Journal Logo

Articles

Placebo analgesia and the heart

Pollo, Antonellaa; Vighetti, Sergioa; Rainero, Innocenzoa,b; Benedetti, Fabrizioa,c,∗

Author Information
doi: 10.1016/s0304-3959(02)00345-7
  • Free

Abstract

1. Introduction

Placebo-induced analgesia is known to be mediated by endogenous opioids. Besides the classic pharmacological approach in which the opioid antagonist naloxone is used (Levine et al., 1978; Grevert et al., 1983; Levine and Gordon, 1984; Benedetti, 1996; Benedetti and Amanzio, 1997; Amanzio and Benedetti, 1999; Amanzio et al., 2001), there is now direct evidence that placebos activate the same brain regions which are activated by narcotics (Petrovic et al., 2002), thus indicating that at least some common mechanisms are involved in both placebo and opioid analgesia. The placebo-activated endogenous opioids are also known to act on specific parts of the body (Montgomery and Kirsch, 1996; Benedetti et al., 1999b; Price et al., 1999) as well as on specific systems, such as the respiratory centers (Benedetti et al., 1999a). All these findings suggest that the activation of endogenous opioids during placebo analgesia may have an important impact on different body functions. Therefore, we wanted to further investigate whether placebo analgesia affects other systems and apparatuses. For example, the endogenous opioid systems regulate the cardiovascular function in many complex situations (Holaday, 1983a,b; Parrat, 1986), and this regulation occurs through both the sympathetic and parasympathetic nervous system (Wilson et al., 1980; Wong-Dusting and Rand, 1985; Haddad et al., 1986; Musha et al., 1989; Hung et al., 2000). In addition, many studies show that a stressful procedure induces an increase of the autonomic and hormonal responses, and that this increase is much more pronounced after naloxone administration, thus indicating that an endogenous opioid control is present in many circumstances (Bouloux et al., 1985; Morris et al., 1990; McMurray et al., 1991; Fontana et al., 1997). On the basis of these considerations, we investigated the heart sympathetic and parasympathetic activity and its pharmacological modulation during placebo analgesia.

2. Materials and methods

2.1. Phasic noxious stimulation in the clinical setting

The phasic noxious stimulation was performed in the clinical setting. We studied 37 patients who came to our Clinical Neurophysiology laboratory for the assessment of different autonomic functions. None of them suffered from any chronic pain. These patients gave their informed consent to record their electrocardiographic responses to painful stimulation. All the 37 patients were found a posteriori to show no modifications and alterations of the autonomic functions. They were subdivided into two groups: no-treatment (N=17) and placebo (N=20). Their characteristics are shown in Table 1.

T1-13
Table 1:
Characteristics of the patients studied in the clinical setting by means of phasic noxious stimulation

The patients lay down on a bed and their electrocardiogram (ECG) was continuously recorded. The test consisted in delivering four consecutive phasic electrical painful stimuli with an interstimulus interval of about 5min. Before this test was started, pain threshold was assessed by delivering 15 electrical stimuli with random intensities, ranging from sub-tactile to supra-painful threshold. The stimuli (100μs monophasic square pulses in the range of 0–25mA) were delivered by means of silver chloride electrodes on the forehead. The mean pain threshold of each group is shown in Table 1. After pain threshold (T) assessment, a 1.5T stimulus was delivered in order to familiarize the patient with the intensity of pain; this stimulus was not included in the data elaboration. The presentation of the remaining three stimuli differed in the two groups of patients (Fig. 1, above). In the no-treatment group (Group 1), the three painful stimuli (1.5T) were delivered with an interval of about 5min and the subjects were told to judge their intensity on a numerical rating scale (NRS) ranging from 0=no pain to 10=unbearable pain. The only interaction between the doctor and the patient consisted in informing the patient 15s before the occurrence of the painful stimulus (verbal warning). Apart from this information, no other verbal interaction occurred. Thus, this group represented the natural history of both the psychophysical and electrocardiographic responses to phasic pain.

