C. Paccalin and M. Jeannerod

Institut des Sciences Cognitives
67 Boulevard Pinel
69675, Bron, France

jeannerod@isc.cnrs.fr

Abstract

Respiration and heart rates were recorded in normal subjects watching effortful actions produced by an actor in front of them. Subjects remained immobile throughout. Two experiments were performed. In Experiment 1, subjects watched a weight lifting performance, either static or dynamic, with increasing weights. In Experiment 2, they watched a walking/running performance on a treadmill moving at increasing speed. In both experiments; no change was found in observers' heart rate. By contrast, consistent changes were found in respiration rate. These changes tended to follow the exercise rhythm of the actor, specially during accelerated running (from 2.5 km/h to 10 km/h) where respiration rate increased linearly with speed of the treadmill. Average maximum increase ranged between 25% and 30% above resting rate. This finding demonstrates activation of central mechanisms related to action performance during observation of effortful actions. It could represent a basis for understanding and imitating actions performed by other people.

Key words: Respiration, Heart, Effort, Imitation, Motor control.
 
Introduction.

Motor execution is associated with cardiovascular and respiratory changes. Although a large amount of these changes seems to be clearly related to adaptation to metabolic needs during production of effort, such as increase in oxygen consumption and evacuation of blood metabolites, this cannot be the only explanation. Indeed, similar, albeit weaker changes are also observed during preparation to effort, i.e., before any muscular contraction has occurred and in the absence of metabolic activity. Adaptation mechanisms thus anticipate forthcoming metabolic needs, with the possible function of shortening the intrinsic delay required for heart and respiration to adapt to effort (1, 3, 26). These results suggest that vegetative mechanisms would be co-activated with motor mechanisms during central preparation of an effortful action, so as to be effective as soon as muscles begin to contract (e.g., 16. See 23 for review).
There is still another condition where heart and respiration have been shown to be modified in the absence of muscle activity. Decety and his co-workers (6) found that mental simulation of the motor action of running on a treadmill at an increasing speed provoked an increase in heart rate, as well as changes in respiratory parameters. In addition, these changes were found to be proportional to the degree of simulated effort. This result was replicated in a further experiment with pedaling against a load at an increasing rate: heart and respiratory rates increased with the pedaling rate, and were higher when the load was increased. The fact that these changes were unrelated to metabolic activity was illustrated by a severe drop in end-tidal CO2 during increased ventilation, indicating that no CO2 was produced at the muscular level. Indeed, NMR spectroscopic analysis of the involved muscles showed lack of change in muscular metabolism during the simulated action (5; see also 27).
In agreement with the above hypothesis of a co-activation of motor and vegetative mechanisms put forward for motor preparation, mental simulation of action was shown to activate central motor structures, including lateral cerebellum, basal ganglia, premotor cortex, posterior parietal cortex (7, 8, 25), and also primary motor cortex itself (20, 24). Mental simulation of action thus corresponds to an endogenous subliminal activation of the motor system, where no overt motor output is produced but where the corresponding vegetative output, devoid of voluntary control, becomes apparent (for a review, see 13).
The present study investigated the possibility that this reasoning could be extended to observation of action, a situation close to mental simulation of action. It has been suggested that, in order to understand actions of other people, one has to internally rehearse the observed movements (see 18). Accordingly, brain activity is modified during observation of action: the pattern of activation resembles that of mental simulation for what concerns premotor and parietal cortex (4, 11). Our working hypothesis, therefore, was that, if present, a concomitant activation of the vegetative system should also become visible in this condition. Two experiments were performed, during which heart and respiration rates were monitored in normal healthy volunteers while they observed an actor performing an effortful action.

Experiment I
Subjects
Eleven healthy male volunteers (medical students or staff members of the Institute) with an average age of 26.18 years (range from 23 to 33 years) participated in this study. None of them regularly exerted sport. All subjects gave their written consent. The experimental protocol was approved by the local Ethical Committee.

Material
Respiratory rate was measured by a mass air flow sensor (Honeywell, Micro switch, AWM 3100V) fixed on a light mask; heart rate was measured by electrocardiogram. Data were acquired using the Dasylab (Data Acquisition Laboratory) software.

