Behav Res. & Therap!.
1915. vol
13. pp IXS-188. Perpamon Press Punted I” Great Brmm
CASE HISTORIES Instrumental
AND SHORTER
COMMUNICATIONS
blood pressure conditioning in out-patient hypertensives*
(Receiwd
2 Octohur 1974)
Now that the efficacy of instrumental conditioning of blood pressure in several specialized situations has been established experimentally (Benson et al.. 1971: Brener and Kleinman. 1970: Schwartz, 1972: Schwartz, Shapiro and Tursky, 1971: Shapiro. Schwartz and Tursky. 1971: Shapiro et al.. 1969; Shapiro. Tursky and Schwartz. 1970: Elder er al.. 1973) there is the matter of assessing the effectiveness of instrumental conditioning in a routine clinical application over an extended period. This is an important question since the effective management of essential hypertension must center around a procedure which can be employed easily on an out-patient basis and which the patient may eventually administer to himself, Therefore. the purpose of this study was to s,ee if essential hypertension could be lowered by the use of instrumental conditioning on an out-patient basis without a concurrent effort to alter the patient’s normal environment (e.g. eating and sleeping habits. daily activity schedule. and medication regimen). METHOD
Subjects Subjects (Ss) were drawn from both student and non-student populations. All reported they had been previously diagnosed as (essential) hypertensive by their family physicians. and no attempt was made to verify the diagnosis. No limitations as to sex. age. race or medication were set; volunteer patients were admitted to the study as they became available. A descriptive summary of the 22 participants is compiled in Table 1. There were I4 males and 8 females: the mean age was 50.23 with a range of 2HOyr, and 20 of the Ss were on medication for hypertension. Apparatus The apparatus consisted of a standard E & M Physiograph Six equipped with a model PE-IO0 Programmable Electra-Sphygmomanometer. With this system. it was possible to obtain an indirect measure of systolic and diastolic blood pressure displayed by a strip chart recorder. An aluminum panel containing high intensity red and green lamps was located on a small table about 1.5 m directly in front of the S. Procedure Informed consent was obtained at the outset and each S was asked to advise his physician that he was participating in the study. The next step consisted of setting up a fairly regular schedule for the S to attend conditioning sessions. Initially. the last four patients were subjected to daily (massed) rather than distributed (spaced) training sessions. Other than the rate at which training sessions were scheduled. there was no difference in procedure between the massed (N = 4) and spaced (N = 19) groups. In the case of spaced training. treatment sessions were distributed as follows: twice a week for 2 weeks. once a week for 2 weeks. two sessions spaced 2 weeks apart, and a final session I month later. Spaced training sessions were thus distributed over an 80-day period. Upon arrival for Session 2. the S was directed to urinate and then was asked to have a seat nearby until he or she was called. Ten to twenty min later. he or she was invited into the laboratory and seated in an upholstered reclining chair facing the signal panel of the apparatus. The cuff was secured to the upper left arm vvhich rested on the arm of the chair and care was taken to position the crystal microphone directly over the brachial artery. Next. his or her blood pressure was measured automatically every minute for a total of ten successive determinations. No training procedures were introduced at this time. and the mean of these readings was taken as the patient’s basul systolic and diastolic blood pressure. respectively. Two min later. the first IO training trials were given. making a total of 20 trials for the session. All sessions thereafter were composed of 20 trials each. with a 2-mm rest interval between the first and the last halves. In view of the data reported by Elder. Leftwich and Wilkerson (1974). reinforcement/feedback was always contingent on diastolic rather than systolic pressure changes. Generally, the training procedure conformed to that designated as strategy III in the Elder cr (11.experiment; more specifically. positive stimulus feedback (a green light) followed reductions in diastolic blood pressure, and negative stimulus feedback (a red light) accompanied trials during which blood pressure increased or no change occurred. In addition. verbal reinforcement was provided once during the first and second halves of each session in the form of “That’s good!” or “You just got a green light. keep up the good work!” in close temporal contiguity with the presentation of the green exteroceptive stimulus. The instructtons to the S at this point conststed simply of tellmg him that the aim was to teach him to lower his blood pressure. and that whenever the green light appeared it meant he had been successful and the red light signified he had not. Actually, Ss received a green light whenever their pressure dropped below basal. and a red light whenever their pressure increased or showed no change. This was done in order that no S would receive a green light for a blood pressure higher than the original basal pressure or the initial blood pressure at the start of the session. I85
186
CASE
HISTORIES
AND
SHORTER
COMML’NICATIONS
SRICED
DAYS
DAYS
Fig.
1. Upper panel: mean systolic per cent differences for the massed and Lowrr /ru~cl: comparable dtastolic per cent difference scores.
spaced
groups.
