Tranquilizer use before and after coronary bypass operation

TRANQUILIZER

USE BEFORE BYPASS

LAURENS

AND

AFTER

CORONARY

OPERATION

D. YOUNG,” JOSEPH J. BARBORIAK~ and ALFRED J. ANDERSON+

(Received

22 December

1987; uccepted

in revised form

23 June 1988)

Abstract-High rates of tranquilizeruse in coronary disease have been previously reported. In patients being evaluated for coronary disease, tranquilizer use has been related to myocardial infarction, angina pectoris, neurotic traits, employment and the use of cardioactive and other medications. In thiq study. assessing biological, psychological and social responses of 1046 male patients to a questionnaire variables before and after coronary bypass surgery were related to tranquilizer use. The effect of CABG was to decrease tranquilizer use. Tranquilizer use prior to coronary bypass operations was related to neuroticism, myocardial infarction and taking cardiac and analgesic medication. Similar findings were noted after coronary bypass operations except that angina pectoris was added as a contributor to tranquilizer use. The multivariate analysis tended to confirm the univariate findings, except that neurotic traits did not make an independent contribution. These results indicate that in patients undergoing variables other than coronary atherosclerosis determine coronary evaluations. cardiac disease tranquilizer use. Neurotic traits play a secondary role. Relief of angina pectoris appears to be particularly important in reducing tranquilizer USC. The finding of angina pectoris after coronary hypnsc operation may have special significance for the continuation of tranquilizer use.

INTRODUCTION PATIENTS with known or suspected coronary disease report relatively high rates of tranquilizer use [ 11. In previous studies of a coronary arteriography population 121, our group has noted rates of tranquilizer use exceeding that of other medication including that of cardioactive medications. Factors including age, gender, previous myocardial infarction, neuroticism, angina pectoris, and using other medications of somatic versus may play a role in tranquilizer use. The relative importance psychological factors is still unclear. Therefore, studies of the relative contribution of somatic and psychological variables to tranquilizer use are of potential value. Coronary artery bypass graft surgery (CABG) is generally effective in relieving the symptoms of coronary artery disease. particularly angina pectoris. We have available a population of patients with coronary disease who completed a questionnaire. Their pre- and post-coronary bypass responses afford an opportunity to analyze the effect of CABG on trarlquilizer use. We were particularly interested in the relationship of pain relief by this operation to tranquilizer use.

METHODS

The coronary and an CABS CABS

study population was composed of 1046 men who had been referred for known 01. possible artery disease. All \uhjects completed a questionnaire at the time of coronary artcriography Identical follow-up questionnaire, an average of 24 months later. All patients in the study had performed between the time of the original questionnaire and the follow-up. The majority had withln several days of completing the Initial questionnaire. The population wa$ restricted to

*Division of General Hospital Psychiatry, Department of Psychiatry: tDepartment of Pharmacology; ZDepartment of Medicine. Medical College of Wisconsin, 8701 Watertown Plank Road. Milwaukee, WI 53226. U.S.A. 475

476

LAURENS D. YOUNG CI crl

men, because the operative rate of men is six times that of women and because some of the response scales, described later in this study, were constructed using a male population. Excluding women from the study was thought warranted as the introduction of possible additional variance could obscure small variable effects and the small reduction in population size would not measurably affect the probability of significant findings.

The previously described questionnaire includes risk factors and other clinical information pertinent [3]. The behavioral and psychosocial items in the part. several items were combined to form clinical These scale scores were then analyzed with respect independent variable, tranquilizer use was defined patient has been or is still taking tranquilizers questionnaire.

over SO0 items covering general health, daily activity. to coronary heart disease and other physical illness scale are the main focus of this study. For the most scales that were then used as dependent variables. to tranquilizer use. For the purpose of this study, the as a positive response to a question indicating that a over a 3 month period prior to completing the

