- Email: [email protected]
Heartbeat perception, instructions, and biofeedback in the control of heart rate
179
International Journal of Psychophysiology, 11 (1991) 179-193 c; 1991 Elsevier Science Publishers B.V. 0167~8760/91/$03.50
PSYCHO
00341
Heartbeat perception, instructions, and biofeedback in the control of heart rate Vilfredo De Pascalis, Giovanni Palumbo and Viviana Ronchitelli DIpnrtimento
di Psicologia, University of Rome, ‘La Sapienza’, (Accepted
Key wordr:
Heartbeat
perception;
2 October
Rome (It&)
1990)
Heart rate: HR-feedback;
Instruction
The present study was designed to examine the possibility that individual differences in heartbeat perception and instructions to control heart rate (HR) may influence the acquisition of voluntary control. Good (n = 20) and poor (n = 20) perceivers of cardiac activity were selected on the basis of their performance according to Whitehead et al. (1977) heartbeat discrimination procedure. Measures of state and trait anxiety (the State-Trait Anxiety Inventory Form X-l and Form X-2) and Tellegen’s Absorption Scale (TAS) were used to assess emotionality and absorptive ability. Good and poor heartbeat perceivers (a) were given non-motivating instructions to try to either increase or decrease their heart rate (HR) with, or (b) without the use of HR-feedback. and (c) were given motivating instructions to try to either increase or decrease their HR with, or (d) without HR-feedback. Heart rate. skin conductance (SC), and EMG activity were monitored. Subjects were also requested to indicate the cognitive strategies used during their HR control training. No relationship between heartbeat perception and state-trait anxiety measures was found. The results did not support the idea that individual differences in heartbeat perception are related to individual differences in HR-control. They did indicate, however, that motivating instructions improve the capacity to increase or decrease HR. Subjects were able to voluntarily increase or decrease their HRs with or without a feedback signal. However, more pronounced HR increases were obtained in the feedback as compared with the no-feedback condition. SC and EMG activity were in accordance with arousal levels demanded by HR decrease and increase tasks. Subjects used cognitive strategies concerning activation responses during HR-increase and relaxation responses during HR-decrease conditions.
INTRODUCTION Brener (1974, 1977) has suggested that the voluntary control of autonomic responses is possible when a person can discriminate differential activity of that response. In particular, he has proposed that the specificity with which a person complies with instructions to voluntary control an autonomic response is determined by the extent to which the person can discriminate modifications of that response independently of other ongoing activities. According to Brener the use of an exter-
Correspondence: V. De Pascalis, Dipartimento Via degli Apuli 8, 00185 Roma, Italy.
di Psicologia,
nal biofeedback signal improves response discrimination so that voluntary response control is facilitated. Moreover, according to Brener, persons who are more able than others to discriminate a visceral response (or persons who are given discrimination training to improve their discrimination capacity) should also be more able to control that response. This assumption has produced a noticeable interest in biofeedback research, stimulating the elaboration of appropriate methods for assessing individual differences in the degree of visceral perception. The majority of the studies have considered the self-perception of heart rate as an index of visceral self-perception, and this research usually has been concerned with the relationship between self-
180
awareness and self-control of heart beat (or heart rate). One line of research is focused on the examination of predictive relationships between subjects’ discrimination performance prior to any training and their performance on the control of HR using biofeedback training. According to Brener’s assumptions the process of learning to control HR involves organizing an appropriate response image. The extent to which subjects possess such an image before any training (i.e., their level of HR discrimination performance) may be expected to influence their control of HR during subsequent HR-biofeedback training. Among the reports which have verified this assumption, some have observed significant positive relationships between.pre-training discrimination and performance on later attempts to increase HR, but not with attempts to decrease HR (McFarland, 1975; Clemens & MacDonald, 1978). Using a similar method to McFarland’s (1975), De Pascalis, Alberti and Pandolfo (1984) failed to find a relationship between HR-perception and performance on HR increase trials, but, contrary to previous findings, a significant and positive correlation with performance on attempts to decrease HR was found. Conflicting results were also found in three experiments presented in the study of Whitehead, Drescher, Heiman & Blackwell (1977) wherein a negative correlation in one case, and no predictive relationships between cardiac discrimination and HR control in the other cases, were found. No predictive relationships, in this respect, were reported by Lacroix and Gowen (1981). Lacroix (1981) reviewed the literature regarding Brener’s theory and found generally positive evidence with respect to the assumption that the control of a response implies an ability to discriminate that response. On the other hand, the two assumptions: (a) that the ability to discriminate a response is a sufficient condition for control of that response, and (b) that control of a response becomes more specific over the course of training, were not supported by the experimental evidence. However, this author also observed that a large differentiation in the methods among investigations made comparisons among studies difficult. A number of objective techniques for the as-
sessment of visceral self-perception have been developed (Brener & Jones, 1974; McFarland, 1975; Whitehead et al., 1977; Ashton, White and Hodgson, 1979; Schandry & Specht, 1981). However, Yates, Jones, Marie and Hogben (1985) found that the heartbeat was judged as being coincident with light flashes occurring between 200 and 400 ms after the R-wave. These authors found marked individual differences in the patterns of responding, indicating that a single criterion for heartbeat detection is not a satisfactory means of assessing detection ability. In this study we chose to measure heartbeat (HB) perception using Whitehead’s technique, because HB perception measured with this approach is independent from artifacts that might be attributable to motor behavior or self-instruction. The aim of this experiment was to evaluate Brener’s prediction that good HB perceivers, as compared with poor HB perceivers, would display better performance during training to increase or to decrease their HR when provided with external feedback. Cuthbert, Kristeller, Simons, Hodes, and Lang (1981) reported a series of four experiments devoted to assess the effects of instructions to lower HR on HR change and general arousal reduction. Various conditions of biofeedback, cognitive load, incentive. knowledge of results and the experimenter-subject relationship were tested. These authors reported that informational and motivational components of the stimulus, and the interpersonal relationship between subject and experimenter, were interacting factors for achieving relaxation or HR lowering. These findings demonstrated the power of instructions and psychosocial/motivational variables in producing physiological changes during relaxation experiments. The work by Qualls and Sheehan (1979, 1981) suggested that capacity for absorption (Tellegen and Atkinson, 1974) reliably predicted degree of biofeedback learning in the case in which the task was a decrease in frontalis electromyogram (EMG) activity. In high absorption subjects there was no difference in performance between biofeedback and no-biofeedback tasks, whereas low absorption subjects performed better when provided with feedback. A later study by Offutt and Lacroix
181
(1983) confirmed Qualls and Sheehan’s findings with skin temperature and with different training procedures over the entire range of absorption scores. In the light of these last findings the aim of the present study was also to investigate the effects of motivating and non-motivating instructions on somatic and autonomic changes, while subjects tried to either increase or decrease their heart rate with or without HR-feedback. Skin conductance and frontalis muscle EMG activities were monitored throughout HR control conditions. We expected larger sympathetically-mediated responses and greater degree of HR control when subjects received motivating instructions to control their HR, rather than in those who were not given motivating instructions. The last but not least purpose of this study was to evaluate the relationship between absorption and heartbeat perception defined using the Whitehead et al. (1977) paradigm.
METHOD Subjects 78 female students (19 to 30 year age range) at the University of Rome volunteered to take part in a screening session in which the Whitehead et al. (1977) HB discrimination task was administered. Two extreme discrimination groups were selected on the basis of their performance on the HB discrimination test. The group of good HB perceivers consisted of 20 females achieving a discrimination index of at least 0.25 (approx. 65% correct decisions) or greater, whereas the group of poor HB perceivers consisted of 20 females achieving a discrimination index of 0.1 (approx. 60% correct decisions) or less. (A discrimination index of 0.0 indicates change performance; 50% correct decisions). Mean discrimination index score for the group of good HB perceivers was 0.45 + 0.19 (approx. 82% correct decisions), while mean score for the group of poor HB perceivers was 0.03 + 0.02 (approx. 55% correct decisions). The subjects’ mean age was 23 years, ranging from 19 to 26. Subjects were not allowed to participate if they had a history of heart trouble or were taking any form of medication.
Experimental design The experiment was carried out in three sessions. In the first, subjects (n = 78) were administered a HB discrimination task and divided into good (n = 20) and poor (n = 20) perceivers. In a second session, 5 good and 5 poor HB perceivers were assigned in random order to each of the following four conditions in the study: (a) nonmotivating instructions at increasing (or decreasing) HR with feedback, (b) motivating instructions at increasing (or decreasing) HR with feedback, (c) non-motivating instructions at increasing (or decreasing) HR without feedback, (d) motivating instructions at increasing (or decreasing) HR without feedback. In session 3 subjects were required to decrease HR if they increased HR in the preceding session, and vice versa. The order of administration of the increasing-decreasing-HR session was balanced across subjects. The basic analysis utilized for each physiological variable was an ANOVA or ANCOVA according to the following factorial design: 2 (Good, Poor HB-perception) X 2 (Motivating, Non-motivating instructions) X 2 (Feedback, No-Feedback) x 2 (HR increase, HR decrease) x 5 Trials with repeated measures on the last two factors. Significant effects were assessed using the Greenhouse-Geisser epsilon correction for inflated probability of a type I error. A rejection region at P -C 0.05 was used throughout. Post-hoc comparisons of the means were carried out by Duncan’s Multiple Range Test. Apparatus During each experimental session the subject sat in a comfortable reclining chair. A panel mounted in front of the subject contained two signal lights to indicate the beginning of the HB discrimination task (red-light) and the rest periods (blue-light). During the HB discrimination task a green light flashed in synchrony with the HB for 10-s as described below. The apparatus for HB discrimination contained two push buttons for response indicators during tasks. The electrocardiogram (EKG) was recorded using two electrodes placed on the lateral surfaces of the subject’s upper arms. The EKG and the occurrence of experimental events were detected and displayed
182
by an electronic system * custom built according to Whitehead et al. (1977) method. Interbeat intervals (IBIS) were evaluated to an accuracy of 1 ms. EKG R-waves were detected and timed by feeding the EKG to a Schmitt-trigger and detecting the Schmitt-trigger output. HR. skin conductance and EMG were recorded, during HR biofeedback training, using SATEM biofeedback equipment. A PT601 SATEM module cardiotachometer computed the beats per minute heart rate. HR was evaluated on a beat-by-beat basis by detecting the peripheral pulse. The peripheral pulse was detected from the volar surface of the distal phalanx of the second forefinger of the nondominant hand by a photoelectric plethysmographic transducer. Each amplified pulse signal was detected by a voltage-level detector and the interpulse intervals were evaluated with an accuracy of 1 ms and HR was then expressed in beat/mm with a 5-s interval. HR feedback was delivered through headphones and amplified from an audio amplifier. Audiotory feedback was given by a continuous series of beeps. A high HR level produced a high beep rate. As the HR level declined, the beep rate decreased. The beep rate of the feedback signal was contingent on, and immediately after, an interpulse interval variation (i.e., increase or decrease of HR). Skin conductance (SC). SC was recorded DC from two non-polarizing biopotential electrodes, l cm in diameter, with a KCL electrolyte. Electrodes were fastened over the volar surface of the second and third forefingers of the right hand. They were connected to a SATEM PT401 module that supplied through the electrodes a constant current density of 10 PA/cm to record the skin conductance response (SC). A time constant of 10 s was chosen. Frontal electromyographic (EMG) activity. Frontal EMG was monitored by centering non-polarizable biopotential electrodes 1.5 cm above each
* The system to record the HB discrimination data was designed by Dino Moretti and Pietro Fermani, technicians of the Department of Psychology. University of Rome, “La Sapienza”.