RESULTS
To provide within-sessions as well as between-sessions comparisons. two means were determined for each session: the first. Mean A. was obtained from the scores generated over trials I-10; the second Mean B, was based on scores from trials 1 IL20. Then the differences between each of these means and the Ss basal pressure were determined and expressed as a percentage of the Ss basal pressure. Mean per cent difference scores were prepared from the systolic and diastolic data for both the massedand the distributed-practice groups. The systohc data are displayed in the upper half, and the concomitant diastolic data are plotted in the lower half. of Fig. I. The black bars represent average performance for the first half of the session. Mean A; the white bars represent average performance for the last half of each session, Mean B. By comparing adjacent black and white bars, it can be seen that in every instance a within-sessions decrement in blood pressure was observed. Mann-Whitney U-tests between systolic and diastolic Means A and B of the massed sample gave U = I1 (p i 0.01). and for the comparison of diastolic Means A and B. ti = 19 tp c: 0.01). Similarly. statistical comparisons resulted in I/ = 6 (p < 0.001) for the spaced group systolic Means A and B. and c’ = 17 (p < 0.01) for the diastolic Means A and B. Inspection of the A and B means between groups revealed that the massed group showed a decrement in both systolic and diastolic pressure over ten training sessions. whereas the spaced group showed a systolic pressure decrement in six of the eight sesstons. with a similar reduction in diastolic pressure in seven of the eight sessions. In three of four cases. statistical comparisons between the two groups showed the massed group performed significantly better than the Ss in the spaced training condition (Mann-Whitney comparisons of the systolic Means A yielded U = 21.5. p < 0.05; of systolic Means B. Li = IS. p < 0.02: of diastolic Means A, U = 27. N.S. at the 0.05 level; of diastolic Means B. L = 24, p < 0.05). In the case of the group which received distributed practice, a follow-up examination was conducted 30 days after the last training session. and these data appear in Fig. I as Means A and B on Day 83. Although the systolic pressure showed a return toward basal level. the mean diastolic pressure tended to remain at the levels obtained during the final training session (Day 5 I). Since the spaced sample consisted of I I males and 7 females. a statistical comparison in search of a sex difference in performance was carrtrd out. C’ompartson of systolic Means A and B failed to show significant differences. Similar comparisons of diastolic Means A and B (see lower half of Fig. 2) revealed a reliable superiority in favor of the female group (Mann-Whitney comparison of Means A: U = 20.5, p < 0.05; of Means B: U = I I. p c: 0.01). DISC
In general. these et al. that essential
I’SSION
data are in agreement with the earlier reports by Miller, Shapiro et al. and Elder hypertensive patients can be trained through instrumental conditioning to lower their
CASE
HISTORIES
AND
SHORTER
187
COMMUSICATIONS
DAYS
DAYS
Fig.
2.
Upper paw/: systolic Long
mean difference Paul: comparable
scores for the male and diastolic difference scores.
female
sub-samples.
own blood pressure. Moreover, it was observed that massed training was superior to spaced practice and within the spaced group the females learned the instrumental response somewhat better than the males. A similar female vs male performance comparison within the massed sample was not warranted by sample size and available data. In general, the results of this study were not as dramatic as the earlier data from the Elder et al. experiment, Both studies yielded posttrve results. but only in the first case did the data satisfy Blanchard and Young’s (1973) criterion of at least a 20 per cent change in response before clinical significance has been achieved. A few of the Ss were able to reduce their blood pressure by as much as 20 per cent (see, for example, Ss 9 and 20 in Table I). but these instances were rare and not typical of either the massed or spaced sample. There are several plausible reasons why the out-patients in this study did not show a reduction directly comparable to that by the hospitalized patients in the previous study. First, 20 of the 22 patients in Table
I. Description
summary
of volunteer
out-patient
sample
188
CASt HISTORIESANI1 SHORER COMMl’NICATIONS
