- Email: [email protected]
Pulmonary changes induced by frontal EMG training
3
Bdogical Ps.schologv 18 (1984) 3-10 North-Holland
PULMONARY Andrew Depurtment Accepted
CHANGES
HARVER
INDUCED
BY FRONTAL
EMG TRAINING
*
** and Harry KOTSES
of Psychology, Ohio University, Athens, OH 45701, U.S.A for publication
20 June 1983
Earlier research suggested that the effects of facial muscle tension changes on other responses are not widespread but limited to a rather narrow set of pulmonary events. Further evidence in support of the specificity of the facial muscle-pulmonary relationship was provided in the present study by monitoring changes in several responses as a function of muscle tension training. Feedback training for increases and decreases in muscle tension at both facial and limb muscle sites was given to adult males. The effects of these manipulations on PEFR, RR, and HR were examined. Increases in facial muscle tension resulted in PEFR decreases whereas increases in limb muscle tension did not. Decreases in facial muscle tension were not observed as a function of training and no PEFR changes resulted from these conditions. Neither RR nor HR were related to the facial EMG changes observed during feedback training. These observations demonstrated the specificity inherent in the relationship between facial muscle tension and PEFR. and lent support to the hypothesis that these two responses are linked reflexively.
1. Introduction In a series of related studies (Glaus and Kotses, in press; Kotses, Glaus, Bricel, Edwards and Crawford, 1978; Kotses, Glaus, Crawford, Edwards and Scherr, 1976) a systematic relationship was described between changes in facial muscle tension and changes in peak expiratory flow rate (PEFR), the latter an indirect measure of airways resistance. Both normal adults and asthmatic children trained to decrease facial muscle tension exhibited increases in PEFR. A reciprocal relationship also was demonstrated in normal adult males: individuals trained to increase facial muscle tension exhibited decreases in PEFR (Glaus and Kotses, in press).
* The research described in this article was supported, in part. by grant HL 27402 from the Division of Lung Diseases, National Institute of Heart, Lung. and Bood. Submitted by A.H. in partial fulfillment for the degree M.S., Department of Psychology, Ohio University. We wash to thank Alan J. Fridlund and an anonymous reviewer for their comments on an earlier version of this manuscript. ** Address requests for reprints to: Andrew Harver, Department of Psychology, Ohio University. Athens, Ohio 45701, U.S.A. 0301-0511/84/$3.00
0 1984, Elsevier Science Publishers
B.V. (North-Holland)
4
A. Harver and H. Kotses / Induced pulmonury
changes
The relationship between facial muscle tension and airways resistance does not appear to be a part of a more general relationship between muscle tension and respiration. Conditioned increases and decreases of forearm tension. for example, had no effect upon PEFR (Glaus and Kotses. in press). Additionally, changes in facial muscle tension were not related to changes in respiration rate (Glaus and Kotses, in press). Although these latter observations suggested that the association between facial muscle tension and bronchomotor tone is highly specific the degree to which the association is limited is unclear. The present experiment was concerned with the limits of the relationship between facial muscle tension and bronchomotor tone. This question was of interest because the mechanism or mechanisms mediating the association between facial muscle tension and bronchomotor tone will assume vastly different forms depending on the limits of the association. Accordingly, cardiovascular and pulmonary responses were examined during training for both facial muscle tension increases and decreases. For comparative purposes, cardiovascular and pulmonary responses also were examined during training of limb muscle tension increases and decreases.