eyebrow, with a reference electrode in the center of the forehead. The diameter of the electrodes was 1.4 cm. A PT311S-SATEM EMG biofeedback module with standard amplification (EMG level from 0.1 to 10 microvolts and 70 to 1000 Hz bandwidth) was employed. Frontal EMG levels were integrated across 5 s. Materials The State-Trait Anxiety Inventory (STAI) Form X-l (State Anxiety) and X-2 (Trait Anxiety) (Spielberger, Gorsuch and Lushene, 1970) and the Tellegen Absorption Scale (TAS, Tellegen and Atkinson, 1974) were used to assess emotionality and absorptive ability of the subjects, respectively. Following HR feedback training, subjects were also requested to fill out Blanchard, Scott, Young, and Edmundson’s (1974) list regarding the cognitive strategies that they had used during their HR-control training. Procedure Subjects were screened and introduced to the experimental procedures during session 1. They were introduced to the experimental setting, showed the monitoring equipment, and were also informed about the phases of the experimental procedure. They were not, however, informed about specific experimental hypotheses. Subjects were then asked to complete the STAI From X-l to evaluate their state anxiety in the laboratory setting. Following completion of the inventory, subjects were required to sit in the reclining chair quietly during the session. Subjects were given taped instructions explaining the heartbeat discrimination task, according to the Whitehead et al. (1977) and Montgomery and Jones (1984) methods. They were instructed to press one of two buttons after each trial to indicate the occurrence of delayed (‘false’) or immediate (‘true’) HR feedback. The HB discrimination task consisted of 60 heartbeat discrimination trials each 10-s long. Subjects received two blocks of 30 trials each, with a 2-min rest period between. Half of the trials consisted of immediate feedback and the other half delayed feedback. and these were randomly distributed within each block. Immediate or true HB feedback was a 0.1-s flash occurring 128 ms
183
after each EKG R-wave *, delayed or false feedback was a light flash occurring 384 ms after each R-wave. The electronic system first presented to the subjects a blue light to indicate that the HB discrimination trial would soon begin. After the 10-s perception trials a yellow light came on to signal to the subjects that it was time to respond by pressing one of the two buttons to indicate whether they believed immediate or delayed feedback had occurred. The subjects were allowed as much time as they required to respond and the yellow light remained until they responded. The next trial began as soon as the subject responded. Subjects were given two push-buttons (one on the right and the other on the left arm of a recliner chair) labeled ‘true’ and ‘false’. During heartbeat trials they were required to respond by pressing the ‘true’ button if they felt that the HR light-signal they received was immediate or true heartbeat feedback; in the cases in which they felt that the HR signal was delayed or false, they were required to press the ‘false’ button. Subjects were not informed about their HB perception performance, nevertheless they were informed that they would obtain one point for each correct decision and that the subjects who exhibited the largest scores would be participants in the HR biofeedback training experience. At the end of the HB discrimination task, subjects were required to complete the trait anxiety inventory (STAI, Form X-2) and the Tellegen Absorption Scale (TAS). Heart-activity perception was measured using the following formula (Montgomery and Jones, 1984): HDI = (Nt/NT - Ntf/NF), where Nt/NT represents the proportion of immediate feedback trials to which the subject responds ‘true’ (Nt = number of ‘true’ responses given from the subjects; NT = number of actual ‘true’ trials) and Ntf/N F is the proportion of delayed feedback trials to which the subject responds ‘true’ (Ntf = number of delayed feedback trials that subject
* Whitehead et al. (1977) have assumed that the perception of heart contraction requires a delay of 128 ms (immediate or true feedback). This assumption is supported by the fact that mechanical contraction of the heart requires approx. 100 ms (150 ms for the blood pulse wave) to reach the wrist or neck following the R-wave.
identifies as ‘true’, N F = number of actual ‘false’ trials). As suggested by Montgomery and Jones (1984) this index is similar to Clemens’ (1979) d’ index of perceptual sensitivity in signal detection theory. HDI is taken here as an absolute value between 0 and 1; higher values indicate an increased heart-activity perception. This index is a linear probability measure which can be analyzed by means of parametric statistics. Sessions 2 and 3 were to test heart rate control, one for HR increase and the other for HR decrease. Following attachment of the photoelectric plethysmographic finger transducer and EMG and skin conductance electrodes, the experimenter gave the subjects tape-recorded experimental instructions according to the requirement of each experimental session. Common instructions to all subjects included a 5-min explanation about the focus of the research. The purpose of this period was to allow subjects to become accustomed to the physiological recording devices that had been attached to them. Subjects were also told to avoid holding their breath, breathing deeply or breathing in any other peculiar way, and they were requested to breathe at a steady regular rate. They were also requested not to tense or flex muscles and to move as little as possible in their chairs. After these instructions, resting level pulse rate was measured for a 3-min initial period. Following this period, the subjects’s instructions differed depending on the condition of the experimental design to which they had been assigned. Subjects received a combination of the instructions, described below, appropriate for their condition: HB perception factor. Subjects were divided into two groups, one of good HB perceivers (N = 20 Ss) and the other of poor HB perceivers (N = 20 Ss). Each of these subjects perceived one combination of the experimental treatments. Feedback-no feedback factor. Subjects either were provided with biofeedback and instructed to use it in attempting to control their heart rates or were not provided with biofeedback. In the feedback condition (FB) subjects were told that they would hear beeps through the headphones, and that the rate of the beeps would be determined by their HR. They were instructed to regulate HR using the beep as pulse rate information (i.e., a
184
beep rate increase represented a faster HR, and a beep rate decrease represented a slower HR). The acoustic intensity of the feedback signal was adjusted to the subject’s requirement. Subjects in the no feedback (NoFB) condition were not provided with any HR feedback and were specifically told to regulate HR by means of their own mental strategies. Motivating and non-motivating instructions. Subjects were given either motivating or non-motivating instructions before beginning each HR-control task. Non-motivated subjects were told that each person is able to acquire good HR control, and that they must not become worried if they did no produce very large HR changes because biofeedback research has demonstrated that even small changes are significant. The motivated subjects were given instructions emphasizing that, according to biofeedback research, persons who have learned to increase or to decrease their HR in the laboratory, are also able to control it in a stressful real-life situation. Moreover, they were informed that the acquisition of this capacity could reduce the probability of heart injures. To enhance the motivating effect, they were also told that there is evidence showing that a person obtains a good level of HR control when he/she shows a good level of task engagement. Increase-decrease factor. Subjects were instructed to try to either increase or decrease their HR. Subjects in each experimental condition received five 3-min training periods separated by 1-min rest periods. Training periods began with the illumination of a small red light in front of each subject. The red light was switched off at the
TABLE
I
Correlation matrix for state anxie+ (X-l), trait unxieiy (X-2), Tellegen absorption scale (TAS) and heartbeat discrimrnation index (HDI) Variables
X-l
x-2
0.501 * *
x.2
TAS
0.109
0.273
HDI
0.078
~ 0.100
* P < 0.05 * * P i 0.01 (two-tailed
test, df = 76).
TAS
end of each training period. The illumination of a small blue light indicated a rest period. Following the completion of the last period, subjects in all groups were given Blanchard et al.‘s list to assess the cognitive strategies they used during the training periods. Subjects were also asked to report thoughts or activities they had experienced during HR increase or HR decrease conditions.
RESULTS Anxiety, absorption and heartbeat perception To evaluate the influence of State Anxiety and Trait Anxiety (STAI, Forms X-l and X-2) and Absorptive ability (TAS) on HB perception scores, a multiple regression analysis across the overall group (N = 78) was carried out. No significant relationship was found. None of the zero-order correlation coefficients computed with respect to HDI scores was found to be significant (see Table I). The partial correlations revealed that anxiety and absorption variables accounted for a negligible portion of the total variance (less than 2%). Furthermore, to evaluate if good (N = 20) and poor (N = 20) perceivers differed in their anxiety or absorption levels, three separate one-way ANOVAs were performed. No significant differences were observed between groups for each of the variables considered (Trait Anxiety: Good = 41.0, S.D. = 9.25, Poor = 43.0, S.D. = 8.5; State Anxiety: Good = 40.7, S.D. = 8.4. Poor = 38.1, SD. = 7.3; TAS: Good = 7.2, S.D. =2.0, Poor= 8.2, S.D. = 2.7). Groups of subjects who scored high and low on each measure of anxiety and of absorption were then taken from the pool of all subjects (n = 78) and analyzed in three one-way ANOVA’s with respect to HB discrimination scores. No significant differences (P > 0.05) for HB discrimination scores between high and low subjects in trait anxiety, state anxiety and absorption were found (F(1.76) = 1.11. 2.98 and 0.17; P > 0.05 for Trait and State anxiety, and Absorption, respectively).
* PO.131
Physiological measures and correction for initial vulues Heart rate (HR), skin conductance (SC) and
T5
T4
T3
T2
Tl
Trials
Groups: HR.D
NoFB
1.8
(1.5)
5.2
(1.4)
(1.1)
- 1.9
1.5
(1.1)
0.3
(1.4)
3.9
(1.5)
(1.0)
(1.0)
-1.3
(1.2)
- 0.2
(1.3)
-0.7
(1.5)
-0.3
(1.2)
4.9
(1.3)
(1.5)
(1.2)
2.5
- 2.3
HR-I
0.4
HR-D
0.2
HR.I
(1.6)
- 0.8 (1.1)
5.8
(1.2)
6.8
(1.9)
- 4.8
3.4 (0.9)
(1.8)
1.8
3.5 (1.2)
- 5.8 (1.1)
- 3.7 (1.2)
- 3.2 (1.1)
- 3.6 (0.9)
0.6 (1.3)
7.1 (1.2)
HR.D
HR.I
(1.8)
- 2.3
(1.8)
3.4
1.2 (1.2)
1.1 (1.0)
5.2 (1.3)
1.5 (1.2)
5.3 (1.1)
HR.I
HR.D
(1.0)
- 3.6
(0.9)
0.4
(1.1)
- 2.7
(1.3)
- 3.4
(1.5)
- 0.9
(1.4)
3.7
(1.3)
2.9
(1.3)
3.0
(1.3)
4.4
(1.4)
3.0
HR.I
FB
4.4
3.4
3.4 (1.2) 2.7 (1.1)
- 1.1 (1.4) - 4.2 (1.6)
- 1.1 (1.7) - 0.2 (1.8)
1.5 (1.4)
(1.1)
- 3.2
(1.3)
1.6 (1.7)
1.2
(1.3)
(1.5)
(1.6)
-5.2
(1.4)
4.9
HR.I
- 4.3
(1.7)
(1.9)
- 2.4
- 1.2 (1.5)
HR.D
HR.I
FB
(1.5)
- 2.9
(1.4)
-5.0
(1.5)
- 3.3
(1.4)
-2.1
(1.5)
- 2.2
HR-D
Motivated subjects
(1.0)
- 1.5
(1.0)
-1.5
(1.1)
1.2
HR-D
NoFB
subjects
Non-motiuated
NoFB
FB
FB
Non-motivated
subjects
Poor HB perceiuers
Good HB perceivers Motivated subjects
(1.3)
2.0
(1.2)
-1.4
(1.1)
-0.3
(1.2)
0.7
(1.5)
9.1
HR.I
NoFB
(0.8)
-2.1
(1.1)
0.7
(1.0)
-0.7
(0.9)
~ 1.4
(1.2)
-1.7
HR-D
Heart rate change (bpm) duringJive trials (Tl, T2, T3, T4, T5) J rom Jive pre-trial baselines in HR-increase (HR.1) and HR.decrease (HR-D) tasks for Good and Poor HB perceiuers who were provided and not provided with HR.feedback (FB, NoFB conditions) and whom were given non-motiuating and motivating instructions during HR-control tasks. Standard errors in parentheses.
