RESPIRATORY AND HEART RATE INDTCES 0F REACTION TIME Thesis for the Degree of Ph. D. MICHIGAN STATE UNWERSTTY STEPHEN W. FORGES 192’0 «yr-02".“qu ' "‘ V I- .31-! 3;,- 14 R ‘ Michigan Static University .n‘ J /I//I//// /I////II // / 1/ //IT I////I///fl fl/Ifl/f/IITI/I/l/l/I 3 1293 10225 5554 This is to certify that the thesis entitled Respiratory and Heart Rate Indices of Reaction Time presented by Stephen W. Forges has been accepted towards fulfillment of the requirements for Ph. D. degree in Psychology Major pr e sor August 14, 1970 Date 0-169 Hi 1 mm“ Eu Mg; ABSTRACT RESPIRATORY AND HEART RATE INDICES OF REACTION TIME By ‘59 Stephen W. Forges .e 0‘ The respiratory and heart rate (HR) indices of attention to reaction time (RT) signal and control nonsignal stimuli were in— vestigated in male college students. The two conditions produced differential respiratory and HR responses. Heart rate accelerated in.response to the warning and respond signals. The HR acceleration to the respond signal was concordant with increases in HR variability, respiratory frequency and amplitude. In constrast to the mean HR, which did not change in anticipation of the respond signal, respiratory frequency increased and respiratory amplitude and HR variability de- creased. In response to the nonsignal stimuli HR decelerated. Mean HR variability prior to the trial onset and mean magnitude of the re— duction of HR variability in anticipation of the respond signal were highly significant correlates of RT when the preparatory interval (PI) was of variable duration. When the PI was of fixed duration HR variability was not correlated with performance. Although HR decelerated prior to the respond signal, the magnitude of deceleration was not related to performance. The data support a two component hypothesis of attention: the first, a phasic reflexive response dependent upon the specific stimulus change and characterized by a directional HR response; the second, a_tonic instrumental response related to attentional performance and characterized by a decrease in HR variability. RESPIRATORY AND HEART RATE INDICES OF REACTION TIME By. (i‘ 0* ‘ .,\\\ Stephen WJ‘Porges A THESIS Submitted to Michigan State University in partial fulfillnent of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1970 To a girl named Sue ACKNOWLEDGMENTS Sincere appreciation is expressed to Dr. Hiram E. Fitzgerald, Chairman of the dissertation ccmndttee, whose concern, guidance, and encouragement have contributed to the success of this research. I would like to thank the members of the conndttee, Dr. Paul Bakan, Dr. Glenn I. Hatton, Dr. Lawrence Messé, and Dr. Mark Rilling, for their personal concern and support during my graduate studies. Special thanks are expressed to Dr. Lawrence Messe for statisti- cal consultation and to Dr. Gerald M. Gillmore for computer assistance. TABLE OF CONTENTS List of tables.............................. ...... v- List of figures.... ............... . ............... vi Introduction.... .................................. 1 Method.................... ........ . ...... . ........ 5 Results.................................... ....... 8 Discussion.......................................23 List of references........... .................... 28 Appendices A. Instructions............ ............... 30 B. Analyses of variance sunnery tables....32 iv LIST OF TABLES Table Page 1. Analysis of variance summary table for 32 respiration frequency 2. Analysis of variance summary table for 33 respiration amplitude 3. Analysis of variance summary table for 3A mean heart rate 4. Analysis of variance sunnery table for 35 mean heart rate variance 5. Analysis of variance sunnery table for 36 second-by-second heart rate 6. Analysis of variance summary table for 37 reaction tine LIST OF FIGURES Figure 1. Mean reSpiration frequency as a function of periods for the two tasks 2. Mean respiration amplitude as a function of periods for the two tasks 3. Mean heart rate as a function of periods for the two tasks A. Mean heart rate variance as a function of periods for the two tasks 5. Mean heart rate change during the prepara- tory interval as a fUnction of seconds for each group 6. Mean heart rate change during the prepara- tory interval during Trial 1, 5, and 10 for the fixed PI RT group 7. Mean reaction time as a function of trials for the two RT groups vi Page 10 12 1A 16 17 l9 INTRODUCTION Considerable research has been involved in the identification of heart rate (HR) components of attention (Lacey, 1959, 1967; Graham & Clifton, 1966; Porges & Raskin, 1969). These investigations of the HR indices of attention have lead to attenpts to relate HR responses, most commonly HR deceleration, with sensorimotor performance (Obrist , Webb , & Sutterer, 1969; Webb & Obrist, 1970; Lewis & Wilson, 1970). Although HR deceleration is a reliable response during specific aspects of at— tention (Lacey, 1967; Porges & Raskin, 