2. Methods Sixty Ohio University male students, who ranged in age between 18-25, volunteered to participate in the study. All subjects were nonsmokers and all received course credit for participating in the study. Students were assigned randomly to one of six experimental groups based on the type of training administered: Frontal increase, frontal decrease, frontal noncontingent, limb increase, limb decrease, limb noncontingent. Facial and limb muscle activity (EMG) and heart rate were recorded with standard Ag/Ag-Cl Beckman electrodes filled with Beckman electrolyte and attached to the skin with Beckman adhesive collars. Facial muscle activity was picked up from subjects in the frontal groups by electrodes placed on the forehead above each eyebrow in accordance with the recommendations of Davis (1952). Forearm muscle activity was picked up from subjects in the limb groups by electrodes placed over the left brachioradialis muscle. For all subjects, heart rate was recorded through electrodes positioned over the right arm and left leg, and respiration was monitored with a strain gauge. All electrophysiological responses, as well as integrated EMG activity, were displayed on a Beckman Type RM Dynograph recorder. Peak expiratory flow rate was measured with a Wright Peak Flow Meter. Auditory feedback was provided in binary form to subjects in both the increase and decrease conditions by a Med-Associates ANL-190 voltage controlled oscillator and audio amplifier. During training, appropriate responding was not signalled, whereas incorrect responding resulted in the presentation of a 250 Hz tone. Noncontin-
A. Harver and H. Korses / Induced pulmonary
changes
5
gent subjects heard a tape recording of the feedback tone consisting of random presentations of the feedback signal. Feedback was delivered to subjects over Koss headphones. Each student participated in one recording session which consisted of a five-minute baseline period followed by a twenty-minute feedback period. Subsequent to entering the laboratory, the subject was greeted, and was asked to sign a consent form agreeing to participate in the experiment. The session began with instruction in the use of the Wright Peak Flow Meter. Following this, electrodes were attached to skin surfaces and three peak flow measures were taken, at one minute intervals, with the subject standing. The largest of these was recorded for analyses. A strain gauge then was positioned around the subject’s chest. The subject was asked to sit quietly during the baseline period and to attempt to keep the tone off as much as possible during the feedback period. After feedback training three additional peak flow measures were taken as before. Integrated EMG, average heart rate, and average respiration rate were determined for each minute of the baseline period and for each minute of the twenty-minute feedback period. Scores were analyzed with univariate analyses of variance testing effects due to Condition (increase, decrease, noncontingent) and Minutes (five or twenty) with repeated measures on the second factor, for both the frontal and the limb groups. Peak flow data were analyzed separately for both groups with an analysis of variance testing effects due to Condition (increase, decrease, noncontingent) and Period (pretraining, posttraining).
3. Results ’ 3. I. Frontal groups During the training period subjects in the frontal increase condition exhibited higher EMG levels than did subjects in the noncontingent condition. The EMG levels of subjects in the frontal decrease condition did not differ from those of subjects in the noncontingent condition. These observations, illustrated in fig. 1, were supported by a reliable Condition X Minutes interaction (F(38,513) = 2.23, MS, = 58.06) yielded by the frontal EMG feedback analysis. A test of the interaction (Cicchetti, 1972) indicated that the level of EMG activity of subjects in the increase condition was reliably different from the level of EMG activity in the noncontingent condition in minutes 11 through 20. Differences were not obtained at any point between the level of EMG activity exhibited by subjects in the decrease condition compared with those exhibited by subjects in the noncontingent condition. The analysis of ’ An alpha level of p i 0.05 was used in all statistical
analyses.
6
32. -
G 0 z 0” z 3 0 > :: 0
31. 30. 29.28. 27. -
-
INCREASE
-
NONCONTIN~E~T
-
DECREASE
26. 25. -24. 23. 22.21. 20.-
17. 16. 13. t&J 14.-
BLOCKS OF FOUR MINUTES Fig. 1. Facial EMG for the three experimental four minute blocks during feedback training.
conditions
averaged
over subjects
and successive
variance also yielded a reliable effect for Minutes (F(19513) = 4.05. IWS, = 105.21) which was due primarily to increases in muscle tension exhibited by subjects in the increase condition. Subjects trained to increase facial muscle tension showed a sigltificant decline in peak flow scores following training. This observation was supported by a reliable effect for Period (F(1,27) = 9.66, MS, = 806.67) and by a marginally significant Condition x Period interaction (F(2,27) = 3.22, MS, = 268.52) yielded by the peak flow analysis, as well as by a test of the interaction (Cicchetti, 1972). The PEFR changes for all groups are shown in fig. 2. Differences in neither respiration rate nor heart rate were observed during feedback training between subjects in the three experimental conditions. However, heart rate levels of subjects in all conditions increased across the minutes of training, independent of feedback contingency. This latter observation was supported by a main effect for Minutes (F(9,513) = 1.89, MS, = 11.41) yielded by the feedback training analysis. 3.2. Limb g-rozips As shown in fig, 3, subjects in the limb increase condition exhibited higher EMG levels than did subjects in the limb noncontingent condition during feedback training; subjects in the limb decrease condition did not differ
z”71 7cm
EMG
(MICROVOLT
SECONDS)
t
2 :>
3: f. :::
i\
:$ :.a
1
:$
:a: ::,
8
A. Haruer and H. Kotses / induced pulmonary
changes
reliably in terms of EMG activity from subjects in the noncontingent condition. These observations were supported by a reliable main effect for Condition (F(2,27) = 14.72, MS, = 122736.46) and by a reliable Condition X Minutes interaction (F(38,513) = 3.69, MS, = 1189.06) yielded by the feedback training analysis. A test of the interaction (Cicchetti, 1972) indicated that the level of EMG activity in the increase condition was reliably different from the level of EMG activity in the noncontingent condition in minutes 2 through 20; the EMG levels of subjects in the decrease condition did not differ reliably from those of subjects in the noncontingent condition at any point. The analysis of variance also yielded a reliable Minutes effect ( F(19,513) = 3.13, MS, = 1189.03) which appeared to be the result of the large increases in muscle tension exhibited by subjects in the increase condition. Peak flow scores exhibited by subjects in the three experimental conditions were not related to the hmb muscle tensicn changes observed; scores obtained after feedback training did not differ reliably from those obtained before training (fig. 2). Differences in neither respiration rate nor heart rate were observed between subjects in the three experimental conditions. However, heart rate levels of subjects in all conditions increased across the minutes of training, independent of feedback contingency. This latter observation was supported by a reliable effect for Minutes (F(l9,513) = 1.92, MS, = 9.47) yielded by the feedback analysis.
