.
An Inhibitory Effect of Halothane on Norepinephrine-induced Thermogenesis
In the current investigation, it was observed that halothane led to a large inhibition of norepinephrine-induced thermogenesis, and it was verified that this inhibition was not secondary to hypothermia or hypoxia.
As expected from earlier observations,  norepinephrine injection in cold-acclimated hamsters elicited a large metabolic response: here a fourfold higher metabolic rate was observed (Figure 1(A) and Figure 2(A)). The response declined rapidly to basal levels but an identical response could be elicited by a second norepinephrine injection; thus, there was no attenuation of the response, due to the first stimulation.
For investigation of the effect of halothane on the norepinephrine-induced increase in oxygen consumption, hamsters were anesthetized with an induction dose of 3% halothane and a maintenance dose of 1.5% halothane (Figure 1(B) and Figure 2(B)). During anesthesia, the resting metabolic rate was decreased by about 20%, which is similar to what has been seen in adult humans [36–38] and dogs. [39–41].
A most remarkable effect of halothane anesthesia was observed on the metabolic response to norepinephrine injection: it was reduced by not less than 70%(Figure 1(B) and Figure 2(B)). Thus, in halothane-anesthetized hamsters, only a minor thermogenic response from brown adipose tissue could be evoked, even with the present method that circumvents the sympathetic pathway and in which the adrenergic receptors in the tissue are directly stimulated with exogenous norepinephrine.
However, when the hamster had recovered from anesthesia, a large response to norepinephrine was observed (Figure 1(B) and Figure 2(B)). This response was not different from that observed in control animals. Thus, the inhibitory effect of halothane was found to be fully reversible, provided that sufficient time was allowed for recovery (recovery time for the halothane effect in isolated brown-fat cells was about 30 min. ).
Because the major part of the thermogenic response of hamsters to injected norepinephrine is believed to derive from their brown adipose tissue (see introduction), the results obtained here in intact animals are those that would be predicted from earlier investigations of isolated hamster brown-fat cells. In these cells, a large and reversible inhibition of norepinephrine-induced thermogenesis by halothane was observed.  Accordingly, the reactions of the intact animal can be explained on the basis of the results of the in vitro experiments. In other words, there is not, in the intact animal, any significant additional norepinephrine-induced thermogenesis that is not as sensitive to halothane as is the response of the isolated brown fat cells.
Dose-dependence of Halothane Effect
The results obtained here were observed with clinically relevant doses of halothane. To characterize the concentration dependence of the halothane effect, experiments were performed with different concentrations of halothane. As seen in Figure 3, the effect of halothane on the norepinephrine-induced thermogenic response was dose-dependent. From the curve, it can be seen that a halothane concentration of less than 1.0% led to a 50% reduction of the norepinephrine-induced oxygen consumption. Thus, in this respect, the sensitivity of the animals to halothane was similar to that earlier observed with isolated brown-fat cells, in which an IC50of 0.7% halothane was found. .
Inhibitory Effect of Halothane under Normothermic Conditions
As pointed out above, the resting metabolic rate was somewhat (about 20%) decreased during halothane anesthesia. In the absence of a parallel reduction in heat loss from the animals, it would be predicted that this reduction in resting metabolic rate would lead to a decreasing body temperature during anesthesia. In some anesthetized hamsters, the body temperature was recorded and was found to decrease steadily by about 3 degrees Celsius/h. Heat production in brown adipose tissue is a chemical reaction and would as such be expected to show a high temperature dependence. Although the lipolytic step may be less temperature sensitive than most chemical processes,  it has e.g., earlier been observed that in isolated brown-fat cells, the norepinephrine-stimulated rate of oxygen consumption was reduced to 67% at 27 degrees Celsius (as compared with the rate at 37 degrees Celsius).  Although the body temperature only had decreased by 2–3 degrees Celsius by the time of norepinephrine injection, it could be envisaged that the reduced response to norepinephrine during halothane exposure to some extent could be secondary to the reduced body temperature of the hamsters. Therefore, the effect of halothane was reinvestigated under conditions in which the body temperature was kept high by maintaining a high ambient temperature (30 degrees Celsius) in the metabolic chamber; such a high chamber temperature normally cannot be used during a norepinephrine injection experiment, as experience has shown that such conditions may be detrimental to the hamsters, probably because they cannot then eliminate the large amount of extra heat that normally is elicited by the norepinephrine injection. The increased ambient temperature was found to be adequate to ensure normothermic temperatures in the hamsters during halothane anesthesia: the body temperature was monitored under these conditions and was constantly 37–38 degrees Celsius. It was observed that even when the animals were thus kept normothermic, the norepinephrine-induced increase in metabolic rate was nonetheless very low, only about 20% of that observed in control hamsters above (cf. Figure 1). (The fact that the halothane-anesthetized animal was able to tolerate the experiment in the high environmental temperature was in itself an indication of a reduced heat production under these conditions). Thus, the inhibitory effect of halothane on norepinephrine-induced thermogenesis was not secondary to the slightly lower body temperature found in the anesthetized animals in the experimental conditions routinely used (see also the rat experiments below, in which external heating was used).