TABLE II
1X6
frontalis EMG amplitude were quantified in 5-s time periods, and expressed in beats/min, microsiemens and microvolt, respectively. Each physiological measure was analyzed using Analysis of Covariance (ANCOVA) or Analysis of Variance (ANOVA) with a Split-Plot Factorial Design (see Kirk. 1968); ANCOVA was used in the cases in which the ‘Law of Initial Values’ (LIV, Lacey 1956; Lacey & Lacey, 1962) was present. LIV was determined to be present if, and only if, the correlation coefficient r evaluated between prestimulus level (s), and a poststimulus minus prestimulus difference level (d) was significant (Benjamin, 1967). ANOVA was used in the cases in which LIV was not considered to be present, that is. I was not significant. Two values of r were obtained for each variable: one in HR increase and the other in HR decrease condition. These values for HR data were not significant ( -0.08 and ~ 0.11. df= 38. P > 0.05, in HR increase and HR decrease condition. respectively). The correlation coefficient was significant in HR-increase for SC activity. and in HR-decrease condition for EMG activity (SC: 0.33, P < 0.05 and 0.08 N.S.: EMG: -0.03. N.S. and -0.38 P -C 0.05. for HR-increase and HR-decrease respectively). Effects of heurtheut perception, instructions und feedback WI HR control An analysis of variance on the mean HR scores for the minute preceding each trial showed that these means did not differ significantly between groups or interact with conditions. The only significant effect was a highly significant increase in HR over the 2nd, 3rd and 4th pre-trial baseline periods as compared with the 1st one (F(4.128) = 5.3, P < 0.001). In view of this. an attempt was made to allow for habituation and chance fluctuations in HR by subtracting these mean baseline HRs, preceding each trial, from the trial means to give the basic data for the main analysis (see McCanne and Sandman, 1975, for the application of this procedure). Table II reports HR difference scores across experimental conditions. The ANOVA for these HR data showed some significant interaction effects. The most powerful effects were for HR-control ( F(1.32) = 56.5, P < O.OOl), and for the Instruc-
tions x HR-control x Trial interaction ( F(4.128) = 5.6, P < 0.001). The first effect indicated (Duncan-test. P < 0.01) that, with respect to the running baseline periods, there were significant HR changes according to task requirements (i.e., the HR increased during HR-increase and decreased during HR-decrease conditions). The second interaction indicated that HR-control over trials was influenced by Instruction types. More specifically, motivated subjects as compared with non-motivated subjects, were more able to obtain HR changes in both HR-increase and HR-decrease trials. In particular, motivated subjects were able to obtain a significant increment or decrement of their HR across each of the five experimental trials with respect to baseline periods (motivated subjects: 8.6. 2.4. 2.4. 2.5. 2.9. for HR-increase trials, and ~ 1.1. - 1.7. - 2.8, - 1.8. - 3.6 for HR-decrease trials). This was not found to be true for non-motivated subjects who showed no HR change over trials 1, 3 and 4 during HR-increase, or over trials 1, 3, and 5 during HR-decrease task (non-motivated subjects; 0.7, 4.0. 0.6, 0.7, 2.1 for HR-increase trials and 0.6. -2.6, 0.6, - 2.7. -0.3, for HR-decrease trials). Finally. the Feedback x HR-control interaction was significant (F(1,32) = 7.8. P < 0.009). Comparisons among the means indicated that there were no differences in HR-reduction between Feedback and No-Feedback during the HR-decrease condition ( - 1.7 and ~ 1.4 for HR decrease/FB and HR-decrease/ No-Feedback conditions, respectively). In contrast, during HR-increase. the Feedback effect was more pronounced than NoFeedback (3.8 and 1.1 for HR increase/FB and HR increase/No Feedback conditions. respectively). No main or interactional effects for HBperception or Feedback factors were found. SC dutu. An analysis of variance on the mean SC activity for the baseline periods preceding each trial, indicated that motivated subjects had higher levels of SC activity than the non-motivated ones (16.8 vs. 11.4, respectively: F(1.32) = 7.2, P < 0.02). The other effect was for baseline periods. wherein a significant decrease of SC activity over the 3rd, 4th and 5th periods, as compared with the 1st and 2nd ones, was observed (15.1. 15.1. 13.5. 13.5. 13.2, respectively; F(4,128) = 17.0, P <
III
T5
T4
T3
T2
Tl
Trials
Groups.. HR-D
(0.6) - 1.1 (0.6) - 0.5 (0.6)
0.5 (1.2) - 0.5 (1.4) 1.4 (1.4)
- 0.8
(1.1)
~ 3.8
(0.9)
- 2.4
(1.0)
0.5
(1.2)
0.3
(0.9)
0.5
(0.8)
-1.0
(0.1)
(1.2)
-1.2
- 0.7
1.0
(1.1)
0.1 (0.8)
1.2 (1.1)
(1.1)
(0.1)
- 3.4
~ 0.4
2.2
HR.I
(1.1)
NoFB
HR.I
HR.D
FB
(1.4)
- 0.2
(1.4)
0.2
(1.4)
PO.2
(1.3)
0.4
(1.2)
1.7
HR-I
FB
(1.1)
- 0.4
(1.1)
0.0
(1.2)
1.5
(1.1)
0.0
(1.1)
~ 0.3
HR-D
NoFB
(1.2)
- 0.2
(1.3)
- 0.3
(1.2)
0.0
(1.2)
~ 0.4
(1.2)
0.8
HR-I
HR-D
2.2
(1.8)
- 0.7
(0.8)
1.9
(0.8)
0.3 (1.8)
(0.7)
- 2.8
(0.7)
~ 4.5
(1.1)
-1.5
(1.8)
-0.5
(1.8)
~ 0.6
(1.8)
- 0.2
HR.1
FB
- 0.3 (1.5)
0.0
(1.4)
0.0
(1.4)
-0.7
(1.4)
-0.8
(1.4)
-0.2
HR.I
NoFB
(1.1)
(1.1)
0.4
(1.1)
-1.2
(1.1)
1.4
(1.1)
- 2.6
HR.D
subjects
Poor HB percewers Non-motivated
Good HB perceiuers
Non-motruated subjects
Motivated SubJects
(1.1)
-2.0
(1.1)
- 0.2
0.1 (1.1)
(1.1)
(1.1)
(1.1) - 0.8
0.4
0.5
0.6 (1.1)
(1.1)
(1.1)
-0.5
(1.2)
-1.6
(1.1)
0.1
(1.2)
- 4.0
3.5 (1.3)
1.2
HR-D
subjects
(1.1)
HR-I
FB
~ 0.5
-1.7 (0.9)
HR-D
Motwated
(1.3)
- 0.5
(1.2)
-0.3
(1.3)
0.2
(1.3)
- 0.9
(1.3)
0.6
HR.I
NoFB
(1.4)
- 0.3
(1.4)
~ 0.7
(1.4)
0.7
(1.3)
- 1.2
(1.3)
-0.5
HR.D
Skin conductance change (pS) dunng five rrrals (Tl, T2, T3, T4, TS) from flue pre-trral baselines m HR.increase (HR.I) and HR.decrease (HR-D) tasks for Good and Poor HB perceivers who were provided and not provided with HR-feedback (FB, NoFB conditions) and whom were given non-motwatlng and motwating instructions during HR-control tasks. Standard errors in parentheses.
TABLE
IV
T5
T4
T3
T2
Tl
Trrals
Groups..
0.4 (0.6) - 0.2 (0.5)
0.4
(0.9) 0.3
(0.9)
1.5
2.2
(1.3)
(0.6)
0.7
0.3 (0.6)
(1.3)
(0.8)
(1.3)
-0.2
(1.0)
2.8
0.1
(0.8)
-0.07
(0.8)
(1.0)
0.0
(0.8) -0.5
0.5 (0.9)
- 0.2
1.1 (1.0)
(0.8)
1.4 (0.9)
~ 0.6
- 0.4 (0.8)
0.1
(1.0)
0.0
(0.9)
0.1
(0.5)
0.4
(0.7)
-1.3
(0.6)
-0.3
(0.6)
-0.4
(0.5)
-0.1
(0.6)
(1.2)
-1.3
(1.3)
0.6
(1.2)
1.3
(1.1)
0.7
1.2
0.7 (0.8)
(1.4)
(0.7) ~ 3.7
(1.3)
- 1.3
0.4 (0.7)
(1.2)
(0.5) - 1.6
(1.1)
-1.4
(1.5)
- 0.4
0.3 (0.5)
- 3.9
0.0 (1.1)
-0.1 (0.9)
0.8
(1.5)
HR.D
HR.I
HR-D
HR-I
HR.D
HR.I
(0.9)
~ 0.4
(0.9)
- 0.5
(0.9)
0.0
(0.8)
- 0.2
(0.9)
- 0.6
(0.9)
1.4
(0.9)
0.4
(0.9)
-0.4
(0.9)
-0.1
(0.8)
0.1
HR.1
NoFB
HR.I
HR.I
HR.D
FB
NoFB
FB
NoFB
FB HR.D
Non-motlr~ated subyxts
Non-motrocrted subjects
Poor HR perceuxrs Motwated subjecls
Good HB percewers
(0.8)
-1.2
(0.8)
-0.2
(0.6)
-1.0
(0.6)
PO.4
(0.8)
-1.2
HR.D
(1.0)
- 0.3
(1.1)
0.9
(1.0)
0.8
(1.1)
0.6
(1.1)
0.1
HR-I
FB
(1.0)
- 0.4
(1.1)
- 0.4
(1.0)
- 0.4
(1.1)
-1.2
(1.1)
-0.1
HR.D
Motioated subjects
(0.8)
0.1
(0.8)
0.1
(0.8)
0.5
(0.8)
0.5
(0.9)
- 0.3
HR.I
NoFB
(1.0)
~ 1.1
(1.1)
- 0.8
(1.0)
- 0.7
(1.1)
- 0.7
(1.1)
-0.5
HR.D
hasehnes ,,I HR.mcrease (HR.I) and HR.d ecreaw(HR.D) tasks for Good and Poor HB EMG amplttude change (pV) durrngfwetrials (TI, TZ, T.$ T4, TS) f ram fivepi-e-tnal perceroers rtmhowere prouded and not prorvded w:h HR.feedback (FB, NoFB condrtrons) und whom UWP grwn non-motwutrng cmd motwatmg mstructrons durrng HR-control tasks. Standard errors m parentheses.
TABLE
189
0.001). No other effects were found. SC baseline mean scores preceding each trial were substracted from trial means to give the basic data (difference scores) for statistical analysis. Difference scores for SC activity are reported in Table III. Owing to the fact that during the HR-increase task SC baseline levels were found to be related to SC change, an ANCOVA, using resting scores as covariate, was performed for SC difference scores across HR-increase and HR-decrease trials. A main effect for Instruction was found (F(1,31) = 4.2, P < 0.05) indicating a significant lowering of SC level during task performance for non-motivated subjects, while the motivated subjects did not show a significant change (-0.6 and 0.03, respectively). The Instruction X HR control X Trial interaction was significant (F(4.128) = 7.7, P < 0.001). This interaction indicated that with respect to baseline levels non-motivated subjects reduced their SC level over trials 1, 4 and 5 (- 1.0, -0.4, -0.6, -1.6 and -1.2, from the 1st to the 5th trial, respectively) in the HR-decrease condition. Meanwhile, no corresponding differences were observed between trials for motivated subjects with the exception of the 2nd trial in which a reduction of SC activity was observed (0.6, - 1.5, 0.5, -0.5, -0.4, from the 1st to the 5th trial, respectively). Moreover, HB perception x Instruction x HR control X Trial was significant (F(4,128) = 3.6, P < 0.001). Inspection of the means showed different trends between good and poor perceivers across HR-decrease trials for non-motivating and motivating instructions. In particular, good perceivers who were not motivated in HR reduction, significantly (P < 0.01) lowered SC activity across the last 4 trials with respect to the corresponding baseline levels (-0.37, -1.0, -0.9, -3.0, -1.4 from the 1st to the 5th trial, respectively), while they did not show SC change when they were motivated in controlling HR (-0.25, -0.3, 0.5, 0.1, -0.5 from the 1st to the 5th trial, respectively). Poor HB perceivers who were not motivated to reduce HR, in contrast, significantly reduced SC activity during the first and .fifth trials (- 2.2, 0.4, 0.3, - 0.2, - 1.0, respectively) while poor perceivers, who were motivated to reduce HR, increased SC activity during the first trial and reduced it during the
second - 0.3, tween across actional
and fourth trials (1.5, -2.6, 0.4. -1.1, respectively). No differential trends begood and poor HB perceivers were found HR-increase trials. No other main or intereffects were found. EMG data. The analysis of variance across baseline EMG scores did not evidence any main or interactional effect. As with HR and SC scores, the EMG difference scores were used as basic data for statistical analysis. Because EMG amplitude in the baseline condition was found to be related to EMG change during the HR-decrease task, an ANCOVA, using initial baseline EMG scores as covariates, was performed. There was a Trial x HR control interaction (F(4,128) = 7.24, P < 0.001). This interaction indicated a decrease of EMG amplitude during the HR-decrease condition in respect to baseline levels ( - 0.7, - 0.6, - 0.6, - 0.4, -1.1 from the 1st to the 5th trial, respectively), and in contrast a tendency of EMG amplitude to increase across the trials during the HR-increase task was found. Finally, Instruction x Trial (F(4,128) = 4.9, P < 0.05) and HB perception x Instruction X Trial interactions (F(4,128) = 4.4, P < 0.05) were significant. The last interaction indicated that good HB perceivers who were motivated in HR-control significantly lowered their EMG amplitude with respect to the baseline condition across four of the five trials (1.74, - 0.76, - 1.0, - 0.8, - 2.5 respectively), while no significant reductions across trials for non-motivated subjects were found (-0.08, -0.1, -0.4, 0.1, - 0.12, respectively). No significant differences across trials were found for the remaining experimental conditions. EMG difference scores during the experimental conditions are reported in Table IV. Cognitive mediation
Subjects were encouraged to try to either increase or decrease their HR by using their own cognitive strategies, but no mention was rn’ .1C about the type of strategy that they could adopt in the task. At the conclusion of each HR-control session, subjects filled out the 15-items list of cognitive strategies of Blanchard et al. (1974). Subjects were requested to indicate the cognitive
strategies used in controlling their HR, and when the strategy adopted by the subject was not listed on the questionnaire, they were requested to report it on the sheet. The items contained three categories of responses: (1) activation responses (cognitive activities concerning thoughts about motor activity, anxiety, fear. anger, or aggressive ideas; 6 items); (2) relaxation or trying to relax situations (5 items); (3) self-instruction about the physiological response to be controlled (focusing attention on body antecedents related to HR change, focusing attention on feedback signal, etc.; 4 items). To evaluate any differences between the number of activation, relaxation, and self-instruction responses reported by motivated versus nonmotivated subjects, two separate x2 tests were carried out for Feedback and No-feedback conditions, respectively. No significant differences were found (P < 0.05). When activation, relaxation, and self-instruction responses were compared in HRincrease versus HR-decrease tasks, significant differences for FB (x’ = 32.79. P < 0.001)and for NoFB conditions (x’ = 32.54, P -c0.001)were found. These results indicated that during the HR-increase task subjects reported significantly more activation than relaxation responses. while during the HR-decrease task they reported more relaxation than activation responses. Percentages of reported strategies, during HR-increase, were 6.8%. 52.7%, and 41.2% for relaxation, activation and self-instructions. respectively. During the HRdecrease task. corresponding percentages were 41.5%. 19.9% and 38.6%.