Procedure
Subjects sat in a comfortable armchair. They watched an actor placed at 2.5 m. This actor performed weight lifting movement sequences as described below in a special ergometer. The same actor was used throughout the experiment for all the subjects. Subjects heart and respiratory rates were monitored. In addition, the armchair was placed on a force platform (sensitivity, 10 gr) for detecting possible movements.
 The experiment started with a 1 minute relaxation period, during which the actor remained immobile. This period was followed by 4 movement sequences, the order of which was randomly determined for each subject.
- Dynamic sequence without weight : the actor  alternated upright and squat positions for 1 minute at a rate of 20 alternations per minute
- Dynamic sequence with weight :  the actor performed the same alternating movements for 1 minute with a 24 kg bar on his shoulders (the bar was guided by 2 vertical rails), at a rate of 15 alternations per minute
- Static sequence with the 24 kg bar : the actor lifted the bar of 24 kg and hold it on his shoulders for 15 seconds
- Static sequence with a 50 kg bar : the actor performed the same movement but with a 50 kg bar.
 The instantaneous frequency of heart and respiratory rates were measured. Data were analyzed by t test, ANOVA and multiple regression. The instruction given to the subjects was to watch the actor as if they had to perform the same movement after the observation session. They were instructed not to talk or move during the session. They were explained the aim of the study only after the session was completed.

Results
Heart rate : Watching a human actor performing a weight lifting task under different conditions produced no significant effect on the heart rate of the observer. This result was consistent across all subjects. This was true for both observation of dynamic and static sequences (ANOVA: lifting condition: F(2,20)=0.91, p<0.83; holding condition: F(2,20)=1.96, p<0.17).

Respiration rate : An increase in respiratory rate with respect to the relaxation period was observed in all subjects (except one in the static condition) during observation of the actor’s movements  (Figure 1). The average increase was about 29.8 % (from 12.69 cycles per minute (c/min), sd 3.76,  up to 16.47 c/min, sd 3.18) during alternating movements without bar, and about 33.2% (from 12.69 c/min, sd 3.76,  up to 16.90 c/min, sd 3.35) with the 24 kg bar. (ANOVA: F(2,20)=23.56,  p<0.01; Multiple regression: F(1,31)=8.01, =0.45, p<0.01,).
 A less marked increase was observed during the static sequences, from 12.69 c/min, sd 3.76 during relaxation up to 15.10 c/min, sd 4.90 during observation with the 24 kg bar and 15.14 c/min,  sd 5.60 with the 50 kg bar. (ANOVA: F(2,20)=5.57, p<0.01) (Figure 2). There was no relationship to the intensity of the observed effort (Multiple regression: F(1,31)=1.44, p<0.23).

      Figure 1 here

      Figure 2 here

Experiment 2
Subjects
Fourteen healthy subjects (7 males and 7 females) members of the Institute volunteered in the second experiment. They were not involved in regular sport activities. None of them were involved in the first experiment. Mean age was 28.07 years, sd 5.0 (Women: 27.43, sd 4.86, Men: 28.71, sd 5.44).There was no difference between the 2 groups for the age (t test: p<0.66).

Material
It was the same as for Experiment 1, except that no force platform was used. The absence of movement of the subjects was controlled visually.

Procedure
Subjects sat in a comfortable armchair in a relaxed position, in front of a white screen on which a videofilm was projected. The chair was at 3 meters from the screen; picture size was 2x3 m, so that the actor displayed in the film appeared in real size (1.70 m). The film was a set of 10 one minute sequences, as detailed below. Between each sequences, there was a 30 seconds  relaxation sequence displaying a motionless neutral object (a plant).The serial order of the 10 sequences was randomly determined for each subject. Subjects were instructed to watch the actor as if they had to perform the same exercise after the session. They were also instructed not to talk or move during the session. They were informed about the aim of the study after the data recording.
 The 10 film sequences were target and control sequences. The 6 target sequences displayed the actor on a treadmill that was either motionless (rest sequence) or moving at a constant velocity (2.5, 7 or 10 km/h), with or without a slope of 15 degrees (dynamic sequences). Finally, one sequence displayed the actor walking on a flat treadmill progressively accelerating from 0 to 10 km/h over one minute (acceleration sequence).
 Four more sequences were used as controls : empty motionless treadmill, empty moving treadmill, neutral background with mental calculation, metronome beating at 150 beats/minute.  Data were analyzed by t test, ANOVA and multiple regression.

Results
Heart rate : No systematic change in heart rate was observed across target and control conditions. This result was consistent across all subjects. The ANOVA showed an effect of target sequences (F(3,39)=7.95, p< 0.01), but heart rate did not change as a function of increasing velocity of the actor’s running (multiple regression: F(1,54)=0.77, p< 0.38).