Inhibitory Effect of Halothane in Animals Whose Lungs Were Mechanically Ventilated
Thermogenesis in brown adipose tissue results from combustion of fatty acids.  An adequate supply of oxygen, therefore, is essential for unrestricted brown-fat heat production, and a large fraction of the cardiac output (about 1/3) may be directed to the brown adipose tissue during norepinephrine-stimulated thermogenesis.  The circulatory system responds in a way that ensures a high rate of delivery of oxygen to the tissue even under adverse conditions. Thus, when the oxygen content of the blood is reduced by hemodilution, blood flow to brown adipose tissue is augmented, compensating for the reduced blood oxygen content.  It also is known from experiments with isolated brown-fat cells that norepinephrine-induced heat production proceeds undiminished by reduction in oxygen tension until only one-fifth of the atmospheric oxygen pressure remains (also, our unpublished observation). Therefore, it would seem that hypoxia would seldom be a problem for brown adipose tissue function. There are, nevertheless, several observations of a reduced heat production from brown adipose tissue during hypoxic hypoxia, [46–49] and this reduction is not due to a sympathetic inhibition but to a lack of oxygen for combustion. .
Because anesthesia with halothane in air may lead to hypoxia in spontaneously breathing animals, the possibility could not be eliminated that a hypoxic state could contribute to or explain the reduced thermogenic response to norepinephrine observed during halothane anesthesia. Therefore, a series of control experiments were performed in animals whose lungs were mechanically ventilated, with repeated measurements of blood gases being performed to ensure that the animals were not hypoxic when norepinephrine was injected. For technical reasons, it was necessary to perform these control experiments in rats.
It was verified that the responses in spontaneously breathing rats were similar to those observed in spontaneously breathing hamsters. Norepinephrine injection in cold-acclimated rats (Figure 4(A)) led to the expected and well known increase [26,31] in the rate of oxygen consumption, and there was no tendency to a reduced response to a second norepinephrine injection (Figure 4(A)). It then was investigated whether an inhibitory effect was observable in spontaneously breathing rats (Figure 4(B)) anesthetized with halothane. It was found that halothane severely attenuated the response to norepinephrine also in the rat.
The main control experiments were performed with rats whose lungs were mechanically ventilated. In these rats, the body temperature was kept constant at 37 degrees Celsius (Table 1). It is seen in Table 1that, with the ventilatory conditions used, the blood gas values were fairly stable during the experiment. The PaCO2was maintained at about 3.7 kPa (28 mmHg), which was slightly less than that observed in unanesthetized rats (about 5 kPa (38 mmHg)), [50,51] The blood pH was stable at 7.4, which is the value found in unanesthetized rats. [50,51] The PaO2values were around 10 kPa (75 mmHg), which is only slightly less than the values found in awake rats (11–12 kPa (83–90 mmHg)). [50,51] It should be noted that at the observed pH of 7.4, and with the observed PaO2values of 9–10.5 kPa (68–79 mmHg) and PaCO2of 3.7 kPa (28 mmHg), rat hemoglobin shows 82–92% saturation. [52–54] Because of the shallow slope of the hemoglobin saturation curve at these oxygen tensions, the degree of saturation (and thus the total oxygen content of the blood) is only marginally increased between 10 kPa (75 mmHg) oxygen and 12 kPa (90 mmHg) oxygen. [52–54] Thus, a hypoxic state was not found in the mechanically ventilated rat neither before nor after norepinephrine injection.
However, as shown in Figure 4(C), even in these rats, the thermogenic response to norepinephrine injection was virtually eliminated. It was concluded that the attenuation or elimination of norepinephrine-induced thermogenesis, even in spontaneously breathing animals, was not secondary to hypothermia or hypoxia but was a true inhibition by halothane.
Halothane Inhibition of Physiologically Induced Nonshivering Thermogenesis
In the current investigation, injection of norepinephrine has been used to elicit nonshivering thermogenesis. In infants, nonshivering thermogenesis is physiologically induced as a response to an experienced cold stress. There are observations in the literature which, based on the present observations, may be interpreted to imply that such physiologically induced nonshivering thermogenesis is halothane-sensitive. Thus, the metabolism of newborn lambs (that have a high amount of brown adipose tissue  and a high capacity for nonshivering thermogenesis  that is probably activated under standard experimental conditions) is reduced by as much as 50% during halothane anesthesia, [57,58] and similarly, the metabolism of cold-acclimated animals in the cold is dramatically reduced by halothane anesthesia [59–61](also, our unpublished observations). Thus, it would seem that physiologically induced nonshivering thermogenesis is equally sensitive to halothane as is norepinephrine-induced non-shivering thermogenesis. However, the interpretation of halothane effects on physiologically induced thermogenesis is complicated because of halothane effects on thermoregulatory centers,  and only through analysis of norepinephrine-induced thermogenesis, as performed here, is it possible to ascertain that the halothane effects are directly on the heat-generating mechanism of brown adipose tissue.
The results presented here demonstrate that halothane markedly inhibits norepinephrine-induced thermogenesis. This inhibition is dose-dependent and reversible, and it is not secondary to hypothermia or hypoxia. Given that norepinephrine-induced thermogenesis mainly derives from brown adipose tissue, the results were those predicted from in vitro experiments with isolated brown-fat cells.  Because norepinephrine-induced thermogenesis may be equated with nonshivering thermogenesis, the results predict that when nonshivering thermogenesis is the predominant means of thermoregulatory heat production, a competent thermogenic response cannot be elicited during halothane anesthesia. Thus, the results may explain why newborn infants during halothane anesthesia do not demonstrate a thermogenic response to the cold experienced when there is an increase in the temperature gradient between the body and environment. .
The effects of halothane are observed with clinically relevant doses. Because thermoregulatory heat production in newborns and infants is due to nonshivering thermogenesis in brown adipose tissue, an inhibition of nonshivering thermogenesis must occur during anesthesia of such patients. This inhibition may explain, at least partly, the fact that newborns and infants are more susceptible to hypothermia during anesthesia than are adults, despite their potentially greater capacity for nonshivering thermogenesis.
The authors thank Stig Sundelin, for assistance on instrumentation, and Millvej Svensson, for practical assistance.