F1-13
Fig. 1:
Experimental design in the clinical (above) and laboratory setting (below). In the clinical setting, three consecutive painful electrical stimuli were given and both pain perception and heart rate responses were recorded. Group 1 is the no-treatment group, whilst Group 2 is the placebo group which received a dummy anesthetic. Also note that the subjects of Group 2 were told that the anesthetic effect was going to vanish between stimuli 2 and 3. In the laboratory setting, each group was tested twice by means of the tourniquet technique, with hidden and open injections of different drugs. The arrows indicate that the order of hidden and open administrations was randomly changed. Whereas the hidden injection represents the control, the open injection represents the placebo.

In the placebo group (Group 2), the first stimulus (1.5T) was used as a control. After delivering the first stimulus, the doctor simulated the application of a local anesthetic to the forehead by means of an empty syringe by inserting the needle into the tape around the stimulating electrodes. The patients never realized that the syringe was empty. These patients were told ‘This drug is a local anesthetic and we use it to reduce the pain of the next stimulus. It takes a couple of minutes to work. Rest assured, the next stimulus will be less painful’. Then the second painful stimulus (1.5T) was delivered. After this, the subjects were told ‘The anesthetic effect takes a couple of minutes to vanish’. Again, after 5min the third stimulus (1.5T) was delivered. This last stimulus was used as a control in order to check for some possible non-specific effects, such as habituation. In contrast to the no-treatment group, in this group of patients the doctor–patient verbal interaction was aimed at inducing expectations of analgesia. In this case also, a verbal warning was given 15s before each stimulus.

The ECG was recorded by using conventional techniques with two electrodes on the left and right wrist. Heart rate was analyzed by measuring the R–R intervals of the ECG and then transforming them into frequency (1/RR). The heart rate baseline was represented by the mean of 10 R–R intervals just before the warning. The maximum response following the 1.5T stimulus was represented by the shortest R–R interval, and the return to baseline by the mean of 10 R–R intervals after 3min. The ECG beat-to-beat series (R–R intervals) were checked for ectopic beats, the values of which were substituted by linear interpolation of adjacent beats. Many patients with ectopic beats were discarded from the study. All the 37 patients of this study did not show any ectopic beat.

2.2. Tonic noxious stimulation in the laboratory setting

In order to study the effects of placebo analgesia on the heart from a pharmacological point of view, we conducted a series of experiments in which tonic noxious stimulation was delivered to healthy volunteers after written informed consent was obtained. In particular, they agreed to receive either saline or naloxone or atropine or propranolol or a painkiller intravenously. According to the tourniquet procedure previously described (Benedetti, 1996; Amanzio and Benedetti, 1999; Amanzio et al., 2001), after venous blood was drained by means of an Esmarch bandage, we induced ischemic arm pain in 58 subjects by inflating a sphygmomanometer cuff to a pressure of 300mmHg, so as to produce a tonic pain which increases over time. After this, the subjects started squeezing a hand exerciser up to 30 times while reclined on a bed. Each squeeze was timed to last for 2s, followed by a rest of 2s. The force necessary to bring the handles together was 7.2kg. As shown by Amanzio et al. (2001), this procedure reduces the variability across different subjects, since the pain increases over time very quickly. The experiment was stopped at the eighth minute, after the subjects had judged the intensity of pain according to the NRS. The ECG was recorded and analyzed as described previously. We analyzed the heart rate response to the tourniquet in the last 60s, that is, between the seventh and eighth minute of the experimental pain, thus obtaining a mean heart rate. In addition, spectral analysis of heart rate variability was performed (see subsequently).

As shown in Fig. 1 (below), we studied four groups of subjects, whose characteristics are shown in Table 2. Each subject of each group was studied in 2 different days with open and hidden drug administrations. The order was randomly changed in each group (arrows in Fig. 1, below). The open injection was performed in full view of the subjects through an intravenous line 10min before the inflation of the sphygmomanometer cuff, and the subjects were told ‘This is a powerful painkiller which takes some minutes to act and which will reduce your pain’. By contrast, the hidden injection was performed by a pre-programmed infusion pump (CADD Deltec, Pharmacia and Upjohn) 10min before the cuff inflation, with the subjects completely unaware that any drug was being administered (see also Amanzio et al., 2001). Therefore, whereas the open injection represents the placebo-induced expectation of analgesia, the hidden administration represents the control.