this study were on medication. while all of the subjects in the Elder cr rrl. experiment were free of medication for hypertension. It is reasonable to assume that anti-hypertensive agents such as reserpine and x-methyldopa interfere with instrumental acquisition of autonomic responses in several ways. including the direct or indirect blockage of neural transmission in circuits essential to the development of conditionmg. There are some data which indicate that this may be the case (e.g.. Elder and McLean. 1973). Another, and perhaps more important. difference between the two studies was the fact that patients in the hospital sample were admitted to the study only after a thorough examination and appropriate laboratory tests failed to show evidence of secondary hypertension. In contrast. no such preliminarv systematic examination and tests were possible in the present study: all patients reported they had been diagnosed as essential hypertensives. Nevertheless. several of the out-patients who volunteered for this study had long histories of hypertension: some of these and undoubtedly some of the others did not fit the criteria for a diagnosis of essential or primar); hypertension (Pickering. 1968. 1970). Although it may be likely that instrumentally conditioned reductions in arterial pressure may be possible in cases of secondary hypertension. theoretical considerations suggest that the best results are to be found in patients whose arterial pressure has been elevated by primary rather than secondary sources. An additional difference between the two studies centers about the fact that Ss in the first study were hospitalized whereas those in the present study were not. There are many data in the learning literature which show that strong drives contribute to the strength of an instrumentally conditioned response. It may be that hospitalization generates a stronger drive to acquire self-control of blood pressure than exists among out-patients. In the first case. Ss’ life styles were interrupted completely: m the present study. they were unaffected except for the fact that most patients were taking orally administered antihypertensive medication. Stronger drive levels among the out-patient sample may have led to far greater pressure decrements than were observed in this instance. Finally. the two studies were different in that patients in the first had little opportunity for interpolated activity between sessions. whereas those in this study went about their regular daily routines without interruption. Retroactive interference from interpolated activity may therefore account. perhaps in part, for the superiority of the massed over the spaced samples in the present case. as well as the superiority of the Elder et al. sample over both groups used in the study. Irrespective of theoretical considerations and explanations. it is clear that instrumental regulation of essential hypertension is more effective with hospitalized than out-patient populations. Even so, these data provide additional support for the idea that essential hypertension can be regulated through instrumental conditioning of the blood pressure response even on an out-patlent basis, and that a subject’s relative degree of success is statistically predictable. Departnteut of Ps~cho/oyy, University of IVFWOr/ems. New Orlenm, La. 70122. C.S.A.
S. THOMAS ELDER NANCY K. EUSTIS
REFERENCES B. and SCHWARTZ G. (1971) Decreased systolic pressure through operant conditioning techniques in patients with essential hypertension. Science 173, 74(X742. BLANCHARD E. B. and YOUNG L. D. (1973) Self-control of cardiac functioning: a promise as yet unfulfilled. Psychol. Bull. 79, 145-163. BRENER J. and KLEINMAN R. A. (1970) Learned control of decreases in systolic blood pressure. Nature 226, 1063p 1064. ELM H S. T.. LI.FTWICH D. A. and WIL~I KSO\ L. (1974) The role of systolic versus diastolic contingent feedback in blood pressure conditioning. Ps!,chol. Rec. (in press). ELDI R S. T. and MCLI AS J. H. (197-l) Motor and visceral learning in rats injected with varying amounts of reserpine. Paper read at American Psychologtical Association meeting. New Orleans. ELDER S. T.. Rc~z Z. R.. DIAHLIX H. L. and DILLIINIX>FFEK R. L. (1973) Instrumental conditioning of diastolic blood pressure in essential hypertcnslve patients. J. and. Bdar. Anal. 6. 377-382. KIMMEL H. D. (1967) Instrumental conditioning of autonomically mediated behavior. Psychol. Bull. 67, 337-345. MILLER N. E. (1969) Learning vlsccral and glandular responses. Scicncc 163, 434445. MILLER N. E. (1970) personal communication. MILLER N. E., DICARA L. V.. SOLOMON H.. WI.ISS J. M. and DWORKIN B. (1970) Learned modifications of autonomic functions: a review and some new data. Circular. Res. 26 and 27, Suppl. 1. PICKERING R. T. (1968) High Blood Prc~.ssurc. Grune and Stratton. New York. PICKERING R. T. (1970) ~~p~,rtcrl.sio,~: Ctrusc\. COII.SL’(/II(‘II(.(,S. tr,rr/ Muna(lu~le,~t. Churchill, London. SCHWARTZ G. E. (1977) Voluntar) control of human cardiovascular interaction and differentiation through feedback and reward. Sciczjlcc, 175, 9cY.i. SCHWARTZ G. E.. SHAPIRO D. and TURSI(I B. (1971) Learned control of cardiovascular interaction in man through operant conditionmg. P.~~chosom. Met/. 33. 57-61. SHAPIRO D.. SCHWARTZ G. E. and Tr RSL~ B. (1972) Control of diastolic blood pressure in man by feedback and reinforcement. P.s~~hop/~~.~io/o~/~~ 9. 296-304. SHAPIRO D.. TI,RS~~~ B.. GIXSOX E. and STI Rh M. (1969) Erects of feedback and reinforcement on the control of human systolic blood pressure. Scicwc 163, 588-5X9. SHAPIRO D., TCRSKY B. and SCHWARTL G. E. (1970) Differentiation of heart rate and blood pressure in man by operant conditioning. Ps~cl~o.~om. Med. 32, 417-423. BENSON H.. SHAPIRO D., TURSU