CONCLUSIONS The results of this study do not support a significant relationship between Whitehead’s HB perception measure and HR control. This negative finding is inconsistent with a prediction from Brener’s (1974, 1977) theory: that the presence of an appropriate response image should be a sufficient condition for control of that response. Our findings suggest that HB awareness prior to any training is not predictive of the performance on voluntary HR control. Reports in which measures of cardiac awareness were derived from HR track-
ing procedures have observed a significant relationship between initial discrimination performance and performance on subsequent attempts to increase HR. but not with performance on attempts to decrease HR (Kleinman, 1970; McFarland & Campbell, 1975; McFarland, 1975). Similar findings were obtained by Clemens & McDonald (1976) using the Brener-Jones procedure for assessing cardiac awareness. These findings contrasted with those obtained in one of our studies (De Pascalis et al.. 1984) wherein HB perception was evaluated with McFarland’s (1975) method. The HB perception/HR-increase relationship was nonsignificant, while it was significant for HR-decrease. Subjects who achieved high HB-perception scores decreased HR significantly better than low scorers. Moreover, conflicting results in three other experiments using the Whitehead procedure were also obtained (Whitehead. Drescher, Heiman & Blackwell, 1977): a negative correlation between cardiac discrimination and control in one case and no relationship in two cases. This suggests that the processes tapped by the Whitehead procedure and the Brener-Jones or HR tracking procedures are dissimilar. One possible explanation of the negative finding, as obtained in this study may be given if we assume that HR control is mainly dependent on efferent processes. Lacroix (1981) in his review of the literature concerning Brener’ theory, observed that there was little convincing experimental evidence to confirm Brener’s view. Lacroix presented a two-process theory of the acquisition of autonomic control in which the development of control of a response consists primarily in the identification of the efferent behavioral programmes already in the subject’s repertoire. Afferent processes play a role in the cases in which behavioral programmes, effective in the control of the target response, are not available to the subject. Discrepant results among studies may be dependent on the different methods used with respect to the measured responses, biofeedback training procedures, and discrimination testing procedures. In respect to this last, Ross and Brener (1981) evaluated the extent to which the Brener and Jones and Whitehead procedures probe the same general capacity in cardiac awareness. The
191
authors observed that subjects used different strategies in the solution of the two procedures: (1) the Brener and Jones procedure tended to induce an ‘active’ solution strategy which had a significant respiratory component characterized by a deep breath at the beginning of each discrimination trial and then holding the breath and remaining quiescent for the remainder of the trial; (2) the Whitehead procedure, in contrast, required the subject to adopt a strategy discouraging the production of HR variations and reduced sensory input. These authors observed that subjects who adopted an active discrimination strategy displayed significantly greater voluntary HR increase, in a post-discrimination test of HR control, than did subjects who adopted a passive discrimination strategy. Some other experiments have demonstrated that among light flashes (Yates et al., 1985) or tones (Brener and Kluvitse, 1988) occurring at intervals of 0, 100, 200, 300 and 400 ms after each R-wave, the stimuli that were 200 and 300 ms after the R-wave were judged to be most contiguous with heartbeat sensations. These findings are in contrast with the Whitehead et al. (1977) definition of heartbeat (i.e., a stimulus presented 128 ms after the R-wave was defined as a heartbeat, while a stimulus delivered 384 ms after R-wave was defined as occurring outside the events of the cardiac cycle). On the other hand, the experiment of Jones, Collins, Dabkowski and Jones (1988) has demonstrated that the ability to make exteroceptive temporal discriminations of the 128 and 384 ms time delays has little to do with the ability to make similar interoceptive discriminations with Whitehead’s paradigm. In view of the fact that Whitehead’s cardiac awareness procedure cannot be reduced to a simple discrimination of an exteroceptive stimulus pair, the different results among studies may be explained in terms of the different perceptual components measured with the Whitehead and Brener and Jones procedures. The lack of a relationship between HB perception and HB control found in this study is not attributable to the use of female subjects. Although a number of early reports found that males may be superior to females in making HB discriminations with Whitehead’s procedure (White-
head et al., 1977; Wildman & Jones, 1982) and during cardiac discrimination learning with knowledge-of-results training (Katkin, Blascovich & Goldband, 1981; Katkin, 1985) more recent studies (Jones, Jones, Rouse, Scott, & Caldwell, 1987; Rouse, Jones and Jones, 1988) have shown that gender differences in HB perception may be dependent on difference in body composition. Although we did not record respiration, we may exclude in this experiment a mediating role for respiration activity in the evaluation of cardiac awareness, because there is no evidence for such a relationship with Whitehead’s procedure (Jones et al., 1987; Rouse et al., 1988). SC and EMG activities were not found to be dependent on individual differences in HB perception, except for good perceivers who were motivated to reduce their HR; they evidenced no change in SC levels and a pronounced reduction in EMG activity across trials with respect to baseline levels. Although it was found that subjects increase or decrease their HR with or without feedback signal, analysis of variance also showed that the feedback condition was more effective than instructions alone in the degree of HR increase. In accordance with earlier reports, no differences between feedback and no-feedback conditions for HR-reduction trials were found (Blanchard & Young, 1972; Bouchard & Granger, 1977; Manuck, Levenson, Hinrichsen & Gryll, 1975; Levenson, 1976; Bouchard & Labelle, 1982). Results of this study confirm other findings (Cuthbert et al., 1981) suggesting that instructions play a principal role in the voluntary control of heart rate. Subjects who were given motivating instructions obtained greater HR changes, either among HR-increase or HR-decrease trials, as compared with subjects who were given nonmotivating instructions. SC and EMG activities during HR-reduction trials were parallel to the low arousal level in accordance with task requirements. Non-motivated subjects, whom we expected to be less activated, showed lower SC levels and no difference in EMG activity, as compared with baseline levels, in the HR-decrease condition. In this condition, motivated subjects exhibited a reduction in EMG activity and no change in SC levels. SC and EMG levels did not differ apprecia-
192
bly between non-motivated and motivated while subjects attempts to decrease HR.
groups
REFERENCES Ashton. R.. White, K.D.. and Hodgson. G. (1979) Sensitivity to heart rate: A psychophysical study. P.yvchoph?;siologv. 16: 463-466. BenJamin, L.S. (1967) Facts and artifacts in using analysis of covariance to ” undo” the law of initial values. Psychophysrolog\, 4: 187-202. Blanchard, E.B.. and Young, L. (1972) Relative efficacy of visual and auditory feedback for self-control of heart rate. .I. Gen. PsychoI.. 87: 195-202. Blanchard. E.B., Scott, R.W., Young, L.D.. and Edmundson, E.D. (1974) Effect of knowledge of response on the selfcontrol of heart rate. Psv~hoph_v.stolo~v. 11: 251-264. Bouchard. M.A., and Labelle, J. (1982) Voluntary heart rate deceleration: A critical evaluation. Biofeedback und Self Regulation. 7: 121-137. Bouchard. M.A.. and Granger. L. (1977) The role of instructions versus instructions plus feedback in voluntary heart rate slowing. PsJ,chophysrolom, 14: 475-482. Brener, J. (1974) A general model of voluntary control applied to the phenomena of learned cardiovascular change. In P.A. Obrist. A.H. Black. J. Brener, and L.V. DiCara (Eds.), Cardiocmc. Psychophysiol.. Chicago: Aldine. Pp. 365-391. Brener. J. (1977) Sensory and perceptual determinants of voluntary vlsceral control. In G.E. Schwartz and J. Beatty (Eds.). Btofeedbock: Theon und research. New York: Academic Press. pp. 29-66. Brener. J.. and Jones. J.M. (1974) Interoceptive discrimination in intact humans: Detection of cardiac activity. Phvsiol. Behar:.. 13: 763-767. Brener, J.. and Kluvitse, C. (1988) Heartbeat detection: Judgement of the simultaneity of external stimuli and heartbeats. P.yychoph.vsioloR). 25: 554-561. Clemens. W.J. (1979) Assessment, learning and retention of heart beat discrimination. Psychophysiology, 16: 333-338. Clemens. W.J., and MacDonald, D.F. (1976) Relationship betw’een heart beat discrimination and heart rate control (Abstract). Ps~~choph,~stologv. 13: 176. Cuthhert, B., Krlsteller. J., Simons, R., Hodes, R.. and Lang, P.J. (1981) Strategies of arousal control: Biofeedback, meditation, and motivation. J. L-p. Psycho/.: Cert., 110: 518-546. De Pascalis. V.. Alberti. M.L. and Pandolfo, R. (1984) Anxiety, perceptlox and control of heart rate. Percept. Mot. Skilh. 59: 203~211. Jones, G.E.. Jones, K.R., Rouse, C.H.. Scott, D.M., and Caldwell. J. (1987) The effect of body position on the perception of cardiac sensations: An experiment and theoretical implications. Psychoph_vstologv, 24: 300-311. Jones, G.E.. Collins, S.W.. Dahrowski. E.A.. and Jones. K.
(1988) The discriminahility of temporal patterns used in the Whitehead discrimination procedure. Ps.~,choph_vsioloRv, 25: 547-553. Katkin, E.S. (1985) Blood, sweat, and tears: Individual differences in autonomic self-perception. Psychoph_vsio/ogv, 22: 125-137. Katkin. ES. Blascovich, J.. and Goldband, S. (1981) Empirical assessment of visceral self-perception: Individual and sex differences in the acquisition of heartbeat discrimination. J. Personality Sot. PsychoI., 19: 568. Kirk. R.E. (1968) in Experimentul desrgn: Procedures for the hehoc~iorol sctences, Brooks/Cole Publishing Company, Belmont, pp. 245-318. Kleinman, R.A. (1970) The development of voluntary cardiovascular control. Unpublished doctoral dissertatmn. University of Tennessee. Lacey, J.1. (1956) The evaluation of autonomic responses: Toward a general solution. Ann. N. Y. Acad. Ser.. 67: 123-163. Lacey. J.I., and Lacey, B.C. (1962) The law of initial value in the longitudinal study of autonomic responses and response patterns over a four year interval. Ann. N. Y. Acud. Ser., 98: 1257-1290. Lacroix. J.M., and Gowen. A.H. (1981) The acquisition of autonomic control through biofeedback: Some tests of discrimination theory. P.r_vchophysio/ogy. 18: 559-572. Lacroix, J.M. (1981) The acquisition of autonomic control through biofeedback: The case against an afferent process and a two-process alternative. P.yychophysiologv. 18: 573587. Levenson. R.W. (1976) Feedback effects and respiratory involvement in voluntary control of heart rate. PsychophysrO/OR)‘,13: 108-114. Manuck. S.B., Levenson. R.W., Hinrichsen. J.J. and Gryll, S.L. (1975) Role of feedback in voluntary control of heart rate. Percept. Mot. Skilh, 40: 747-752. McCanne, T.R. and Sandman, C.A. (1975) Determinants of human operant heart rate conditioning: A systematic investigation of several methodological issues. J. Camp. Ph,sro/. Psycho/.. 88: 609-b18. McFarland, R.A. (1975) Heart rate perception and heart rate control. Psychophysrolom. 12: 402-405. McFarland. R.A., and Campbell. C. (1975) Precise heart-rate control and heart-rate perception. Percept. Mot., Skills, 41: 730. Montgomery, W.A., and Jones. G.E. (1984) Laterality, emotionality. and heartbeat perception. P.~_vchoph_vsto/ogv. 21: 459-465. Offutt. C., and Lacroix, J.M. (1983) Absorption: A predictor of biofeedback Iearnmg. (Abstract). P~~choph~~,stoloRv, 20: 4. Quails, P.J., and Sheehan. P.W. (1979) Capacity for absorption and relaxation during electromyograph biofeedback and no-feedback conditions. J. Ahnorm. Psycho/.. 88: 652-662. Quails, P.J., and Sheehan, P.W. (1981) Role of the feedback slgnal in electromyograph biofeedback: The relevance of attention. J. E.yp. P.y.vchol.: Gene&, 110: 204-216.
193 Ross, A. and Brener, J. (1981) Two procedures for training cardiac discrimination: A comparison of solution strategies and their relationship to heart rate control. Psychophysiolo~, 18: 62-70. Rouse, C.H., Jones, G.E. and Jones, K.R. (1988) The effect of body composition and gender on cardiac awareness. Psychophymlogy, 25: 400-407. Schandry, R., and Specht, G. (1981) The influence of psychological and physical stress on the perception of heartbeats (Abstract). Psychophysiology, 18: 154. Spielberger. C.D., Gorsuch, R.. and Lushene, R. (1970) The State-Trait Anxiety Inoentoty (STAI) Test Manual, Consulting Psychologists Press, Palo Alto.
Tellegen. A., and Atkinson, G. (1974) Openness to absorbing and selfaltering experiences (‘absorption’). a trait related to hypnotic susceptibility. J. Abnorm. Psychol., 83: 268-277. Whitehead, W.E., Drescher, V.M., Heiman. P., and Blackwell. B. (1977) Relation of heart rate control to heart beat perception. Biofeedback and Self-Regulation, 2: 371-392. Wildman, H.A., and Jones. G.E. (1982) Reliability of the Whitehead heartbeat discrimination procedure (Abstract). Psychophysiolog);, 19: 592. Yates. A.J., Jones, K.E., Marie, G.V. and Hogben, J.H. (1985) Detection of the heartbeat and events in the cardiac cycle. Psychophvsiologv, 22: 561-567.