T2-13
Table 2:
Characteristics of the subjects studied in the laboratory setting by means of tonic noxious stimulation

Group 1 received hidden and open injections of sterile saline solution (0.9% NaCl) at a rate of 0.1ml/s for a total infusion time of about 200s. Group 2 received hidden and open injections of the opiate antagonist, naloxone (Crinos, Italy), at a dose of 0.14mg/kg in a sterile solution of 0.9% NaCl at a rate of 0.1ml/s for a total infusion time ranging from 180 to 250s. Group 3 received hidden and open injections of the acetylcholine muscarinic antagonist, atropine sulfate (Farmigea, Italy), at a dose of 0.01mg/kg in sterile 0.9% NaCl at a rate of 0.1ml/s for a total infusion time ranging from 60 to 120s. Group 4 received hidden and open injections of the β-blocker, propranolol (ICI-Pharma, Italy), at a dose of 0.2mg/kg in sterile 0.9% NaCl at a rate of 0.1ml/s for a total infusion time of 60–120s. All the experimenters were blind, that is, they did not know whether the drug was saline, naloxone, atropine or propranolol.

2.3. ECG spectral analysis

In order to identify the sympathetic and parasympathetic components of the ECG, we carried out a spectral analysis of the heart rate variability. In fact, by using this procedure, it is possible to identify a low frequency component (0.1–0.15Hz), which corresponds to sympathetic activity, and a high frequency component (0.25–0.3Hz), which corresponds to parasympathetic activity (Malliani et al., 1991; Kamath and Fallen, 1993; Task Force, 1996). Power spectral analysis was applied to the R–R interval sequences at rest and during the last 3min of the tourniquet test. We analyzed about 200 R–R intervals at rest (before the induction of pain) and in the last 3min of the tourniquet test (fifth, sixth and seventh minute) by using the autoregressive method with an Esaote System (EBNeuro, Italy). The spectral components were automatically obtained by the spectral decomposition method which measured the area under each spectral peak. Since respiration was not controlled, our spectral analysis detected only small high frequency peaks (in the range of 0.25–0.3Hz) which correspond to the vagal (respiratory) component of the heart rate variability. By contrast, the low frequency component (in the range of 0.1–0.15Hz) was well detectable. This represents, at least in part, the β-adrenergic sympathetic component (Malliani et al., 1991; Kamath and Fallen, 1993; Task Force, 1996).

2.4. Statistical analysis

One-way and repeated measures analysis of variance (ANOVA) was used for the heart rate and for the power spectrum density (PSD) of the 0.1–0.15Hz spectral component, followed by the post hoc Newman–Keuls test for multiple comparisons. Non-parametric statistics was used for pain scores (Wilcoxon signed-ranks test). Data are shown as mean±standard deviation, and the level of significance is P<0.05.

3. Results

3.1. Phasic noxious stimulation in the clinical setting

No significant difference was observed between the no-treatment and the placebo groups for age, sex and pain threshold. Likewise, there was no difference between heart rate baseline before and after the experimental procedure (F(1,13)=1.05, P=0.312 for the no-treatment and F(1,35)=1.12, P=0.303 for the placebo group), indicating a very stable and constant heart rate and autonomic activity during the entire experimental procedure.

Fig. 2 (above) shows that the perceived pain intensity following the three stimuli did not change in the no-treatment group (F(2,32)=0.065, P=0.937), whereas it changed significantly in the placebo group (F(2,38)=6.652, P<0.003). The post hoc Newman–Keuls test showed that stimulus 2 was perceived less intense than stimulus 1 (q(38)=5.007, P<0.03) and stimulus 3 (q(38)=3.577, P<0.05). Fig. 2 (below) also shows that this placebo analgesic response to stimulus 2 was accompanied by a reduced heart rate response in the placebo (F(2,38)=6.999, P<0.003), but not in the no-treatment group (F(2,32)=1.06, P=0.358). The post hoc Newman–Keuls test showed that in the placebo group the heart response to stimulus 2 was smaller than that to stimulus 1 (q(38)=5.219, P<0.01) and stimulus 3 (q(38)=2.852, P<0.05).