4. Discussion Consistent with earlier findings (Claus and Kotses, in press) we found that facial muscle tension increases were associated with decreases in PEFR whereas limb muscle tension increases were not. The average decline in peak flow scores exhibited by subjects in the facial increase condition was consistent in degree (15.2 liters/min) as well as in direction with reductions observed earlier. Facial tension increases, however, were not related either to respiration rate or to heart rate. The lack of respiratory rate changes in conjunction with the combination of facial muscle tension increases and PEFR decreases confirmed earlier findings (Glaus and Kotses, in press). The lack of concomitant heart rate changes, on the other hand, further supported the hypothesis that facial muscle tension and bronchomotor tone are related in a highly specific manner. Unfortunately, decreases in facial muscle tension were not observed as a function of feedback training. Therefore, under these conditions. changes in PEFR were not anticipated, and none were observed. Possibly, the nature of the feedback signal employed was responsible for our failure to observe facial muscle tension decreases. Feedback, in the present study, was provided in binary form; feedback, in research demonstrating conditioned muscular decreases (e.g. Glaus and Kotses, in press), was presented in analogue form.
A. Harver and H. Kotses / Induced pulmonary
changes
9
Although these two forms of feedback have not been compared in the training of muscular responses, analogue feedback is superior to binary feedback for the control of heart rate (Colgan, 1977; Lang and Twentyman, 1974). Despite our failure to train decreases in facial muscle tension, the present findings concerning the effects of facial tension increases are consistent with predictions from the facial muscle-pulmonary reflex hypothesis described earlier (Glaus and Kotses, in press; Kotses and Glaus, 1981). In that formulation, changes in afferent trigeminal activity, arising from changes in facial muscle tension, serve to influence bronchomotor tone via the modification of efferent vagal activity. A model limited to these characteristics suggests that changes in facial tension will have limited physiological effects. By demonstrating a lack of widespread effects of facial tension increases either on respiratory or cardiovascular activity the present findings supported this supposition. Several other considerations bear upon the facial muscle-pulmonary reflex hypothesis. From a supportive point of view, the reflex hypothesis is consistent with aspects of cranial nerve anatomy (Carpenter, 1976) and knowledge of bronchomotor regulation (Nadel, 1976; Tomori and Widdicombe, 1969). It is consistent also with evidence of reflexive connections between afferent vagal activity and facial motorneuron activity in cats (Tanaka and Asahara, 1980). On the negative side, the hypothesis may require modification for a number of reasons. First, in its present form the hypothesis specifies the site of origin of reflexive activity only in broad terms. To increase the precision of the reflex formulation, it is necessary to examine the effects of tension changes in individual facial muscles on bronchomotor tone. Second, the reflex hypothesis posits no mechanism to bridge the temporal gap from training to testing periods. A mechanism of this sort is needed since bronchomotor changes have not been evaluated during facial training. Third, a range of intervening variables, some of which have been considered earlier (Glaus and Kotses, in press), may in some manner contribute to the relationship between facial tension changes and bronchomotor tone. Clearly, these are major considerations for the reflex hypothesis; none, however, entirely eliminates the reflex possibility.