1969), the studies relating magni— tude of HR deceleration with.simple reaction time (RT) performance have indicated weak relationships. This repeated finding, must, therefore, question the use of HR deceleration as an index of attentional perform- ance. Lacey (1959) proposed that attentive observation of the external environment is accompanied by cardiac deceleration, whereas, situations that require rejection of the external environment (cognitive tasks or tasks requiring attentive observation of internal stimuli) produce HR acceleration. Lacey interpreted this directional HR response as a neuro- physiological link in.an afferent feedback mechanism responsive to the efferent processes of the CNS, and, therefore, capable of influencing behavioral events. Porges and Raskin (1969) substantiated the directional HR response by demonstrating that the direction of the HR response was dependent upon the experimental situation; HR accelerated while §s at- tended internally to their own HR, and HR decelerated while §s attended to intermittent tones or lights. The recording of HR responses during classical conditioning has demonstrated HR deceleration in anticipation of the unconditioned Stlflh ulus (Obrist, Sutterer, & Howard, 1969). Zeaman and Smith (1965) con— firmed the finding that anticipatory HR deceleration occurred independent of direction and magnitude of the cardiac component of the unconditioned response. They reported HR deceleration not only in anticipation of shock, but also in anticipation of both pleasant and unpleasant stimuli. Obrist (1969) reported anticipatory HR deceleration to the unconditioned stimulus and during subject preparation to respond in a simple RT task. Chase, Grahanu and Graham (1968) used a simple RT task and reported two deceleratory HR responses within the PI. The first deceleration followed the onset of the warning signal and the second preceded the onset of the respond signal. They interpreted the first deceleration as an OR to the warning signal and the second deceleration as a conditioned "attention" response enhancing stimulus reception. Other investigators (Lacey, Kagan, Lacey, & Moss, 1963; Obrist, 1963) have also reported deceleratory HR responses in anticipation of various motivationally significant stimuli. Contiguous with the research designating HR deceleration as an auto— nomic component of attention is the research attempting to forge a rela— tionShip between HR deceleration and performance on attention demanding tasks. Obrist, webb, and Sutterer (1969) using a 5 sec. fixed PI in a.sinp1e RT task, found a weak, but significant, correlation between the magnitude of HR deceleration and RT (r = .32, p_<.05). Lewis and Wison (1970) reported that HR deceleration accompanied visual scanning in lilk-month old children. The deceleration was related to the accuracy of the behavioral response, but not to the response latency, a measure of RT. The RT studies reviewed above have utilized fixed PIs. By using a fixed PI, the RT task may be explained in classical conditioning terminol— ogy; the warning and the fixed PI may be interpreted as a compound CS and the respond signal as the U08. Since the conditioned HR response is decelerative, HR deceleration prior to the respond signal would indicate a conditioned anticipatory response, temporally bound to the fixed PI. Performance during simple RT tasks is a function of the PI. Faster RT occurs during conditions in which the temporal relationship between the warning and respond signals is fixed. Thus, when a simple RT task with fixed P13 is used to investigate the relationship between autonomic components of attention and attentional performance, the results may be confbunded by the temporally conditioned autonomic responses. The possi- bility, therefore, exists that the HR deceleration prior to the respond signal is not related to perceptual enhancement of the respond signal but a conditioned anticipatory response. If this is the case, it is not surprising to find a weak relationship between HR deceleration and RT. In addition to the confounding effect of temporal conditioning, the orienting response to the warning signal may influence autonomic responses during the PI. If the PI is of short and constant duration, the conditioned and orienting responses may mask the HR response related to performance. To solve these problems and to enhance the probability of finding a HR response related to attentional performance, the PI may be extended to partition the orienting components and presented for random variable durations to reduce the possibility of temporally conditioned anticipatory responses. Thus, the use Of an extended vari— able PI increases the probability of finding a HR response which would reflect an instrumental act by the organism to lower selected perceptual thresholds. The present experinent was designed to investigate the respiratory and HR indices of attention to