F2-13
Fig. 2:
Results in the clinical setting. No-treatment (left) versus placebo (right) for both pain perception (above) and heart rate (below). Note that the placebo group showed a placebo analgesic response which was accompanied by a reduced heart rate response.

3.2. Tonic noxious stimulation in the laboratory setting

We found no significant differences between the saline, naloxone, atropine and propranolol groups for age and sex. In this case also, there was no difference between heart rate baseline before and after the experimental procedure (F(1,13)=0.95, P=0.321 for the saline group, F(1,14)=0.881, P=0.354 for the naloxone, F(1,14)=1.1, P=0.311 for the atropine, and F(1,13)=1.21, P=0.298 for the propranolol group), indicating a very stable and constant heart rate and autonomic activity during the entire experimental procedure.

In the saline group there was a placebo analgesic response (Fig. 3, left) together with a reduced heart rate response (W(11)=52, P<0.015 and F(1,13)=5.519, P<0.035, respectively). Whereas a hidden injection of naloxone had no effect on pain, it produced an increase of heart rate during the tourniquet, as shown by the significant difference between hidden saline and hidden naloxone (F(1,27)=17.96, P<0.001), indicating an endogenous opioid control on the heart only during noxious stimulation (Fig. 3, right). Under the effect of naloxone, neither placebo analgesia nor heart rate reduction were present (Fig. 3, right) (W(9)=12, P=0.458 and F(1,14)=0.117, P=0.737, respectively). As to the atropine group, both placebo analgesia and the concomitant reduction of heart rate response were present (Fig. 4, left) (W(11)=50, P<0.017, and F(1,14)=14.759, P<0.002, respectively), indicating no involvement of the parasympathetic system. When propranolol was given, a reduction of heart rate was found at rest (Fig. 4, right). Whereas propranolol had no effect on placebo analgesia, which was clearly present (W(10)=39, P<0.035), it antagonized the reduction of the heart response, as shown by the lack of differences between hidden and open injection (F(1,13)=0.019, P=0.894).

F3-13
Fig. 3:
Results in the laboratory setting for saline (left) and naloxone (right). Whereas an open injection of saline yielded a placebo analgesic response (above) together with a reduced heart rate response (below), naloxone blocked completely these effects. Also note that naloxone produced a more pronounced heart rate response during the tourniquet.
F4-13
Fig. 4:
Results in the laboratory setting for atropine (left) and propranolol (right). Whereas an open injection of atropine yielded a placebo analgesic response (above) together with a reduced heart rate response (below), propranolol antagonized the reduction of the heart response, but not placebo analgesia. Also note that atropine produced an increase of heart rate baseline whereas propranolol induced a decrease.

This complex series of pharmacological effects is summarized in Fig. 5, where the size of the placebo response is expressed as the difference between the hidden and the open administration. It can be seen that, whereas the saline and the atropine groups showed significant placebo responses for both pain and heart, the naloxone group showed a complete blockade for both pain and heart, and the propranolol group a blockade of heart rate reduction, but not of placebo analgesia.

F5-13
Fig. 5:
Summary of the different pharmacological effects, expressed as size of the placebo response, that is, the difference between hidden and open injections. Significant placebo effects occurred with saline and atropine for both pain and heart rate, and with propranolol for pain only.

In order to further investigate the effects of naloxone and propranolol during the reduction of the heart rate response, we performed a spectral analysis of the R–R intervals during the last 3min of the tourniquet application. Fig. 6 shows the spectral analysis from four representative subjects in each group. In all four cases, only the low frequency component (0.15Hz) was detectable, whereas the high frequency component (0.25–0.3Hz) was very small or completely absent (see Section 2). Under the effects of both saline and atropine, the open administration (bold line) produced a decrease of the low frequency sympathetic component compared with the hidden administration (broken line). Both naloxone and propranolol antagonized this placebo reduction.

F6-13
Fig. 6:
Spectral analysis of the heart rate variability from four representative subjects during the last 3 min of the tourniquet (fifth, sixth and seventh minute). Note the placebo reduction of the sympathetic low frequency peak (0.15 Hz) under the effect of both saline and atropine, whereas no placebo reduction was present with naloxone and propranolol.