International Journal of Psychophysiology, 11 (1991) 179-193 c; 1991 Elsevier Science Publishers B.V. 0167~8760/91/$03.50
PSYCHO
00341
Heartbeat perception, instructions, and biofeedback in the control of heart rate Vilfredo De Pascalis, Giovanni Palumbo and Viviana Ronchitelli DIpnrtimento
di Psicologia, University of Rome, ‘La Sapienza’, (Accepted
Key wordr:
Heartbeat
perception;
2 October
Rome (It&)
1990)
Heart rate: HR-feedback;
Instruction
The present study was designed to examine the possibility that individual differences in heartbeat perception and instructions to control heart rate (HR) may influence the acquisition of voluntary control. Good (n = 20) and poor (n = 20) perceivers of cardiac activity were selected on the basis of their performance according to Whitehead et al. (1977) heartbeat discrimination procedure. Measures of state and trait anxiety (the State-Trait Anxiety Inventory Form X-l and Form X-2) and Tellegen’s Absorption Scale (TAS) were used to assess emotionality and absorptive ability. Good and poor heartbeat perceivers (a) were given non-motivating instructions to try to either increase or decrease their heart rate (HR) with, or (b) without the use of HR-feedback. and (c) were given motivating instructions to try to either increase or decrease their HR with, or (d) without HR-feedback. Heart rate. skin conductance (SC), and EMG activity were monitored. Subjects were also requested to indicate the cognitive strategies used during their HR control training. No relationship between heartbeat perception and state-trait anxiety measures was found. The results did not support the idea that individual differences in heartbeat perception are related to individual differences in HR-control. They did indicate, however, that motivating instructions improve the capacity to increase or decrease HR. Subjects were able to voluntarily increase or decrease their HRs with or without a feedback signal. However, more pronounced HR increases were obtained in the feedback as compared with the no-feedback condition. SC and EMG activity were in accordance with arousal levels demanded by HR decrease and increase tasks. Subjects used cognitive strategies concerning activation responses during HR-increase and relaxation responses during HR-decrease conditions.
INTRODUCTION Brener (1974, 1977) has suggested that the voluntary control of autonomic responses is possible when a person can discriminate differential activity of that response. In particular, he has proposed that the specificity with which a person complies with instructions to voluntary control an autonomic response is determined by the extent to which the person can discriminate modifications of that response independently of other ongoing activities. According to Brener the use of an exter-
Correspondence: V. De Pascalis, Dipartimento Via degli Apuli 8, 00185 Roma, Italy.
di Psicologia,
nal biofeedback signal improves response discrimination so that voluntary response control is facilitated. Moreover, according to Brener, persons who are more able than others to discriminate a visceral response (or persons who are given discrimination training to improve their discrimination capacity) should also be more able to control that response. This assumption has produced a noticeable interest in biofeedback research, stimulating the elaboration of appropriate methods for assessing individual differences in the degree of visceral perception. The majority of the studies have considered the self-perception of heart rate as an index of visceral self-perception, and this research usually has been concerned with the relationship between self-
180
awareness and self-control of heart beat (or heart rate). One line of research is focused on the examination of predictive relationships between subjects’ discrimination performance prior to any training and their performance on the control of HR using biofeedback training. According to Brener’s assumptions the process of learning to control HR involves organizing an appropriate response image. The extent to which subjects possess such an image before any training (i.e., their level of HR discrimination performance) may be expected to influence their control of HR during subsequent HR-biofeedback training. Among the reports which have verified this assumption, some have observed significant positive relationships between.pre-training discrimination and performance on later attempts to increase HR, but not with attempts to decrease HR (McFarland, 1975; Clemens & MacDonald, 1978). Using a similar method to McFarland’s (1975), De Pascalis, Alberti and Pandolfo (1984) failed to find a relationship between HR-perception and performance on HR increase trials, but, contrary to previous findings, a significant and positive correlation with performance on attempts to decrease HR was found. Conflicting results were also found in three experiments presented in the study of Whitehead, Drescher, Heiman & Blackwell (1977) wherein a negative correlation in one case, and no predictive relationships between cardiac discrimination and HR control in the other cases, were found. No predictive relationships, in this respect, were reported by Lacroix and Gowen (1981). Lacroix (1981) reviewed the literature regarding Brener’s theory and found generally positive evidence with respect to the assumption that the control of a response implies an ability to discriminate that response. On the other hand, the two assumptions: (a) that the ability to discriminate a response is a sufficient condition for control of that response, and (b) that control of a response becomes more specific over the course of training, were not supported by the experimental evidence. However, this author also observed that a large differentiation in the methods among investigations made comparisons among studies difficult. A number of objective techniques for the as-
sessment of visceral self-perception have been developed (Brener & Jones, 1974; McFarland, 1975; Whitehead et al., 1977; Ashton, White and Hodgson, 1979; Schandry & Specht, 1981). However, Yates, Jones, Marie and Hogben (1985) found that the heartbeat was judged as being coincident with light flashes occurring between 200 and 400 ms after the R-wave. These authors found marked individual differences in the patterns of responding, indicating that a single criterion for heartbeat detection is not a satisfactory means of assessing detection ability. In this study we chose to measure heartbeat (HB) perception using Whitehead’s technique, because HB perception measured with this approach is independent from artifacts that might be attributable to motor behavior or self-instruction. The aim of this experiment was to evaluate Brener’s prediction that good HB perceivers, as compared with poor HB perceivers, would display better performance during training to increase or to decrease their HR when provided with external feedback. Cuthbert, Kristeller, Simons, Hodes, and Lang (1981) reported a series of four experiments devoted to assess the effects of instructions to lower HR on HR change and general arousal reduction. Various conditions of biofeedback, cognitive load, incentive. knowledge of results and the experimenter-subject relationship were tested. These authors reported that informational and motivational components of the stimulus, and the interpersonal relationship between subject and experimenter, were interacting factors for achieving relaxation or HR lowering. These findings demonstrated the power of instructions and psychosocial/motivational variables in producing physiological changes during relaxation experiments. The work by Qualls and Sheehan (1979, 1981) suggested that capacity for absorption (Tellegen and Atkinson, 1974) reliably predicted degree of biofeedback learning in the case in which the task was a decrease in frontalis electromyogram (EMG) activity. In high absorption subjects there was no difference in performance between biofeedback and no-biofeedback tasks, whereas low absorption subjects performed better when provided with feedback. A later study by Offutt and Lacroix
181
(1983) confirmed Qualls and Sheehan’s findings with skin temperature and with different training procedures over the entire range of absorption scores. In the light of these last findings the aim of the present study was also to investigate the effects of motivating and non-motivating instructions on somatic and autonomic changes, while subjects tried to either increase or decrease their heart rate with or without HR-feedback. Skin conductance and frontalis muscle EMG activities were monitored throughout HR control conditions. We expected larger sympathetically-mediated responses and greater degree of HR control when subjects received motivating instructions to control their HR, rather than in those who were not given motivating instructions. The last but not least purpose of this study was to evaluate the relationship between absorption and heartbeat perception defined using the Whitehead et al. (1977) paradigm.
METHOD Subjects 78 female students (19 to 30 year age range) at the University of Rome volunteered to take part in a screening session in which the Whitehead et al. (1977) HB discrimination task was administered. Two extreme discrimination groups were selected on the basis of their performance on the HB discrimination test. The group of good HB perceivers consisted of 20 females achieving a discrimination index of at least 0.25 (approx. 65% correct decisions) or greater, whereas the group of poor HB perceivers consisted of 20 females achieving a discrimination index of 0.1 (approx. 60% correct decisions) or less. (A discrimination index of 0.0 indicates change performance; 50% correct decisions). Mean discrimination index score for the group of good HB perceivers was 0.45 + 0.19 (approx. 82% correct decisions), while mean score for the group of poor HB perceivers was 0.03 + 0.02 (approx. 55% correct decisions). The subjects’ mean age was 23 years, ranging from 19 to 26. Subjects were not allowed to participate if they had a history of heart trouble or were taking any form of medication.
Experimental design The experiment was carried out in three sessions. In the first, subjects (n = 78) were administered a HB discrimination task and divided into good (n = 20) and poor (n = 20) perceivers. In a second session, 5 good and 5 poor HB perceivers were assigned in random order to each of the following four conditions in the study: (a) nonmotivating instructions at increasing (or decreasing) HR with feedback, (b) motivating instructions at increasing (or decreasing) HR with feedback, (c) non-motivating instructions at increasing (or decreasing) HR without feedback, (d) motivating instructions at increasing (or decreasing) HR without feedback. In session 3 subjects were required to decrease HR if they increased HR in the preceding session, and vice versa. The order of administration of the increasing-decreasing-HR session was balanced across subjects. The basic analysis utilized for each physiological variable was an ANOVA or ANCOVA according to the following factorial design: 2 (Good, Poor HB-perception) X 2 (Motivating, Non-motivating instructions) X 2 (Feedback, No-Feedback) x 2 (HR increase, HR decrease) x 5 Trials with repeated measures on the last two factors. Significant effects were assessed using the Greenhouse-Geisser epsilon correction for inflated probability of a type I error. A rejection region at P -C 0.05 was used throughout. Post-hoc comparisons of the means were carried out by Duncan’s Multiple Range Test. Apparatus During each experimental session the subject sat in a comfortable reclining chair. A panel mounted in front of the subject contained two signal lights to indicate the beginning of the HB discrimination task (red-light) and the rest periods (blue-light). During the HB discrimination task a green light flashed in synchrony with the HB for 10-s as described below. The apparatus for HB discrimination contained two push buttons for response indicators during tasks. The electrocardiogram (EKG) was recorded using two electrodes placed on the lateral surfaces of the subject’s upper arms. The EKG and the occurrence of experimental events were detected and displayed
182
by an electronic system * custom built according to Whitehead et al. (1977) method. Interbeat intervals (IBIS) were evaluated to an accuracy of 1 ms. EKG R-waves were detected and timed by feeding the EKG to a Schmitt-trigger and detecting the Schmitt-trigger output. HR. skin conductance and EMG were recorded, during HR biofeedback training, using SATEM biofeedback equipment. A PT601 SATEM module cardiotachometer computed the beats per minute heart rate. HR was evaluated on a beat-by-beat basis by detecting the peripheral pulse. The peripheral pulse was detected from the volar surface of the distal phalanx of the second forefinger of the nondominant hand by a photoelectric plethysmographic transducer. Each amplified pulse signal was detected by a voltage-level detector and the interpulse intervals were evaluated with an accuracy of 1 ms and HR was then expressed in beat/mm with a 5-s interval. HR feedback was delivered through headphones and amplified from an audio amplifier. Audiotory feedback was given by a continuous series of beeps. A high HR level produced a high beep rate. As the HR level declined, the beep rate decreased. The beep rate of the feedback signal was contingent on, and immediately after, an interpulse interval variation (i.e., increase or decrease of HR). Skin conductance (SC). SC was recorded DC from two non-polarizing biopotential electrodes, l cm in diameter, with a KCL electrolyte. Electrodes were fastened over the volar surface of the second and third forefingers of the right hand. They were connected to a SATEM PT401 module that supplied through the electrodes a constant current density of 10 PA/cm to record the skin conductance response (SC). A time constant of 10 s was chosen. Frontal electromyographic (EMG) activity. Frontal EMG was monitored by centering non-polarizable biopotential electrodes 1.5 cm above each
* The system to record the HB discrimination data was designed by Dino Moretti and Pietro Fermani, technicians of the Department of Psychology. University of Rome, “La Sapienza”.