RT signal and control nonsignal stimuli and to relate these indices to RT performance. A second purpose was to investigate the differences between respiratory and HR responses to temporally bound stimuli (fixed interstimulus intervals) and responses to stimuli not temporally bound to the stimuli (variable interstimulus intervals). Thus, changes in HR and respiration could be compared for Se attending to signal or nonsignal stimuli during schedules of fixed or variable interstimulus intervals. Method Subjects. Forty—eight male volunteers from introductory psychology courses at Michigan State University received extra course credit for serving as S8. As they appeared at the laboratory, they were assigned to experimental conditions according to a predetermined random schedule. Apparatus. Stimuli were programmed by means of magnetic tape and were presented automatically (Porges 8: Fitzgerald, in press). Reaction time was measured by a Standard electronic timer. The ambient noise level of the experimental room was 51 db. (re .0002 dynes/cm2). Room temperature was maintained at approximately 700 F. The physiological responses were continuously recorded on a Beck- man Type RS Dynogr’aph at a paper speed of 5mm/sec. The EKG recording sites were cleaned with 70% ethanol prior to the application of the electrodes. Zinc cup electrodes with a surface area of 3.114 sq. cm. and filled with cotton soaked in a 1% ZnSOu solution were used to record HR from EKG lead II. The HR was measured using a Beckman 9857 cardiotachom— eter. Respiration was monitored with a Parks Electronics A-in. mercury strain gage attached attached around the chest by a Velcro fastener. Changes in chest circumference were measured by a Beckman 9875B mercury strain gage coupler. Procedure. Subjects were randomly assigned to one of four groups of 12 §_s each. A two—by-two factorial design was used with the factors being task (RI‘ and control) and schedule of PI (fixed and variable). The RT groups were instructed to respond as rapidly as possible following the termination of the warning signal (green light) that remained illuminated for the duration of the PI. Simultaneous with the termination of the warning signal, the respond signal (red light) was illuminated. The respond signal was terminated when §_pressed a button. The control groups were instructed to merely watch the signal lights. The respond signal was presented for 60 msec. following the termination of the warning signal in the control, since it was only necessary to elicit a response to change in stimulus conditions. The S was seated in a comfortable armchair in a sound—attenuated room. After the pickups were attached, E read the instructions which informed §.of the task and then calibrated the recording equipment. The experimental room was dimmed and the stimuli were presented by the automatic equipment. Each S then received ten trials. The four groups had the same schedule of intertrial intervals, which varied among A5, 60, and 75 sec. according to a predetermined random schedule. The PI duration was 16 sec. for the fixed PI groups and varied among 16, 22, and 28 sec. according to a.predetermined random schedule for the variable PI groups. The RT groups were instructed to press a button on the arm rest at the cessation of the warning signal. The latency between the termination of the warning signal and the button press was recorded as §fs RT. Quantification of the data. The responses during each of the 10 trials were SCOPGd for HR and respiation. Each trial was divided into four periods: a "pre" period consisting of the 8 sec. prior to trial onset, an OR period consisting of the first 8 sec. of the PI, an attend period consisting of the last 8 sec. of the PI, and a respond period consisting of the first 8 sec. following the onset of the respond signal. Respiration rate was analyzed by counting the frequency of initiations and terminations of inspirations in each period. The mean amplitude of complete inspirations in each period was measured in millimeters of pen deflection. The HR.was obtained from a beat-by-beat read out of the cardiotachometer which transformed the R—R intervals into HR. The HR reading for each measurement interval of 1 sec. consisted of the HR in beats/min calculated from the R—R interval which was completed during that second. If more than one R—R cycle was completed during a given second, only the later cycle was scored. The HR response was evaluated by a comparison of mean HR and HR variance during the four periods and by a secondeby-second analysis of HR during the OR and attend periods. Results 'RespiratiOn'frequency. There was a significant increase in respiration frequency from 3.98 during the "pre" period to A.A9 during the respond period [F(3, 132) = 27.7, p_< .0005]. However, as illustrated in Fig. l, the respiration frequency during the four periods differed under the two task conditions [F(3, 132) = l2.A, p