References Carpenter, M.B. (1976). Human Neuroanatomy (7th ed.). Williams and Wilkins: Baltimore. Cicchetti, D.U. (1972). Extension of multiple-range test to interaction tables in the analysis of variance: A rapid approximate solution. Psychological Bulletin. 77, 405-408. Colgan, M. (1977). Effects of binary and proportional feedback on bidirectional control of heart rate. Psychophysiology, 14, 187-191. Davis, J.F. (1952). Manual for Surface Electromyography. Laboratory for Psychological Studies, Allen Memorial Institute for Psychiatry: Montreal. Claus, K.D. and Kotses, H. (1983). Facial muscle tension influences lung airway resistance; limb muscle tension does not. Biological Psychology (in press).
10
A. Harcwr and H. Kotses
/ Induced
pulmonary
changes
Kotses, H. and Claus, K.D. (1981). Applications of biofeedback to the treatment of asthma: A critical review. Biofeedback and Self-Regulation. 6. 573-593. Kotses, H., Claus. K.D.. Bricel, S.K., Edwards, J.E. and Crawford. P.L. (1978). Operant muscular relaxation and peak expiratory flow rate in asthmatic children. Journal of Psychosomatx Research, 22, 17-23. Kotses, H., Claus, K.D., Crawford, P.L., Edwards, J.E. and Scherr, M.S. (1976). Operant reduction of frontalis EMG activity in the treatment of asthma in children. Journal of Psychosomatic Research, 20. 453-459. Lang, P.J. and Twentyman, CT. (1974). Learning to control heart rate: Binary vs. analogue feedback. Psychophysiology, 11, 616-629. Nadel, J.A. (1976). Airways: Autonomic regulation and airway responsiveness. In: Weiss, E.B. and Segal, MS. (Eds.). Bronchial Asthma: Mechanisms and Therapeutics. Little. Brown and Co: Boston. Tanaka, T. and Asahara. T. (1980). Synaptic activation of vagal afferents on facial motorneurons in the cat. Brain Research, 212, 188-193. Tomori. Z. and Widdicombe, J.G. (1969). Muscular, bronchomotor, and cardiovascular reflexes elicited by mechanical stimulation of the respiratory tract. Journal of Physiology. 200. 25-49.
Bdogical Ps.schologv 18 (1984) 3-10 North-Holland
PULMONARY Andrew Depurtment Accepted
CHANGES
HARVER
INDUCED
BY FRONTAL
EMG TRAINING
*
** and Harry KOTSES
of Psychology, Ohio University, Athens, OH 45701, U.S.A for publication
20 June 1983
Earlier research suggested that the effects of facial muscle tension changes on other responses are not widespread but limited to a rather narrow set of pulmonary events. Further evidence in support of the specificity of the facial muscle-pulmonary relationship was provided in the present study by monitoring changes in several responses as a function of muscle tension training. Feedback training for increases and decreases in muscle tension at both facial and limb muscle sites was given to adult males. The effects of these manipulations on PEFR, RR, and HR were examined. Increases in facial muscle tension resulted in PEFR decreases whereas increases in limb muscle tension did not. Decreases in facial muscle tension were not observed as a function of training and no PEFR changes resulted from these conditions. Neither RR nor HR were related to the facial EMG changes observed during feedback training. These observations demonstrated the specificity inherent in the relationship between facial muscle tension and PEFR. and lent support to the hypothesis that these two responses are linked reflexively.
1. Introduction In a series of related studies (Glaus and Kotses, in press; Kotses, Glaus, Bricel, Edwards and Crawford, 1978; Kotses, Glaus, Crawford, Edwards and Scherr, 1976) a systematic relationship was described between changes in facial muscle tension and changes in peak expiratory flow rate (PEFR), the latter an indirect measure of airways resistance. Both normal adults and asthmatic children trained to decrease facial muscle tension exhibited increases in PEFR. A reciprocal relationship also was demonstrated in normal adult males: individuals trained to increase facial muscle tension exhibited decreases in PEFR (Glaus and Kotses, in press).