eyebrow, with a reference electrode in the center of the forehead. The diameter of the electrodes was 1.4 cm. A PT311S-SATEM EMG biofeedback module with standard amplification (EMG level from 0.1 to 10 microvolts and 70 to 1000 Hz bandwidth) was employed. Frontal EMG levels were integrated across 5 s. Materials The State-Trait Anxiety Inventory (STAI) Form X-l (State Anxiety) and X-2 (Trait Anxiety) (Spielberger, Gorsuch and Lushene, 1970) and the Tellegen Absorption Scale (TAS, Tellegen and Atkinson, 1974) were used to assess emotionality and absorptive ability of the subjects, respectively. Following HR feedback training, subjects were also requested to fill out Blanchard, Scott, Young, and Edmundson’s (1974) list regarding the cognitive strategies that they had used during their HR-control training. Procedure Subjects were screened and introduced to the experimental procedures during session 1. They were introduced to the experimental setting, showed the monitoring equipment, and were also informed about the phases of the experimental procedure. They were not, however, informed about specific experimental hypotheses. Subjects were then asked to complete the STAI From X-l to evaluate their state anxiety in the laboratory setting. Following completion of the inventory, subjects were required to sit in the reclining chair quietly during the session. Subjects were given taped instructions explaining the heartbeat discrimination task, according to the Whitehead et al. (1977) and Montgomery and Jones (1984) methods. They were instructed to press one of two buttons after each trial to indicate the occurrence of delayed (‘false’) or immediate (‘true’) HR feedback. The HB discrimination task consisted of 60 heartbeat discrimination trials each 10-s long. Subjects received two blocks of 30 trials each, with a 2-min rest period between. Half of the trials consisted of immediate feedback and the other half delayed feedback. and these were randomly distributed within each block. Immediate or true HB feedback was a 0.1-s flash occurring 128 ms
183
after each EKG R-wave *, delayed or false feedback was a light flash occurring 384 ms after each R-wave. The electronic system first presented to the subjects a blue light to indicate that the HB discrimination trial would soon begin. After the 10-s perception trials a yellow light came on to signal to the subjects that it was time to respond by pressing one of the two buttons to indicate whether they believed immediate or delayed feedback had occurred. The subjects were allowed as much time as they required to respond and the yellow light remained until they responded. The next trial began as soon as the subject responded. Subjects were given two push-buttons (one on the right and the other on the left arm of a recliner chair) labeled ‘true’ and ‘false’. During heartbeat trials they were required to respond by pressing the ‘true’ button if they felt that the HR light-signal they received was immediate or true heartbeat feedback; in the cases in which they felt that the HR signal was delayed or false, they were required to press the ‘false’ button. Subjects were not informed about their HB perception performance, nevertheless they were informed that they would obtain one point for each correct decision and that the subjects who exhibited the largest scores would be participants in the HR biofeedback training experience. At the end of the HB discrimination task, subjects were required to complete the trait anxiety inventory (STAI, Form X-2) and the Tellegen Absorption Scale (TAS). Heart-activity perception was measured using the following formula (Montgomery and Jones, 1984): HDI = (Nt/NT - Ntf/NF), where Nt/NT represents the proportion of immediate feedback trials to which the subject responds ‘true’ (Nt = number of ‘true’ responses given from the subjects; NT = number of actual ‘true’ trials) and Ntf/N F is the proportion of delayed feedback trials to which the subject responds ‘true’ (Ntf = number of delayed feedback trials that subject
* Whitehead et al. (1977) have assumed that the perception of heart contraction requires a delay of 128 ms (immediate or true feedback). This assumption is supported by the fact that mechanical contraction of the heart requires approx. 100 ms (150 ms for the blood pulse wave) to reach the wrist or neck following the R-wave.
identifies as ‘true’, N F = number of actual ‘false’ trials). As suggested by Montgomery and Jones (1984) this index is similar to Clemens’ (1979) d’ index of perceptual sensitivity in signal detection theory. HDI is taken here as an absolute value between 0 and 1; higher values indicate an increased heart-activity perception. This index is a linear probability measure which can be analyzed by means of parametric statistics. Sessions 2 and 3 were to test heart rate control, one for HR increase and the other for HR decrease. Following attachment of the photoelectric plethysmographic finger transducer and EMG and skin conductance electrodes, the experimenter gave the subjects tape-recorded experimental instructions according to the requirement of each experimental session. Common instructions to all subjects included a 5-min explanation about the focus of the research. The purpose of this period was to allow subjects to become accustomed to the physiological recording devices that had been attached to them. Subjects were also told to avoid holding their breath, breathing deeply or breathing in any other peculiar way, and they were requested to breathe at a steady regular rate. They were also requested not to tense or flex muscles and to move as little as possible in their chairs. After these instructions, resting level pulse rate was measured for a 3-min initial period. Following this period, the subjects’s instructions differed depending on the condition of the experimental design to which they had been assigned. Subjects received a combination of the instructions, described below, appropriate for their condition: HB perception factor. Subjects were divided into two groups, one of good HB perceivers (N = 20 Ss) and the other of poor HB perceivers (N = 20 Ss). Each of these subjects perceived one combination of the experimental treatments. Feedback-no feedback factor. Subjects either were provided with biofeedback and instructed to use it in attempting to control their heart rates or were not provided with biofeedback. In the feedback condition (FB) subjects were told that they would hear beeps through the headphones, and that the rate of the beeps would be determined by their HR. They were instructed to regulate HR using the beep as pulse rate information (i.e., a
184
beep rate increase represented a faster HR, and a beep rate decrease represented a slower HR). The acoustic intensity of the feedback signal was adjusted to the subject’s requirement. Subjects in the no feedback (NoFB) condition were not provided with any HR feedback and were specifically told to regulate HR by means of their own mental strategies. Motivating and non-motivating instructions. Subjects were given either motivating or non-motivating instructions before beginning each HR-control task. Non-motivated subjects were told that each person is able to acquire good HR control, and that they must not become worried if they did no produce very large HR changes because biofeedback research has demonstrated that even small changes are significant. The motivated subjects were given instructions emphasizing that, according to biofeedback research, persons who have learned to increase or to decrease their HR in the laboratory, are also able to control it in a stressful real-life situation. Moreover, they were informed that the acquisition of this capacity could reduce the probability of heart injures. To enhance the motivating effect, they were also told that there is evidence showing that a person obtains a good level of HR control when he/she shows a good level of task engagement. Increase-decrease factor. Subjects were instructed to try to either increase or decrease their HR. Subjects in each experimental condition received five 3-min training periods separated by 1-min rest periods. Training periods began with the illumination of a small red light in front of each subject. The red light was switched off at the
TABLE
I
Correlation matrix for state anxie+ (X-l), trait unxieiy (X-2), Tellegen absorption scale (TAS) and heartbeat discrimrnation index (HDI) Variables
X-l
x-2
0.501 * *
x.2
TAS
0.109
0.273
HDI
0.078
~ 0.100
* P < 0.05 * * P i 0.01 (two-tailed
test, df = 76).
TAS
end of each training period. The illumination of a small blue light indicated a rest period. Following the completion of the last period, subjects in all groups were given Blanchard et al.‘s list to assess the cognitive strategies they used during the training periods. Subjects were also asked to report thoughts or activities they had experienced during HR increase or HR decrease conditions.
RESULTS Anxiety, absorption and heartbeat perception To evaluate the influence of State Anxiety and Trait Anxiety (STAI, Forms X-l and X-2) and Absorptive ability (TAS) on HB perception scores, a multiple regression analysis across the overall group (N = 78) was carried out. No significant relationship was found. None of the zero-order correlation coefficients computed with respect to HDI scores was found to be significant (see Table I). The partial correlations revealed that anxiety and absorption variables accounted for a negligible portion of the total variance (less than 2%). Furthermore, to evaluate if good (N = 20) and poor (N = 20) perceivers differed in their anxiety or absorption levels, three separate one-way ANOVAs were performed. No significant differences were observed between groups for each of the variables considered (Trait Anxiety: Good = 41.0, S.D. = 9.25, Poor = 43.0, S.D. = 8.5; State Anxiety: Good = 40.7, S.D. = 8.4. Poor = 38.1, SD. = 7.3; TAS: Good = 7.2, S.D. =2.0, Poor= 8.2, S.D. = 2.7). Groups of subjects who scored high and low on each measure of anxiety and of absorption were then taken from the pool of all subjects (n = 78) and analyzed in three one-way ANOVA’s with respect to HB discrimination scores. No significant differences (P > 0.05) for HB discrimination scores between high and low subjects in trait anxiety, state anxiety and absorption were found (F(1.76) = 1.11. 2.98 and 0.17; P > 0.05 for Trait and State anxiety, and Absorption, respectively).
* PO.131
Physiological measures and correction for initial vulues Heart rate (HR), skin conductance (SC) and
T5
T4
T3
T2
Tl
Trials
Groups: HR.D
NoFB
1.8
(1.5)
5.2
(1.4)
(1.1)
- 1.9
1.5
(1.1)
0.3
(1.4)
3.9
(1.5)
(1.0)
(1.0)
-1.3
(1.2)
- 0.2
(1.3)
-0.7
(1.5)
-0.3
(1.2)
4.9
(1.3)
(1.5)
(1.2)
2.5
- 2.3
HR-I
0.4
HR-D
0.2
HR.I
(1.6)
- 0.8 (1.1)
5.8
(1.2)
6.8
(1.9)
- 4.8
3.4 (0.9)
(1.8)
1.8
3.5 (1.2)
- 5.8 (1.1)
- 3.7 (1.2)
- 3.2 (1.1)
- 3.6 (0.9)
0.6 (1.3)
7.1 (1.2)
HR.D
HR.I
(1.8)
- 2.3
(1.8)
3.4
1.2 (1.2)
1.1 (1.0)
5.2 (1.3)
1.5 (1.2)
5.3 (1.1)
HR.I
HR.D
(1.0)
- 3.6
(0.9)
0.4
(1.1)
- 2.7
(1.3)
- 3.4
(1.5)
- 0.9
(1.4)
3.7
(1.3)
2.9
(1.3)
3.0
(1.3)
4.4
(1.4)
3.0
HR.I
FB
4.4
3.4
3.4 (1.2) 2.7 (1.1)
- 1.1 (1.4) - 4.2 (1.6)
- 1.1 (1.7) - 0.2 (1.8)
1.5 (1.4)
(1.1)
- 3.2
(1.3)
1.6 (1.7)
1.2
(1.3)
(1.5)
(1.6)
-5.2
(1.4)
4.9
HR.I
- 4.3
(1.7)
(1.9)
- 2.4
- 1.2 (1.5)
HR.D
HR.I
FB
(1.5)
- 2.9
(1.4)
-5.0
(1.5)
- 3.3
(1.4)
-2.1
(1.5)
- 2.2
HR-D
Motivated subjects
(1.0)
- 1.5
(1.0)
-1.5
(1.1)
1.2
HR-D
NoFB
subjects
Non-motiuated
NoFB
FB
FB
Non-motivated
subjects
Poor HB perceiuers
Good HB perceivers Motivated subjects
(1.3)
2.0
(1.2)
-1.4
(1.1)
-0.3
(1.2)
0.7
(1.5)
9.1
HR.I
NoFB
(0.8)
-2.1
(1.1)
0.7
(1.0)
-0.7
(0.9)
~ 1.4
(1.2)
-1.7
HR-D
Heart rate change (bpm) duringJive trials (Tl, T2, T3, T4, T5) J rom Jive pre-trial baselines in HR-increase (HR.1) and HR.decrease (HR-D) tasks for Good and Poor HB perceiuers who were provided and not provided with HR.feedback (FB, NoFB conditions) and whom were given non-motiuating and motivating instructions during HR-control tasks. Standard errors in parentheses.