Dependent variables The following clinical scales, which have been developed in our laboratory and used in numerous previous studies, were used as dependent variables. Anginu peckris was defined dichotomously as chest pain relieved by rest or nitroglycerin and brought on by exertion; whereas chest pain alone was persistent pain, irrespective of onset or relief. Myocardicrl itlfurction was defined by the usual clinical criteria including electrocardiographic findings, enzyme elevation and symptom picture. Coronrrr~ occlusion level was measured by coronary arteriography using the Sonea or Judkins techniques and quanitified by a previously described system [3]. To&d exwci.sr was assessed by an accumulation of responses assessing the patient’s daily activity level. Several items from the questionnaire were used as scales to assess behavioral and social variables. The rleurofic fruit scale (Table I) is a 20 item true-false questionnaire that has been used as an indicator of psychological distress in previous epldemiological studies [a]. The neurotic trait scale has some clinical validity in being significantly related to pertinent subscales of the Hopkins’ Symptom Checklist-90 [5]. The ungcr scale is a five item subscale of neurotic trait items that has not yet been clinically validated. but has face validity. It was included because of interest in the relationship of hostility and anger to coronary disease. The five items listed in Table II comprise this scale. The a~~nle scale (Table III) is a list of items which the patient may agree or disagree with. This continuously scored scale has fact validity in describing the psychosocial disintegration, which has been suggested to be part of the clinical picture of coronary artery disease [h]. The CPA-6 scale (not shown) is a continuous measure of coronary prone attitudes, based on questionnaire items that have been shown reliably associated with common measures of type A [7]. The alcohol, education and smoking scales arc derived from self-explanatory questionnaire items, that have been used in modified forms in previous studies [Xl.

1. 2. 3. 4. 5. 6. 7. 8. 9.

IO. 11. 12. 13. 13. IS. 16. 17. 18. I’). 20.

I often do things on the spur of the moment without stopping to think. I usually don’t like to talk much unless I un with people I know very well. I am easily sidetracked from things 1 start to do. Often I feel too tired even to do the things I like to do. I’m never quite satisfied with what I do. I have periods of days. weeks or months when I can’t get going. Even when other people praise my work. I am still dissatisfied. I have a hard time making up my mind about things I should do. It’s hard for me to feel close to others. On the whole, life gives me a lot of pleasure. I keep putting things off, and 1 don’t get as much done a\ others do. I lo\e my temper easily. Often, when I’m with a group of people, I feel left out - cvcn if they arc friends Much of the time. I’m not sure what 1 really want. I don’t enjoy many of the things other people seem to like. It seems to me that other people tind it easier to decide what is right than I do. Almost every time I finish doing something. I feel I could have done it better. I find it easy to drop or brcah with a friend. 1 complain a lot. I tend to keep people at a &stance.

ot mine

Tranquilizers

and coronary

TABLE H-ANGER

477

bypass

SCALE

1. I’m never quite satisfied with what I do. 2. I’m sometimes cross and grouchy for no good

3. Even when others praise 4. I lose my temper easily. 5. I complain a lot.

my work,

reason. I am still dissatisfied.

TABLE III.-ANOMIE

1. 2. 3. 4. 5.

ITEMS

With everything so uncertain these days, it almost seems anything could happen. What is lacking in the world today is the old kind of friendship that lasted a lifetime. With everything in a state of disorder, it is hard for a person to know where he stands. The trouble with the world today is that most people don’t believe in anything. People were better off in the old davs when evervone knew how he was exoected to act

Medication items Finally, a group of responses by the patients concerning the use of various medications have been included as dependent variables. These questions are phrased, as much as possible, in lay language and occasionally brand names of commonly prescribed items are used to better identify some drugs. These items were reported either singly or as groups such as antacids. In some cases, several individual medications were analyzed both individually and as groups, for example, nitrates, digitalis, and quinidine were analyzed individually and also as ‘cardiac drugs’ and water pills and blood pressure pills individually analyzed. were also grouped as ‘antihypertensives’. Originally patients before and after coronary bypass operation were divided into those taking tranquilizers and those not taking tranquilizers. Both continuous and dichotomous variables were then analyzed for significance by means of one way analysis of variance [9, lb]. After the univariate analysis, those variables that had a statistically significant relationship to tranquilizer use were entered into a stepwise multivariate regression equation in order of their univariate significance level. Those with the strongest relationship to tranquilizer use were taken first, those with the next strongest next and so on [ll]. This procedure was repeated until all variables that played a role in the regression equation were accounted for. Three stepwise multiple regression analyses were performed using variables as follows: (1) pre CABG tranquilizer use using pre CABG variables (pre x pre), (2) post CABG use using pre variables (post X pre) and (3) post CABG tranquilizer use using post CABG variables (post x post).