Respiration rate : The main result of this experiment was the increase in respiratory rate while observing the actor performing walking or running at increasing speed (Figure 3). The ANOVA showed an effect of velocity of the treadmill (F(3,39)=9.55, p<0.01). The increase in respiration frequency was correlated with running velocity (multiple regression: F(1,54)=3.88, ?= 0.25, p<0.05). This result was consistent across all subjects. Typically, the average increase during observation of running at 10 km/h was about 25.4% with respect to observation of the rest sequence (from 14.40 c/min, sd 4.05  in rest period to 17.46 c/min, sd 4.86 at 10 km/h).

      Figure 3 here

 This result was further confirmed by detailed analysis of respiration changes in frequency during the acceleration sequence, where the actor progressively accelerated from 0 up to 10 km/h. As shown by Figure 4, respiration frequency increased linearly over time during observation of this sequence (ANOVA: F(3,39)=8.20, p< 0.01; Multiple regression: F(1,54)=7.53, ?=0.35, p<0.01).

      Figure 4 here

Effect of control sequences. There was no difference in respiration frequency between the rest sequence (actor motionless) and the sequence showing the empty treadmill (t test: p<0.52). Respiratory rate increased during observation of the metronome sequence (average frequency, 15.99 c/min, sd 3.90), with respect to the rest sequence (t test: p <0.01). It also increased during the sequence with mental calculation (average frequency, 16.43 c/min, sd 4.59, t test: p < 0.01).

Discussion
The main result from this experiment is that the respiratory rate of an immobile observer was increased while he/she watched an actor performing an effortful action. This finding adds to the situations where respiration was found to be modified in the absence of production of physical effort, like motor preparation or motor imagery. By contrast, heart rate was not significantly influenced by observation of the action. Although we have no direct explanation for this dissociation between heart and respiration, it is not entirely surprising. Wuyam et al (27) found no changes in heart rate with respect to rest during mental simulation of running in trained athletes, whereas respiration rate and total ventilation were increased. In Decety et al (6) mental simulation study also, changes in heart rate were far less marked than changes in the respiratory function.
The changes in respiration rate during observation of action were of a relatively small amplitude. On average, they amounted to about 30% of resting rate in Experiment 1 and 25% in Experiment 2. In individual subjects, however, changes by up to 126% could be observed. These values can be compared with those observed in imagined exercise. In this condition, an increase of 20% was considered as a "good response" by Wuyam et al (27). Those are probably conservative values, for at least two reasons. First, we measured respiration rate using a light face mask: this method, also used by the above authors, might have reduced the magnitude of change in rate (see 19). Second, breathing frequency is only one aspect of respiratory function: other more sensitive parameters, such as total ventilation or end-tidal PCO2, which were not measured in the present study, could have revealed larger respiratory changes.
 The main point to be discussed is whether the observed changes in breathing were specifically related to the content of the visual scene displaying effortful actions, or could be explained by non specific factors. It is known that respiratory control is under the influence of mental conditions : ventilation decreases during meditation (14) or conversely, increases during mental activity (e.g., mental arithmetic) and audiovisual stimulation (17). Indeed, the present experiment tends to confirm these results, as respiration rate increased during two of the control conditions, mental arithmetic and listening to the metronome. It can be conjectured, however, that, if respiration is influenced by observation of effort, it should relate, in one way or another, to the degree of that effort. Such a relation was the key argument for relating vegetative changes during motor simulation to the motor content of the mental images (5, 6, 27). In the present experiments, this relation was clearly found during observation of running (Experiment 2). Respiration rate was higher during observation of an actor walking at 7 km/h than at 2.5 km/h. Furthermore, during the sequence with constant acceleration from 0 to 10 km/h, we found a linear relationship of respiration rate to the actor’s running speed (Figure 4). This finding is a strong argument for an effect of the parameters of the observed movements (e.g., speed, frequency of leg movements, etc) on breathing.
 It remains to be determined why this relation did not hold in Experiment 1, specially in the dynamic sequences when the actor alternated squat and upright positions without or with a loaded bar. In fact, there was another difference between these sequences, namely, that the actor alternated positions at a rate lower with the bar than without the bar (15 and 20 alternations per minute, respectively). Thus the possibility exists that the frequency of the actor's movements was the critical parameter. This possibility would also explain why, in the static sequences where the actor stood still while holding the bar, the increase in respiration rate was not influenced by observation of holding bars of increasing weights. Finally, the lack of difference in respiration rate between observing the actor running at a speed of 7 km/h and 10 km/h (Figure 3) can rely on the same explanation : on the film shown to the subjects, the frequency of leg movements increased from 80 cycles per minute at 2.5 km/h to 130 at 7km/h but only to 152 at 10 km/h. This relatively small difference may not have been detected.
 The finding that the breathing frequency of an observer can be modulated by watching rhythmical movements, like walking or running, performed by another person is an interesting one. It adds to well known examples of motor facilitation by observation, the most typical one being contagion of yawning (e.g., 21). These data therefore support the hypothesis that perceiving an action triggers a neural state where the structures potentially involved in executing that action are facilitated. Indications as to the localization of these structures are provided by neuroimaging studies, as mentioned in the Introduction. In addition, a study by Fadiga et al (9) where motor cortex was stimulated transcranially in normal subjects during observation of grasping movements directly demonstrated activation of cortical areas controlling motor output. The set of muscles activated by the stimulus was the same as that recorded while the subject himself actually performed the movement. This result demonstrates an increased excitability of the motor system during observation of actions, quite similar to that described during mental simulation. This facilitation would represent the neural basis for important functions such as imitation or learning by observation (see 10).
 Breathing frequency of the observer thus provides a good index of what is perceived from observed actions. The fact that respiration rate covaries mainly with the actor’s movement frequency is in line with the finding that breathing is normally coordinated with exercise rhythm (during walking or bicycling for example: 2, 12, 15, 22). Hence, the action perception system of the observer would monitor the actor's rhythmical performance, resulting in activation of motor structures involved in re-enacting this same performance. Whereas motor output remains below threshold for producing muscular discharges, this is not the case for respiration: changes in respiration rate in this condition represent an automatic counterpart of this subliminal activation.
 