TABLE II
1X6
frontalis EMG amplitude were quantified in 5-s time periods, and expressed in beats/min, microsiemens and microvolt, respectively. Each physiological measure was analyzed using Analysis of Covariance (ANCOVA) or Analysis of Variance (ANOVA) with a Split-Plot Factorial Design (see Kirk. 1968); ANCOVA was used in the cases in which the ‘Law of Initial Values’ (LIV, Lacey 1956; Lacey & Lacey, 1962) was present. LIV was determined to be present if, and only if, the correlation coefficient r evaluated between prestimulus level (s), and a poststimulus minus prestimulus difference level (d) was significant (Benjamin, 1967). ANOVA was used in the cases in which LIV was not considered to be present, that is. I was not significant. Two values of r were obtained for each variable: one in HR increase and the other in HR decrease condition. These values for HR data were not significant ( -0.08 and ~ 0.11. df= 38. P > 0.05, in HR increase and HR decrease condition. respectively). The correlation coefficient was significant in HR-increase for SC activity. and in HR-decrease condition for EMG activity (SC: 0.33, P < 0.05 and 0.08 N.S.: EMG: -0.03. N.S. and -0.38 P -C 0.05. for HR-increase and HR-decrease respectively). Effects of heurtheut perception, instructions und feedback WI HR control An analysis of variance on the mean HR scores for the minute preceding each trial showed that these means did not differ significantly between groups or interact with conditions. The only significant effect was a highly significant increase in HR over the 2nd, 3rd and 4th pre-trial baseline periods as compared with the 1st one (F(4.128) = 5.3, P < 0.001). In view of this. an attempt was made to allow for habituation and chance fluctuations in HR by subtracting these mean baseline HRs, preceding each trial, from the trial means to give the basic data for the main analysis (see McCanne and Sandman, 1975, for the application of this procedure). Table II reports HR difference scores across experimental conditions. The ANOVA for these HR data showed some significant interaction effects. The most powerful effects were for HR-control ( F(1.32) = 56.5, P < O.OOl), and for the Instruc-
tions x HR-control x Trial interaction ( F(4.128) = 5.6, P < 0.001). The first effect indicated (Duncan-test. P < 0.01) that, with respect to the running baseline periods, there were significant HR changes according to task requirements (i.e., the HR increased during HR-increase and decreased during HR-decrease conditions). The second interaction indicated that HR-control over trials was influenced by Instruction types. More specifically, motivated subjects as compared with non-motivated subjects, were more able to obtain HR changes in both HR-increase and HR-decrease trials. In particular, motivated subjects were able to obtain a significant increment or decrement of their HR across each of the five experimental trials with respect to baseline periods (motivated subjects: 8.6. 2.4. 2.4. 2.5. 2.9. for HR-increase trials, and ~ 1.1. - 1.7. - 2.8, - 1.8. - 3.6 for HR-decrease trials). This was not found to be true for non-motivated subjects who showed no HR change over trials 1, 3 and 4 during HR-increase, or over trials 1, 3, and 5 during HR-decrease task (non-motivated subjects; 0.7, 4.0. 0.6, 0.7, 2.1 for HR-increase trials and 0.6. -2.6, 0.6, - 2.7. -0.3, for HR-decrease trials). Finally. the Feedback x HR-control interaction was significant (F(1,32) = 7.8. P < 0.009). Comparisons among the means indicated that there were no differences in HR-reduction between Feedback and No-Feedback during the HR-decrease condition ( - 1.7 and ~ 1.4 for HR decrease/FB and HR-decrease/ No-Feedback conditions, respectively). In contrast, during HR-increase. the Feedback effect was more pronounced than NoFeedback (3.8 and 1.1 for HR increase/FB and HR increase/No Feedback conditions. respectively). No main or interactional effects for HBperception or Feedback factors were found. SC dutu. An analysis of variance on the mean SC activity for the baseline periods preceding each trial, indicated that motivated subjects had higher levels of SC activity than the non-motivated ones (16.8 vs. 11.4, respectively: F(1.32) = 7.2, P < 0.02). The other effect was for baseline periods. wherein a significant decrease of SC activity over the 3rd, 4th and 5th periods, as compared with the 1st and 2nd ones, was observed (15.1. 15.1. 13.5. 13.5. 13.2, respectively; F(4,128) = 17.0, P <
III
T5
T4
T3
T2
Tl
Trials
Groups.. HR-D
(0.6) - 1.1 (0.6) - 0.5 (0.6)
0.5 (1.2) - 0.5 (1.4) 1.4 (1.4)
- 0.8
(1.1)
~ 3.8
(0.9)
- 2.4
(1.0)
0.5
(1.2)
0.3
(0.9)
0.5
(0.8)
-1.0
(0.1)
(1.2)
-1.2
- 0.7
1.0
(1.1)
0.1 (0.8)
1.2 (1.1)
(1.1)
(0.1)
- 3.4
~ 0.4
2.2
HR.I
(1.1)
NoFB
HR.I
HR.D
FB
(1.4)
- 0.2
(1.4)
0.2
(1.4)
PO.2
(1.3)
0.4
(1.2)
1.7
HR-I
FB
(1.1)
- 0.4
(1.1)
0.0
(1.2)
1.5
(1.1)
0.0
(1.1)
~ 0.3
HR-D
NoFB
(1.2)
- 0.2
(1.3)
- 0.3
(1.2)
0.0
(1.2)
~ 0.4
(1.2)
0.8
HR-I
HR-D
2.2
(1.8)
- 0.7
(0.8)
1.9
(0.8)
0.3 (1.8)
(0.7)
- 2.8
(0.7)
~ 4.5
(1.1)
-1.5
(1.8)
-0.5
(1.8)
~ 0.6
(1.8)
- 0.2
HR.1
FB
- 0.3 (1.5)
0.0
(1.4)
0.0
(1.4)
-0.7
(1.4)
-0.8
(1.4)
-0.2
HR.I
NoFB
(1.1)
(1.1)
0.4
(1.1)
-1.2
(1.1)
1.4
(1.1)
- 2.6
HR.D
subjects
Poor HB percewers Non-motivated
Good HB perceiuers
Non-motruated subjects
Motivated SubJects
(1.1)
-2.0
(1.1)
- 0.2
0.1 (1.1)
(1.1)
(1.1)
(1.1) - 0.8
0.4
0.5
0.6 (1.1)
(1.1)
(1.1)
-0.5
(1.2)
-1.6
(1.1)
0.1
(1.2)
- 4.0
3.5 (1.3)
1.2
HR-D
subjects
(1.1)
HR-I
FB
~ 0.5
-1.7 (0.9)
HR-D
Motwated
(1.3)
- 0.5
(1.2)
-0.3
(1.3)
0.2
(1.3)
- 0.9
(1.3)
0.6
HR.I
NoFB
(1.4)
- 0.3
(1.4)
~ 0.7
(1.4)
0.7
(1.3)
- 1.2
(1.3)
-0.5
HR.D
Skin conductance change (pS) dunng five rrrals (Tl, T2, T3, T4, TS) from flue pre-trral baselines m HR.increase (HR.I) and HR.decrease (HR-D) tasks for Good and Poor HB perceivers who were provided and not provided with HR-feedback (FB, NoFB conditions) and whom were given non-motwatlng and motwating instructions during HR-control tasks. Standard errors in parentheses.
TABLE
IV
T5
T4
T3
T2
Tl
Trrals
Groups..
0.4 (0.6) - 0.2 (0.5)
0.4
(0.9) 0.3
(0.9)
1.5
2.2
(1.3)
(0.6)
0.7
0.3 (0.6)
(1.3)
(0.8)
(1.3)
-0.2
(1.0)
2.8
0.1
(0.8)
-0.07
(0.8)
(1.0)
0.0
(0.8) -0.5
0.5 (0.9)
- 0.2
1.1 (1.0)
(0.8)
1.4 (0.9)
~ 0.6
- 0.4 (0.8)
0.1
(1.0)
0.0
(0.9)
0.1
(0.5)
0.4
(0.7)
-1.3
(0.6)
-0.3
(0.6)
-0.4
(0.5)
-0.1
(0.6)
(1.2)
-1.3
(1.3)
0.6
(1.2)
1.3
(1.1)
0.7
1.2
0.7 (0.8)
(1.4)
(0.7) ~ 3.7
(1.3)
- 1.3
0.4 (0.7)
(1.2)
(0.5) - 1.6
(1.1)
-1.4
(1.5)
- 0.4
0.3 (0.5)
- 3.9
0.0 (1.1)
-0.1 (0.9)
0.8
(1.5)
HR.D
HR.I
HR-D
HR-I
HR.D
HR.I
(0.9)
~ 0.4
(0.9)
- 0.5
(0.9)
0.0
(0.8)
- 0.2
(0.9)
- 0.6
(0.9)
1.4
(0.9)
0.4
(0.9)
-0.4
(0.9)
-0.1
(0.8)
0.1
HR.1
NoFB
HR.I
HR.I
HR.D
FB
NoFB
FB
NoFB
FB HR.D
Non-motlr~ated subyxts
Non-motrocrted subjects
Poor HR perceuxrs Motwated subjecls
Good HB percewers
(0.8)
-1.2
(0.8)
-0.2
(0.6)
-1.0
(0.6)
PO.4
(0.8)
-1.2
HR.D
(1.0)
- 0.3
(1.1)
0.9
(1.0)
0.8
(1.1)
0.6
(1.1)
0.1
HR-I
FB
(1.0)
- 0.4
(1.1)
- 0.4
(1.0)
- 0.4
(1.1)
-1.2
(1.1)
-0.1
HR.D
Motioated subjects
(0.8)
0.1
(0.8)
0.1
(0.8)
0.5
(0.8)
0.5
(0.9)
- 0.3
HR.I
NoFB
(1.0)
~ 1.1
(1.1)
- 0.8
(1.0)
- 0.7
(1.1)
- 0.7
(1.1)
-0.5
HR.D
hasehnes ,,I HR.mcrease (HR.I) and HR.d ecreaw(HR.D) tasks for Good and Poor HB EMG amplttude change (pV) durrngfwetrials (TI, TZ, T.$ T4, TS) f ram fivepi-e-tnal perceroers rtmhowere prouded and not prorvded w:h HR.feedback (FB, NoFB condrtrons) und whom UWP grwn non-motwutrng cmd motwatmg mstructrons durrng HR-control tasks. Standard errors m parentheses.
TABLE
189
0.001). No other effects were found. SC baseline mean scores preceding each trial were substracted from trial means to give the basic data (difference scores) for statistical analysis. Difference scores for SC activity are reported in Table III. Owing to the fact that during the HR-increase task SC baseline levels were found to be related to SC change, an ANCOVA, using resting scores as covariate, was performed for SC difference scores across HR-increase and HR-decrease trials. A main effect for Instruction was found (F(1,31) = 4.2, P < 0.05) indicating a significant lowering of SC level during task performance for non-motivated subjects, while the motivated subjects did not show a significant change (-0.6 and 0.03, respectively). The Instruction X HR control X Trial interaction was significant (F(4.128) = 7.7, P < 0.001). This interaction indicated that with respect to baseline levels non-motivated subjects reduced their SC level over trials 1, 4 and 5 (- 1.0, -0.4, -0.6, -1.6 and -1.2, from the 1st to the 5th trial, respectively) in the HR-decrease condition. Meanwhile, no corresponding differences were observed between trials for motivated subjects with the exception of the 2nd trial in which a reduction of SC activity was observed (0.6, - 1.5, 0.5, -0.5, -0.4, from the 1st to the 5th trial, respectively). Moreover, HB perception x Instruction x HR control X Trial was significant (F(4,128) = 3.6, P < 0.001). Inspection of the means showed different trends between good and poor perceivers across HR-decrease trials for non-motivating and motivating instructions. In particular, good perceivers who were not motivated in HR reduction, significantly (P < 0.01) lowered SC activity across the last 4 trials with respect to the corresponding baseline levels (-0.37, -1.0, -0.9, -3.0, -1.4 from the 1st to the 5th trial, respectively), while they did not show SC change when they were motivated in controlling HR (-0.25, -0.3, 0.5, 0.1, -0.5 from the 1st to the 5th trial, respectively). Poor HB perceivers who were not motivated to reduce HR, in contrast, significantly reduced SC activity during the first and .fifth trials (- 2.2, 0.4, 0.3, - 0.2, - 1.0, respectively) while poor perceivers, who were motivated to reduce HR, increased SC activity during the first trial and reduced it during the
second - 0.3, tween across actional
and fourth trials (1.5, -2.6, 0.4. -1.1, respectively). No differential trends begood and poor HB perceivers were found HR-increase trials. No other main or intereffects were found. EMG data. The analysis of variance across baseline EMG scores did not evidence any main or interactional effect. As with HR and SC scores, the EMG difference scores were used as basic data for statistical analysis. Because EMG amplitude in the baseline condition was found to be related to EMG change during the HR-decrease task, an ANCOVA, using initial baseline EMG scores as covariates, was performed. There was a Trial x HR control interaction (F(4,128) = 7.24, P < 0.001). This interaction indicated a decrease of EMG amplitude during the HR-decrease condition in respect to baseline levels ( - 0.7, - 0.6, - 0.6, - 0.4, -1.1 from the 1st to the 5th trial, respectively), and in contrast a tendency of EMG amplitude to increase across the trials during the HR-increase task was found. Finally, Instruction x Trial (F(4,128) = 4.9, P < 0.05) and HB perception x Instruction X Trial interactions (F(4,128) = 4.4, P < 0.05) were significant. The last interaction indicated that good HB perceivers who were motivated in HR-control significantly lowered their EMG amplitude with respect to the baseline condition across four of the five trials (1.74, - 0.76, - 1.0, - 0.8, - 2.5 respectively), while no significant reductions across trials for non-motivated subjects were found (-0.08, -0.1, -0.4, 0.1, - 0.12, respectively). No significant differences across trials were found for the remaining experimental conditions. EMG difference scores during the experimental conditions are reported in Table IV. Cognitive mediation
Subjects were encouraged to try to either increase or decrease their HR by using their own cognitive strategies, but no mention was rn’ .1C about the type of strategy that they could adopt in the task. At the conclusion of each HR-control session, subjects filled out the 15-items list of cognitive strategies of Blanchard et al. (1974). Subjects were requested to indicate the cognitive
strategies used in controlling their HR, and when the strategy adopted by the subject was not listed on the questionnaire, they were requested to report it on the sheet. The items contained three categories of responses: (1) activation responses (cognitive activities concerning thoughts about motor activity, anxiety, fear. anger, or aggressive ideas; 6 items); (2) relaxation or trying to relax situations (5 items); (3) self-instruction about the physiological response to be controlled (focusing attention on body antecedents related to HR change, focusing attention on feedback signal, etc.; 4 items). To evaluate any differences between the number of activation, relaxation, and self-instruction responses reported by motivated versus nonmotivated subjects, two separate x2 tests were carried out for Feedback and No-feedback conditions, respectively. No significant differences were found (P < 0.05). When activation, relaxation, and self-instruction responses were compared in HRincrease versus HR-decrease tasks, significant differences for FB (x’ = 32.79. P < 0.001)and for NoFB conditions (x’ = 32.54, P -c0.001)were found. These results indicated that during the HR-increase task subjects reported significantly more activation than relaxation responses. while during the HR-decrease task they reported more relaxation than activation responses. Percentages of reported strategies, during HR-increase, were 6.8%. 52.7%, and 41.2% for relaxation, activation and self-instructions. respectively. During the HRdecrease task. corresponding percentages were 41.5%. 19.9% and 38.6%.