* The research described in this article was supported, in part. by grant HL 27402 from the Division of Lung Diseases, National Institute of Heart, Lung. and Bood. Submitted by A.H. in partial fulfillment for the degree M.S., Department of Psychology, Ohio University. We wash to thank Alan J. Fridlund and an anonymous reviewer for their comments on an earlier version of this manuscript. ** Address requests for reprints to: Andrew Harver, Department of Psychology, Ohio University. Athens, Ohio 45701, U.S.A. 0301-0511/84/$3.00
0 1984, Elsevier Science Publishers
B.V. (North-Holland)
4
A. Harver and H. Kotses / Induced pulmonury
changes
The relationship between facial muscle tension and airways resistance does not appear to be a part of a more general relationship between muscle tension and respiration. Conditioned increases and decreases of forearm tension. for example, had no effect upon PEFR (Glaus and Kotses. in press). Additionally, changes in facial muscle tension were not related to changes in respiration rate (Glaus and Kotses, in press). Although these latter observations suggested that the association between facial muscle tension and bronchomotor tone is highly specific the degree to which the association is limited is unclear. The present experiment was concerned with the limits of the relationship between facial muscle tension and bronchomotor tone. This question was of interest because the mechanism or mechanisms mediating the association between facial muscle tension and bronchomotor tone will assume vastly different forms depending on the limits of the association. Accordingly, cardiovascular and pulmonary responses were examined during training for both facial muscle tension increases and decreases. For comparative purposes, cardiovascular and pulmonary responses also were examined during training of limb muscle tension increases and decreases.
2. Methods Sixty Ohio University male students, who ranged in age between 18-25, volunteered to participate in the study. All subjects were nonsmokers and all received course credit for participating in the study. Students were assigned randomly to one of six experimental groups based on the type of training administered: Frontal increase, frontal decrease, frontal noncontingent, limb increase, limb decrease, limb noncontingent. Facial and limb muscle activity (EMG) and heart rate were recorded with standard Ag/Ag-Cl Beckman electrodes filled with Beckman electrolyte and attached to the skin with Beckman adhesive collars. Facial muscle activity was picked up from subjects in the frontal groups by electrodes placed on the forehead above each eyebrow in accordance with the recommendations of Davis (1952). Forearm muscle activity was picked up from subjects in the limb groups by electrodes placed over the left brachioradialis muscle. For all subjects, heart rate was recorded through electrodes positioned over the right arm and left leg, and respiration was monitored with a strain gauge. All electrophysiological responses, as well as integrated EMG activity, were displayed on a Beckman Type RM Dynograph recorder. Peak expiratory flow rate was measured with a Wright Peak Flow Meter. Auditory feedback was provided in binary form to subjects in both the increase and decrease conditions by a Med-Associates ANL-190 voltage controlled oscillator and audio amplifier. During training, appropriate responding was not signalled, whereas incorrect responding resulted in the presentation of a 250 Hz tone. Noncontin-
A. Harver and H. Korses / Induced pulmonary
changes
5
gent subjects heard a tape recording of the feedback tone consisting of random presentations of the feedback signal. Feedback was delivered to subjects over Koss headphones. Each student participated in one recording session which consisted of a five-minute baseline period followed by a twenty-minute feedback period. Subsequent to entering the laboratory, the subject was greeted, and was asked to sign a consent form agreeing to participate in the experiment. The session began with instruction in the use of the Wright Peak Flow Meter. Following this, electrodes were attached to skin surfaces and three peak flow measures were taken, at one minute intervals, with the subject standing. The largest of these was recorded for analyses. A strain gauge then was positioned around the subject’s chest. The subject was asked to sit quietly during the baseline period and to attempt to keep the tone off as much as possible during the feedback period. After feedback training three additional peak flow measures were taken as before. Integrated EMG, average heart rate, and average respiration rate were determined for each minute of the baseline period and for each minute of the twenty-minute feedback period. Scores were analyzed with univariate analyses of variance testing effects due to Condition (increase, decrease, noncontingent) and Minutes (five or twenty) with repeated measures on the second factor, for both the frontal and the limb groups. Peak flow data were analyzed separately for both groups with an analysis of variance testing effects due to Condition (increase, decrease, noncontingent) and Period (pretraining, posttraining).
3. Results ’ 3. I. Frontal groups During the training period subjects in the frontal increase condition exhibited higher EMG levels than did subjects in the noncontingent condition. The EMG levels of subjects in the frontal decrease condition did not differ from those of subjects in the noncontingent condition. These observations, illustrated in fig. 1, were supported by a reliable Condition X Minutes interaction (F(38,513) = 2.23, MS, = 58.06) yielded by the frontal EMG feedback analysis. A test of the interaction (Cicchetti, 1972) indicated that the level of EMG activity of subjects in the increase condition was reliably different from the level of EMG activity in the noncontingent condition in minutes 11 through 20. Differences were not obtained at any point between the level of EMG activity exhibited by subjects in the decrease condition compared with those exhibited by subjects in the noncontingent condition. The analysis of ’ An alpha level of p i 0.05 was used in all statistical
analyses.