 

Aknowledgements
We wish to thank Professor J.R. Lacour who kindly provided access to the ergometer for Experiment 1. M. Thevenet, A. Bernard and J.C. Bazin provided invaluable help in several technical aspects of the experiments. A. Jeuffrin was the actor in Experiment 1. C. P. was supported by Hospices Civils de Lyon and CNRS
 
References

1. ADAMS L., GUZ A., INNES J.A., MURPHY K. The early circulatory and ventilatory response to voluntary and electrically induced exercise in man. J Physiol, 1987,383 :19-30

2. BECHBACHE R.R., DUFFIN J. The entrainment of breathing frequency by exercise rhythm. J Physiol,1977,272:553-561

3. COLLET C. Les signes neurovégétatifs de la programmation du mouvement : application à l’étude de l’anticipation perceptivo-motrice chez l’homme. Thèse, Université Claude Bernard Lyon 1, n°196-95,1995

4. DECETY J., GREZES J., COSTES N., PERANI D., JEANNEROD M., PROCYK E., GRASSI F., FAZIO F. Brain activity during observation of actions Influence of action content and subject’s strategy. Brain, 1997,120,1763-1778

5. DECETY J., JEANNEROD M., DUROZARD D., BAVEREL G. Central activation of autonomic effectors during mental stimulation of motor actions in man. J Physiol, 1993,461 :549-563

6. DECETY J., JEANNEROD M., GERMAIN M., PASTENE J. Vegetative response during imagined movement is proportional to mental effort. Behav Brain Res,1991,51 :1-5

7. DECETY J., PERANI D., JEANNEROD M., BETTINARDI V., TADARY B., WOODS R., MAZZIOTTA JC., FAZIO F. Mapping motor representations with PET. Nature,1994,371:600-602

8. DEIBER MP., IBANEZ V., SADATO N., HALLET M. Cerebral structures participating in motor preparation in humans : a PET study. J Neurophysiol,1996,75 :233-247

9. FADIGA L., FOGASSI L., PAVESI G., RIZZOLATTI G. Motor facilitation during action observation : a magnetic stimulation study. J Neurophysiol,1995,73(6) :2608-2611

10. GALLESE V., GOLDMAN A. Mirror neurons and the simulation theory of mind-reading. Trends Cog Sc,1998,2:493-501

11. GRAFTON S.T., ARBIB M.A., FADIGA L., RIZZOLATTI G.  Localization of grasp representations in humans by PET,II : observation compared with imagination. Exp Brain Res,1996,112:103-111

12. HILL A.R., ADAMS J.M., PARKER B.E., ROCHESTER D.F. Short-term entrainment of ventilation to the walking cycle in humans. J Appl Physiol, 1988,65:570-578

13. JEANNEROD, M. FRAK, V.G. Mental imaging of motor activity in humans. Current Opinions in Neurobiology, 1999, 9: 735-739.