CONCLUSIONS The results of this study do not support a significant relationship between Whitehead’s HB perception measure and HR control. This negative finding is inconsistent with a prediction from Brener’s (1974, 1977) theory: that the presence of an appropriate response image should be a sufficient condition for control of that response. Our findings suggest that HB awareness prior to any training is not predictive of the performance on voluntary HR control. Reports in which measures of cardiac awareness were derived from HR track-
ing procedures have observed a significant relationship between initial discrimination performance and performance on subsequent attempts to increase HR. but not with performance on attempts to decrease HR (Kleinman, 1970; McFarland & Campbell, 1975; McFarland, 1975). Similar findings were obtained by Clemens & McDonald (1976) using the Brener-Jones procedure for assessing cardiac awareness. These findings contrasted with those obtained in one of our studies (De Pascalis et al.. 1984) wherein HB perception was evaluated with McFarland’s (1975) method. The HB perception/HR-increase relationship was nonsignificant, while it was significant for HR-decrease. Subjects who achieved high HB-perception scores decreased HR significantly better than low scorers. Moreover, conflicting results in three other experiments using the Whitehead procedure were also obtained (Whitehead. Drescher, Heiman & Blackwell, 1977): a negative correlation between cardiac discrimination and control in one case and no relationship in two cases. This suggests that the processes tapped by the Whitehead procedure and the Brener-Jones or HR tracking procedures are dissimilar. One possible explanation of the negative finding, as obtained in this study may be given if we assume that HR control is mainly dependent on efferent processes. Lacroix (1981) in his review of the literature concerning Brener’ theory, observed that there was little convincing experimental evidence to confirm Brener’s view. Lacroix presented a two-process theory of the acquisition of autonomic control in which the development of control of a response consists primarily in the identification of the efferent behavioral programmes already in the subject’s repertoire. Afferent processes play a role in the cases in which behavioral programmes, effective in the control of the target response, are not available to the subject. Discrepant results among studies may be dependent on the different methods used with respect to the measured responses, biofeedback training procedures, and discrimination testing procedures. In respect to this last, Ross and Brener (1981) evaluated the extent to which the Brener and Jones and Whitehead procedures probe the same general capacity in cardiac awareness. The
191
authors observed that subjects used different strategies in the solution of the two procedures: (1) the Brener and Jones procedure tended to induce an ‘active’ solution strategy which had a significant respiratory component characterized by a deep breath at the beginning of each discrimination trial and then holding the breath and remaining quiescent for the remainder of the trial; (2) the Whitehead procedure, in contrast, required the subject to adopt a strategy discouraging the production of HR variations and reduced sensory input. These authors observed that subjects who adopted an active discrimination strategy displayed significantly greater voluntary HR increase, in a post-discrimination test of HR control, than did subjects who adopted a passive discrimination strategy. Some other experiments have demonstrated that among light flashes (Yates et al., 1985) or tones (Brener and Kluvitse, 1988) occurring at intervals of 0, 100, 200, 300 and 400 ms after each R-wave, the stimuli that were 200 and 300 ms after the R-wave were judged to be most contiguous with heartbeat sensations. These findings are in contrast with the Whitehead et al. (1977) definition of heartbeat (i.e., a stimulus presented 128 ms after the R-wave was defined as a heartbeat, while a stimulus delivered 384 ms after R-wave was defined as occurring outside the events of the cardiac cycle). On the other hand, the experiment of Jones, Collins, Dabkowski and Jones (1988) has demonstrated that the ability to make exteroceptive temporal discriminations of the 128 and 384 ms time delays has little to do with the ability to make similar interoceptive discriminations with Whitehead’s paradigm. In view of the fact that Whitehead’s cardiac awareness procedure cannot be reduced to a simple discrimination of an exteroceptive stimulus pair, the different results among studies may be explained in terms of the different perceptual components measured with the Whitehead and Brener and Jones procedures. The lack of a relationship between HB perception and HB control found in this study is not attributable to the use of female subjects. Although a number of early reports found that males may be superior to females in making HB discriminations with Whitehead’s procedure (White-
head et al., 1977; Wildman & Jones, 1982) and during cardiac discrimination learning with knowledge-of-results training (Katkin, Blascovich & Goldband, 1981; Katkin, 1985) more recent studies (Jones, Jones, Rouse, Scott, & Caldwell, 1987; Rouse, Jones and Jones, 1988) have shown that gender differences in HB perception may be dependent on difference in body composition. Although we did not record respiration, we may exclude in this experiment a mediating role for respiration activity in the evaluation of cardiac awareness, because there is no evidence for such a relationship with Whitehead’s procedure (Jones et al., 1987; Rouse et al., 1988). SC and EMG activities were not found to be dependent on individual differences in HB perception, except for good perceivers who were motivated to reduce their HR; they evidenced no change in SC levels and a pronounced reduction in EMG activity across trials with respect to baseline levels. Although it was found that subjects increase or decrease their HR with or without feedback signal, analysis of variance also showed that the feedback condition was more effective than instructions alone in the degree of HR increase. In accordance with earlier reports, no differences between feedback and no-feedback conditions for HR-reduction trials were found (Blanchard & Young, 1972; Bouchard & Granger, 1977; Manuck, Levenson, Hinrichsen & Gryll, 1975; Levenson, 1976; Bouchard & Labelle, 1982). Results of this study confirm other findings (Cuthbert et al., 1981) suggesting that instructions play a principal role in the voluntary control of heart rate. Subjects who were given motivating instructions obtained greater HR changes, either among HR-increase or HR-decrease trials, as compared with subjects who were given nonmotivating instructions. SC and EMG activities during HR-reduction trials were parallel to the low arousal level in accordance with task requirements. Non-motivated subjects, whom we expected to be less activated, showed lower SC levels and no difference in EMG activity, as compared with baseline levels, in the HR-decrease condition. In this condition, motivated subjects exhibited a reduction in EMG activity and no change in SC levels. SC and EMG levels did not differ apprecia-
192
bly between non-motivated and motivated while subjects attempts to decrease HR.
groups
REFERENCES Ashton. R.. White, K.D.. and Hodgson. G. (1979) Sensitivity to heart rate: A psychophysical study. P.yvchoph?;siologv. 16: 463-466. BenJamin, L.S. (1967) Facts and artifacts in using analysis of covariance to ” undo” the law of initial values. Psychophysrolog\, 4: 187-202. Blanchard, E.B.. and Young, L. (1972) Relative efficacy of visual and auditory feedback for self-control of heart rate. .I. Gen. PsychoI.. 87: 195-202. Blanchard. E.B., Scott, R.W., Young, L.D.. and Edmundson, E.D. (1974) Effect of knowledge of response on the selfcontrol of heart rate. Psv~hoph_v.stolo~v. 11: 251-264. Bouchard. M.A., and Labelle, J. (1982) Voluntary heart rate deceleration: A critical evaluation. Biofeedback und Self Regulation. 7: 121-137. Bouchard. M.A.. and Granger. L. (1977) The role of instructions versus instructions plus feedback in voluntary heart rate slowing. PsJ,chophysrolom, 14: 475-482. Brener, J. (1974) A general model of voluntary control applied to the phenomena of learned cardiovascular change. In P.A. Obrist. A.H. Black. J. Brener, and L.V. DiCara (Eds.), Cardiocmc. Psychophysiol.. Chicago: Aldine. Pp. 365-391. Brener. J. (1977) Sensory and perceptual determinants of voluntary vlsceral control. In G.E. Schwartz and J. Beatty (Eds.). Btofeedbock: Theon und research. New York: Academic Press. pp. 29-66. Brener. J.. and Jones. J.M. (1974) Interoceptive discrimination in intact humans: Detection of cardiac activity. Phvsiol. Behar:.. 13: 763-767. Brener, J.. and Kluvitse, C. (1988) Heartbeat detection: Judgement of the simultaneity of external stimuli and heartbeats. P.yychoph.vsioloR). 25: 554-561. Clemens. W.J. (1979) Assessment, learning and retention of heart beat discrimination. Psychophysiology, 16: 333-338. Clemens. W.J., and MacDonald, D.F. (1976) Relationship betw’een heart beat discrimination and heart rate control (Abstract). Ps~~choph,~stologv. 13: 176. Cuthhert, B., Krlsteller. J., Simons, R., Hodes, R.. and Lang, P.J. (1981) Strategies of arousal control: Biofeedback, meditation, and motivation. J. L-p. Psycho/.: Cert., 110: 518-546. De Pascalis. V.. Alberti. M.L. and Pandolfo, R. (1984) Anxiety, perceptlox and control of heart rate. Percept. Mot. Skilh. 59: 203~211. Jones, G.E.. Jones, K.R., Rouse, C.H.. Scott, D.M., and Caldwell. J. (1987) The effect of body position on the perception of cardiac sensations: An experiment and theoretical implications. Psychoph_vstologv, 24: 300-311. Jones, G.E.. Collins, S.W.. Dahrowski. E.A.. and Jones. K.
(1988) The discriminahility of temporal patterns used in the Whitehead discrimination procedure. Ps.~,choph_vsioloRv, 25: 547-553. Katkin, E.S. (1985) Blood, sweat, and tears: Individual differences in autonomic self-perception. Psychoph_vsio/ogv, 22: 125-137. Katkin. ES. Blascovich, J.. and Goldband, S. (1981) Empirical assessment of visceral self-perception: Individual and sex differences in the acquisition of heartbeat discrimination. J. Personality Sot. PsychoI., 19: 568. Kirk. R.E. (1968) in Experimentul desrgn: Procedures for the hehoc~iorol sctences, Brooks/Cole Publishing Company, Belmont, pp. 245-318. Kleinman, R.A. (1970) The development of voluntary cardiovascular control. Unpublished doctoral dissertatmn. University of Tennessee. Lacey, J.1. (1956) The evaluation of autonomic responses: Toward a general solution. Ann. N. Y. Acad. Ser.. 67: 123-163. Lacey. J.I., and Lacey, B.C. (1962) The law of initial value in the longitudinal study of autonomic responses and response patterns over a four year interval. Ann. N. Y. Acud. Ser., 98: 1257-1290. Lacroix. J.M., and Gowen. A.H. (1981) The acquisition of autonomic control through biofeedback: Some tests of discrimination theory. P.r_vchophysio/ogy. 18: 559-572. Lacroix, J.M. (1981) The acquisition of autonomic control through biofeedback: The case against an afferent process and a two-process alternative. P.yychophysiologv. 18: 573587. Levenson. R.W. (1976) Feedback effects and respiratory involvement in voluntary control of heart rate. PsychophysrO/OR)‘,13: 108-114. Manuck. S.B., Levenson. R.W., Hinrichsen. J.J. and Gryll, S.L. (1975) Role of feedback in voluntary control of heart rate. Percept. Mot. Skilh, 40: 747-752. McCanne, T.R. and Sandman, C.A. (1975) Determinants of human operant heart rate conditioning: A systematic investigation of several methodological issues. J. Camp. Ph,sro/. Psycho/.. 88: 609-b18. McFarland, R.A. (1975) Heart rate perception and heart rate control. Psychophysrolom. 12: 402-405. McFarland. R.A., and Campbell. C. (1975) Precise heart-rate control and heart-rate perception. Percept. Mot., Skills, 41: 730. Montgomery, W.A., and Jones. G.E. (1984) Laterality, emotionality. and heartbeat perception. P.~_vchoph_vsto/ogv. 21: 459-465. Offutt. C., and Lacroix, J.M. (1983) Absorption: A predictor of biofeedback Iearnmg. (Abstract). P~~choph~~,stoloRv, 20: 4. Quails, P.J., and Sheehan. P.W. (1979) Capacity for absorption and relaxation during electromyograph biofeedback and no-feedback conditions. J. Ahnorm. Psycho/.. 88: 652-662. Quails, P.J., and Sheehan, P.W. (1981) Role of the feedback slgnal in electromyograph biofeedback: The relevance of attention. J. E.yp. P.y.vchol.: Gene&, 110: 204-216.
193 Ross, A. and Brener, J. (1981) Two procedures for training cardiac discrimination: A comparison of solution strategies and their relationship to heart rate control. Psychophysiolo~, 18: 62-70. Rouse, C.H., Jones, G.E. and Jones, K.R. (1988) The effect of body composition and gender on cardiac awareness. Psychophymlogy, 25: 400-407. Schandry, R., and Specht, G. (1981) The influence of psychological and physical stress on the perception of heartbeats (Abstract). Psychophysiology, 18: 154. Spielberger. C.D., Gorsuch, R.. and Lushene, R. (1970) The State-Trait Anxiety Inoentoty (STAI) Test Manual, Consulting Psychologists Press, Palo Alto.
Tellegen. A., and Atkinson, G. (1974) Openness to absorbing and selfaltering experiences (‘absorption’). a trait related to hypnotic susceptibility. J. Abnorm. Psychol., 83: 268-277. Whitehead, W.E., Drescher, V.M., Heiman. P., and Blackwell. B. (1977) Relation of heart rate control to heart beat perception. Biofeedback and Self-Regulation, 2: 371-392. Wildman, H.A., and Jones. G.E. (1982) Reliability of the Whitehead heartbeat discrimination procedure (Abstract). Psychophysiolog);, 19: 592. Yates. A.J., Jones, K.E., Marie, G.V. and Hogben, J.H. (1985) Detection of the heartbeat and events in the cardiac cycle. Psychophvsiologv, 22: 561-567.