6
32. -
G 0 z 0” z 3 0 > :: 0
31. 30. 29.28. 27. -
-
INCREASE
-
NONCONTIN~E~T
-
DECREASE
26. 25. -24. 23. 22.21. 20.-
17. 16. 13. t&J 14.-
BLOCKS OF FOUR MINUTES Fig. 1. Facial EMG for the three experimental four minute blocks during feedback training.
conditions
averaged
over subjects
and successive
variance also yielded a reliable effect for Minutes (F(19513) = 4.05. IWS, = 105.21) which was due primarily to increases in muscle tension exhibited by subjects in the increase condition. Subjects trained to increase facial muscle tension showed a sigltificant decline in peak flow scores following training. This observation was supported by a reliable effect for Period (F(1,27) = 9.66, MS, = 806.67) and by a marginally significant Condition x Period interaction (F(2,27) = 3.22, MS, = 268.52) yielded by the peak flow analysis, as well as by a test of the interaction (Cicchetti, 1972). The PEFR changes for all groups are shown in fig. 2. Differences in neither respiration rate nor heart rate were observed during feedback training between subjects in the three experimental conditions. However, heart rate levels of subjects in all conditions increased across the minutes of training, independent of feedback contingency. This latter observation was supported by a main effect for Minutes (F(9,513) = 1.89, MS, = 11.41) yielded by the feedback training analysis. 3.2. Limb g-rozips As shown in fig, 3, subjects in the limb increase condition exhibited higher EMG levels than did subjects in the limb noncontingent condition during feedback training; subjects in the limb decrease condition did not differ
z”71 7cm
EMG
(MICROVOLT
SECONDS)
t
2 :>
3: f. :::
i\
:$ :.a
1
:$
:a: ::,
8
A. Haruer and H. Kotses / induced pulmonary
changes
reliably in terms of EMG activity from subjects in the noncontingent condition. These observations were supported by a reliable main effect for Condition (F(2,27) = 14.72, MS, = 122736.46) and by a reliable Condition X Minutes interaction (F(38,513) = 3.69, MS, = 1189.06) yielded by the feedback training analysis. A test of the interaction (Cicchetti, 1972) indicated that the level of EMG activity in the increase condition was reliably different from the level of EMG activity in the noncontingent condition in minutes 2 through 20; the EMG levels of subjects in the decrease condition did not differ reliably from those of subjects in the noncontingent condition at any point. The analysis of variance also yielded a reliable Minutes effect ( F(19,513) = 3.13, MS, = 1189.03) which appeared to be the result of the large increases in muscle tension exhibited by subjects in the increase condition. Peak flow scores exhibited by subjects in the three experimental conditions were not related to the hmb muscle tensicn changes observed; scores obtained after feedback training did not differ reliably from those obtained before training (fig. 2). Differences in neither respiration rate nor heart rate were observed between subjects in the three experimental conditions. However, heart rate levels of subjects in all conditions increased across the minutes of training, independent of feedback contingency. This latter observation was supported by a reliable effect for Minutes (F(l9,513) = 1.92, MS, = 9.47) yielded by the feedback analysis.
4. Discussion Consistent with earlier findings (Claus and Kotses, in press) we found that facial muscle tension increases were associated with decreases in PEFR whereas limb muscle tension increases were not. The average decline in peak flow scores exhibited by subjects in the facial increase condition was consistent in degree (15.2 liters/min) as well as in direction with reductions observed earlier. Facial tension increases, however, were not related either to respiration rate or to heart rate. The lack of respiratory rate changes in conjunction with the combination of facial muscle tension increases and PEFR decreases confirmed earlier findings (Glaus and Kotses, in press). The lack of concomitant heart rate changes, on the other hand, further supported the hypothesis that facial muscle tension and bronchomotor tone are related in a highly specific manner. Unfortunately, decreases in facial muscle tension were not observed as a function of feedback training. Therefore, under these conditions. changes in PEFR were not anticipated, and none were observed. Possibly, the nature of the feedback signal employed was responsible for our failure to observe facial muscle tension decreases. Feedback, in the present study, was provided in binary form; feedback, in research demonstrating conditioned muscular decreases (e.g. Glaus and Kotses, in press), was presented in analogue form.