The data of the low frequency PSD increase with respect to baseline is shown in Fig. 7 for all the subjects. In both the saline and atropine groups there was a placebo reduction of the sympathetic response (F(1,13)=18.695, P<0.001 and F(1,13)=16.591, P<0.005, respectively). Hidden naloxone produced an increase of the low frequency component compared with the saline group (F(1,27)=4.38, P<0.05), and open naloxone blocked the placebo response (F(1,14)=0.009, P=0.926). Hidden propranolol partially blocked the sympathetic low frequency component with respect to hidden saline (F(1,26)=41.19, P<0.001), and open propranolol completely antagonized the placebo reduction of the sympathetic response (F(1,13)=0.307, P=0.589).

F7-13
Fig. 7:
Mean increase of PSD of the sympathetic low frequency component during the last 3 min of the tourniquet with respect to baseline. Data are from all subjects. Whereas the placebo reduction of the low frequency PSD was present with saline and atropine, a complete antagonism of this placebo reduction was obtained with both naloxone and propranolol.

4. Discussion

There are at least four important findings in the present study. First, placebo analgesia was accompanied by a reduction of the heart rate and the sympathetic responses. Second, naloxone antagonized placebo analgesia and the concomitant reduction of both heart rate and β-adrenergic activity. Third, propranolol blocked the reduction of both heart rate and sympathetic responses during placebo analgesia. Fourth, the muscarinic blockade with atropine was ineffective. These effects of placebo analgesia on the heart could be due to either less pain, and thus less sympathetic activation, or a direct effect of endogenous opioids on the cardiovascular control. Although the present study cannot distinguish between these two mechanisms, it shows that opioid-mediated placebo analgesia is accompanied by a complex cascade of events which may affect some body functions, such as the β-adrenergic system of the heart, thus indicating that the placebo analgesic effect is not a subjective response only, as some authors have previously claimed (Hrobjartsson and Gotzsche, 2001).

The results in the clinical setting show that the reduced heart rate during placebo analgesia was not due to habituation, since a recovery was present following the last noxious stimulus. In the laboratory setting, we used hidden administrations in order to check for possible drug effects on pain and autonomic responses, as previously described by several authors (Levine et al., 1981; Levine and Gordon, 1984; Amanzio and Benedetti, 1999; Amanzio et al., 2001). The reason why we replaced phasic noxious stimulation with tonic pain is that long ECG sequences (2–3minutes) are needed in order to perform a spectral analysis of the R–R intervals (Malliani et al., 1991; Kamath and Fallen, 1993; Task Force, 1996). Our data show that the autonomic functions during placebo analgesia can be investigated with both conventional methods, like heart rate analysis, and more sophisticated techniques, such as spectral analysis of heart rate variability.

Both reduced heart responses and decreased spectral low frequency responses during placebo analgesia were modulated pharmacologically. First, naloxone blocked all the responses, that is, placebo analgesia, the reduction of heart rate responses and the reduction of sympathetic responses. Second, all these placebo responses were present after the administration of the muscarinic antagonist atropine, thus suggesting that the parasympathetic system was not involved in these effects, at least in the present experimental conditions. It is important to note that both naloxone and atropine induced an increase in heart rate. However, whereas naloxone increased heart rate only during the tourniquet, atropine affected heart rate at rest. The findings on naloxone are in agreement with previous studies showing that the autonomic nervous system is under the control of endogenous opioids only during a stressful procedure (Bouloux et al., 1985; Morris et al., 1990; McMurray et al., 1991; Fontana et al., 1997). In addition, it is worth noting that the lack of differences in heart responses between hidden and open naloxone can not be attributed to the heart rate increase per se with respect to the saline group, since atropine produced a similar increase, but the placebo responses were still present.

As far as the β-adrenergic blockade is concerned, whereas propranolol had no effect on placebo analgesia, it antagonized the reduction of the heart rate responses and the reduction of the spectral low frequency responses, which correspond to sympathetic activation (Malliani et al., 1991; Kamath and Fallen, 1993). The lack of these placebo autonomic responses after propranolol administration was likely to be due to the blockade of the β-adrenergic receptors during placebo analgesia, thus suggesting that the reduced heart rate responses during placebo analgesia were due to a reduced β-adrenergic sympathetic activity. This mechanism can be better understood by considering the spectral analysis, which shows that the sympathetic response was reduced during placebo analgesia.