14. KESTERSON J., CLINCH N.F. Metabolic rate, respiratory exchange ratio, and apneas during meditation. Am J Physiol,1989,256:R632-R638

15. KOHL J., KOLLER E.A., JAGER M. Relation between pedaling and breathing rhythm. Eur J Appl Physiol,1981,47:223-237

16. KROGH A., LINDHART J. The regulation of respiration and circulation during initial stages of muscular work. J Physiol, 1913,47:112-136

17. MADOR M.J., TOBIN M.J. Effect of alterations in mental activity on the breathing pattern in healthy subjects. Am Rev Respir Dis, 1991, 144:481-487

18. MELTZOFF, A.N. Understanding the intentions of others: re-Enactment of intended acts by 18-month-old children. Developmental Psychology, 1995, 31 : 838-850

19. PEREZ, W., TOBIN, M.J. Separation of factors responsible for change in breathing pattern induced by instrumentation. J. Appl Physiol, 1985, 59 : 15-1520.
 
20. PORRO, C.A., FRANCESCATO, M.P., CETTOLO, V., DIAMOND, M.E., BARALDI, P., BAZZOCHI, M., DI PRAMPERO, P.E. Primary motor and sensory cortex activation during motor performance and motor imagery. A functional magnetic resonance study. J Neurosci, 1996, 16 : 7688-7698

21. PROVINE, R.R. Faces as releasers of contagious yawning: An approach to face detection using normal human subjects. Bull. Psychon. Soc., 1989, 27: 211-214.

22. RASSLER B., KOHL J. Analysis of coordination between breathing and walking rhythms in humans. Resp Physiol, 1996, 00:317-327

23. REQUIN J., BRENER, J. RING, C. preparation for action. In: J.R. JENNINGS, L.J.R. COLES (Eds). Handbook of cognitive psychology. Central and autonomic nervous system approaches. New York, Wiley, 1991

24. ROTH M., DECETY J., RAYBAUD M., MASSARELLI R., DELON-MARTIN C., SEGEBARTH C., MORAND S., GEMIGNANI A., DECORPS M., JEANNEROD M. Possible involvement of primary motor cortex in mentally simulated movement :a fMRI study. Neuroreport,1996,7 :1280-1284

25. STEPHAN K.M., FINK G.R., PASSINGHAM D., SILBERSWEIG D., CEBALLOS-BAUMANN A.O., FRITH C.D., FRACKOWIACK R.S.J. Functional anatomy of the mental representation of upper extremity movements in healthy subjects. J Neurophysiol, 1995, 73 : 373-386

26. TOBIN MJ., PEREZ W., GUENTHER SM., D’ALONZO G., DANTZKER DR. Breathing pattern and metabolic behaviour during anticipation of exercise. J Appl Physiol,1986,60 :1306-1312

27. WUYAM B., MOOSAVI SH., DECETY J., ADAMS L., LANSING R.W., GUZ A. Imagination of dynamic exercise produced ventilatory responses which were more apparent in competitive sportsmen. J Physiol,1995,482:713-724
 

 Figure Legends
 

Figure 1

Experiment 1 : Changes in respiration rate (in cycles per minute) during observation of an actor performing during one minute alternating squat and upright positions without or with a loaded bar (24 kg). Alternating frequency was 2O cycles per minute without the bar and 15 cycles per minute with the bar. Mean values from 11 subjects.

Figure 2

Experiment 1 : Changes in respiration rate (in cycles per minute) during observation of an actor lifting and holding for one minute a loaded bar (24 kg or 50 kg). Mean values from 11 subjects.

Figure 3

Experiment 2 : Changes in respiration rate (in cycles per minute) during watching a one minute film sequence displaying an actor running on a treadmill. Respiration rate is plotted against velocity of the treadmill. 0 indicates a sequence where the actors stands upright on the immobile treadmill. Mean values from 14 subjects

Figure 4

Experiment 2 : Changes in respiration rate (in cycles per minute) during watching the Acceleration sequence : the treadmill was accelerated from 0 km/h to 10 km/h within one minute. The sequence has been divided into three 20 second epochs. Mean values from 14 subjects.