A. Harver and H. Kotses / Induced pulmonary
changes
9
Although these two forms of feedback have not been compared in the training of muscular responses, analogue feedback is superior to binary feedback for the control of heart rate (Colgan, 1977; Lang and Twentyman, 1974). Despite our failure to train decreases in facial muscle tension, the present findings concerning the effects of facial tension increases are consistent with predictions from the facial muscle-pulmonary reflex hypothesis described earlier (Glaus and Kotses, in press; Kotses and Glaus, 1981). In that formulation, changes in afferent trigeminal activity, arising from changes in facial muscle tension, serve to influence bronchomotor tone via the modification of efferent vagal activity. A model limited to these characteristics suggests that changes in facial tension will have limited physiological effects. By demonstrating a lack of widespread effects of facial tension increases either on respiratory or cardiovascular activity the present findings supported this supposition. Several other considerations bear upon the facial muscle-pulmonary reflex hypothesis. From a supportive point of view, the reflex hypothesis is consistent with aspects of cranial nerve anatomy (Carpenter, 1976) and knowledge of bronchomotor regulation (Nadel, 1976; Tomori and Widdicombe, 1969). It is consistent also with evidence of reflexive connections between afferent vagal activity and facial motorneuron activity in cats (Tanaka and Asahara, 1980). On the negative side, the hypothesis may require modification for a number of reasons. First, in its present form the hypothesis specifies the site of origin of reflexive activity only in broad terms. To increase the precision of the reflex formulation, it is necessary to examine the effects of tension changes in individual facial muscles on bronchomotor tone. Second, the reflex hypothesis posits no mechanism to bridge the temporal gap from training to testing periods. A mechanism of this sort is needed since bronchomotor changes have not been evaluated during facial training. Third, a range of intervening variables, some of which have been considered earlier (Glaus and Kotses, in press), may in some manner contribute to the relationship between facial tension changes and bronchomotor tone. Clearly, these are major considerations for the reflex hypothesis; none, however, entirely eliminates the reflex possibility.
References Carpenter, M.B. (1976). Human Neuroanatomy (7th ed.). Williams and Wilkins: Baltimore. Cicchetti, D.U. (1972). Extension of multiple-range test to interaction tables in the analysis of variance: A rapid approximate solution. Psychological Bulletin. 77, 405-408. Colgan, M. (1977). Effects of binary and proportional feedback on bidirectional control of heart rate. Psychophysiology, 14, 187-191. Davis, J.F. (1952). Manual for Surface Electromyography. Laboratory for Psychological Studies, Allen Memorial Institute for Psychiatry: Montreal. Claus, K.D. and Kotses, H. (1983). Facial muscle tension influences lung airway resistance; limb muscle tension does not. Biological Psychology (in press).
10
A. Harcwr and H. Kotses
/ Induced
pulmonary
changes
Kotses, H. and Claus, K.D. (1981). Applications of biofeedback to the treatment of asthma: A critical review. Biofeedback and Self-Regulation. 6. 573-593. Kotses, H., Claus. K.D.. Bricel, S.K., Edwards, J.E. and Crawford. P.L. (1978). Operant muscular relaxation and peak expiratory flow rate in asthmatic children. Journal of Psychosomatx Research, 22, 17-23. Kotses, H., Claus, K.D., Crawford, P.L., Edwards, J.E. and Scherr, M.S. (1976). Operant reduction of frontalis EMG activity in the treatment of asthma in children. Journal of Psychosomatic Research, 20. 453-459. Lang, P.J. and Twentyman, CT. (1974). Learning to control heart rate: Binary vs. analogue feedback. Psychophysiology, 11, 616-629. Nadel, J.A. (1976). Airways: Autonomic regulation and airway responsiveness. In: Weiss, E.B. and Segal, MS. (Eds.). Bronchial Asthma: Mechanisms and Therapeutics. Little. Brown and Co: Boston. Tanaka, T. and Asahara. T. (1980). Synaptic activation of vagal afferents on facial motorneurons in the cat. Brain Research, 212, 188-193. Tomori. Z. and Widdicombe, J.G. (1969). Muscular, bronchomotor, and cardiovascular reflexes elicited by mechanical stimulation of the respiratory tract. Journal of Physiology. 200. 25-49.