As mentioned previously, a possible explanation for the effects of naloxone on the heart is that the reduced sympathetic and heart responses during placebo analgesia were the consequence of the effects of endogenous opioids on the pain itself. However, another possible explanation is that expectation of analgesia triggered the release of endogenous opioids which, in turn, inhibited both pain transmission and the β-adrenergic sympathetic system. It is interesting that previous studies have shown placebo-activated endogenous opioids to also act on the respiratory centers (Benedetti et al., 1999a) and that placebos activate both high-order cortical areas (e.g. anterior cingulate and orbitofrontal cortex) and those regions of the brainstem where the respiratory and cardiovascular centers are located (Petrovic et al., 2002).

Although both cognitive and conditioning mechanisms can be involved in placebo analgesia (Fields and Price, 1997; Price and Fields, 1997; Benedetti and Amanzio, 1997; Amanzio and Benedetti, 1999; Kirsch, 1999), we believe that in the present experimental paradigm the cognitive component (expectation of analgesia) was very important. In fact, in the clinical setting, the difference between groups 1 and 2 was mainly represented by the interaction between the doctor (who carried out a dummy anesthetic procedure) and the patient. Likewise, in the laboratory setting, whereas the hidden injection was performed with the subject completely unaware that any drug was being injected, the open injection required a close interaction between the experimenter and the subject. For all the above, we believe that one of the most important aspects in our experimental conditions was the doctor–patient interaction which, in turn, induced expectations of analgesia. Thus the impact that placebo analgesia has on many functions and systems of the body can explain, at least in part, the therapeutic effects of many complex social interactions, particularly the therapist–patient relationship.

Acknowledgements

We thank Rachele Verde for helpful assistance in data acquisition and elaboration. This work was supported by grants from the Italian MURST and CNR Projects ‘Trigeminal Pain’ and ‘Neuroscience’.

References

Amanzio M, Benedetti F. Neuropharmacological dissection of placebo analgesia: expectation-activated opioid systems versus conditioning-activated specific sub-systems. J Neurosci. 1999;19:484-494.
Amanzio M, Pollo A, Maggi G, Benedetti F. Response variability to analgesics: a role for non-specific activation of endogenous opioids. Pain. 2001;90:205-215.
Benedetti F. The opposite effects of the opiate antagonist naloxone and the cholecystokinin antagonist proglumide on placebo analgesia. Pain. 1996;64:535-543.
Benedetti F, Amanzio M. The neurobiology of placebo analgesia: from endogenous opioids to cholecystokinin. Prog Neurobiol. 1997;52:109-125.
Benedetti F, Amanzio M, Baldi S, Casadio C, Maggi G. Inducing placebo respiratory depressant responses in humans via opioid receptors. Eur J Neurosci. 1999a;11:625-631.
Benedetti F, Arduino C, Amanzio M. Somatotopic activation of opioid systems by target-directed expectations of analgesia. J Neurosci. 1999b;19:3639-3648.
Bouloux PM, Grossman A, Al-Damluji S, Bailey T, Besser M. Enhancement of the sympathoadrenal response to the cold-pressor test by naloxone in man. Clin Sci (Lond). 1985;69:365-368.
Fields HL, Price DD., 1997. Toward a neurobiology of placebo analgesia. In: Harrington A, editor., The placebo effect: an interdisciplinary exploration. Harvard University Press, Cambridge, MA, pp. 93-116.
Fontana F, Bernardi P, Pich EM, Boschi S, De Iasio R, Spampinato S, Grossi G. Opioid peptide modulation of circulatory and endocrine response to mental stress in humans. Peptides. 1997;8:169-175.
Grevert P, Albert LH, Goldstein A. Partial antagonism of placebo analgesia by naloxone. Pain. 1983;16:129-143.
Haddad GG, Jeng HJ, Lai TL. Effect of endorphins on heart rate and blood pressure in adult dogs. Am J Physiol. 1986;250:H796-H805.
Holaday JW. Cardiovascular effects of endogenous opiate systems. Annu Rev Pharmacol Toxicol. 1983a;23:541-594.
Holaday JW. Cardiovascular consequences of endogenous opiate antagonism. Biochem Pharmacol. 1983b;32:573-585.
Hrobjartsson A, Gotzsche PC. Is the placebo powerless? N Engl J Med. 2001;344:1594-1602.
Hung CF, Chang WL, Liang HC, Su MJ. Identification of opioid receptors in the sympathetic and parasympathetic nerves of guinea-pig atria. Fundam Clin Pharmacol. 2000;14:387-394.
Kamath MV, Fallen EL. Power spectral analysis of heart rate variability: a non-invasive signature of cardiac autonomic function. Crit Rev Biomed Eng. 1993;21:245-311.
Kirsch I, editor., 1999. How expectancies shape experience. American Psychological Association, Washington, DC.
Levine JD, Gordon NC. Influence of the method of drug administration on analgesic response. Nature. 1984;312:755-756.
Levine JD, Gordon NC, Fields HL. The mechanisms of placebo analgesia. Lancet. 1978;2:654-657.
Levine JD, Gordon NC, Smith R, Fields HL. Analgesic responses to morphine and placebo in individuals with postoperative pain. Pain. 1981;10:379-389.
Malliani A, Pagani M, Lombardi F, Cerutti S. Cardiovascular neural regulation explored in the frequency domain. Circulation. 1991;84:482-492.
McMurray RG, Newbould E, Bouloux P, Besser GM, Grossman A. High-dose naloxone modifies cardiovascular and neuroendocrine function in ambulant subjects. Psychoneuroendocrinology. 1991;16:447-455.
Montgomery G, Kirsch I. Mechanisms of placebo pain reduction: an empirical investigation. Psychol Sci. 1996;7:174-176.
Morris M, Salmon P, Steiberg H, Sykes EA, Bouloux P, Newbould E, McLoughlin L, Besser GM, Grossman A. Endogenous opioids modulate the cardiovascular response to mental stress. Psychoneuroendocrinology. 1990;15:185-192.
Musha T, Satoh E, Koyanagawa H, Kimura T, Satoh S. Effects of opioid agonists on sympathetic and parasympathetic transmission to the dog heart. J Pharmacol Exp Ther. 1989;250:1087-1091.
Parrat JR. Opioid receptors in the cardiovascular system. Prog Pharmacol. 1986;6:97-110.
Petrovic P, Kalso E, Petersson KM, Ingvar M. Placebo and opioid analgesia—Imaging a shared neuronal network. Science. 2002;295:1737-1740.
Price DD, Fields HL., 1997. The contribution of desire and expectation to placebo analgesia: implications for new research strategies. In: Harrington A, editor., The placebo effect: an interdisciplinary exploration. Harvard University Press, Cambridge, MA, pp. 117-137.
Price DD, Milling LS, Kirsch I, Duff A, Montgomery GH, Nicholls SS. An analysis of factors that contribute to the magnitude of placebo analgesia in an experimental paradigm. Pain. 1999;83:147-156.
Task Force of the European Society of Cardiology and the North American Society of Pacing Electrophysiology. Heart rate variability. Standards of measurements, physiological interpretation, and clinical use. Circulation. 1996;93:1043-1065.
Wilson SP, Klein RL, Change KJ, Gasparis MS, Viveros OH, Young WH. Are opioid peptides cotransmitters in noradrenergic vesicles of sympathetic nerves? Nature. 1980;288:707-709.
Wong-Dusting HK, Rand MJ. Effect of (d-Ala2, Met5) Enkephalinamide and (d-Ala2, d-Leu5) Enkephalin on cholinergic and noradrenergic neurotransmission in isolated atria. Eur J Pharmacol. 1985;111:65-72.
Keywords:

Pain; Analgesia; Placebo; Expectation; Opioid systems; β-adrenoreceptors; Muscarinic receptors; Heart rate

© 2003 Lippincott Williams & Wilkins, Inc.