RESULTS

As indicated in Table IV, the overall number of men taking tranquilizers drops after CABG from 367 (35%) before CABG to 140 (13%) after CABG. By univariate analysis, myocardial infarction was positively associated with tranquilizer use both before and after CABG. Chest pain and angina pectoris are related to tranquilizer use only after CABG. The number of subjects complaining of chest pain and angina pectoris after operation was reduced by CABG as expected. TABLE IV.-TRANQUILIZER

No

Pre bypass Yes

367 679 83.1 84. I 72.8 70.8 66.2 41.4 167 + 74 I70 * 72

Number of subjects Chest pain Angina pectoris Myocardial infarction Coronary occlusion *One way analysis

USF FOR CARDIAC VARIAHLWS

of variance.

D*

No

Post bypass YCS

D*

NS NS NS 0.0001 NS

906 42.1 31.S 54.9

140 63.6 5.5.7 67.1

NS 0.0001 0.0001 0.00 I

LAURENS D. YOLING et01.

478 TABLE V.-TRANOUILIZER

cm FOR PSYC~IOS~CIALAND I)EMOGRAPI~IC‘VARIABLES

No

Pre bypass Yes

679 33.6 4.2 f 2.1 4.7 * 6.1 3.9 * 2.0 6.4 i- 3.9 1.8 * 1.3 1.6 F 1.3 13

367 40.3 4.5 i- 2.0 S.6 * 7.0 4.0 * 2.5 7.2 ? 4.1 2.0 + 1.4 I.9 f 1.4 24

On tranquilizers Number of subjects Employment (“A not) Education Alcohol (oz/wk) Anomv scale Neuroiic trait scale CPA-6 scale Anger scale Smoking score (O/6 now) *One way analysis

P* NS 0.001

0.05 NS 0.001 NS 0.02 0.001

No

Post bypass Yes

P*

906 35.7 4.3 + 2.1 4.9 t 5.6 3.7 t 2.4 6.1 Z 4.0 1.3 f 1.1 I.6 2 1.4 14

140 38.6 4.5 -t 2.2 5.0 i 6.5 3.9 i- 2.6 7.3 I 4.6 1,s + 1.2 2.0 + 1.5 18

NS NS NS NS 0.01 NS 0.01 NS

of variance

Table V shows the relationship of psychosocial items and scales to tranquilizer use before and after CABG. Before operation alcohol use, smoking, and education were related to using tranquilizers, but these relationships were not found after CABG. The positive relationship of neurotic traits and anger to tranquilizers was found before and after CABG. Employment, anomie and coronary-prone behavior were not related to tranquilizers before or after CABG. Table VI shows the univariate relationship of other medication use to tranquilizer use. Generally both before and after CABG, the use of other medication is positively related to tranquilizer use. TABLE VI.-MEDIC.ATION

VARIABLESHY IJNIVARIA.IEANALYSIS

Percent On tranquilizers Number of subjects Antacid Digitalis Propanolol Ouinidine Aspirin Nitrate Cardiac Analgesic Antihyper Hypnotic

medication use Pre bypass NV Yes />(N-Y)

67’) 21.4 8.8 3O.Y 2.9 37.4 37.2 3x 47 32 7

367 33.x 11.4 33.6 6.X 37.h 45.x 14 s4 30 20

0.01 NS NS 0.0 I NS 0.01 0.05 0.03 NS 0.0001

No YOh 18.0 21.3 20.2 7.1 47.3 15.7 34 55 41 4

Post bypass Yes p(N-Y) 140 26.4 36.4 37.9 12.Y 58.6 36.4 56 71 3Y 13

0.03 0.001 0.0s 0.03 0.02 0.0001 0.0 1 0.001 NS 0. OOOI

Table VII shows the major variables included in the three multivariate analyses, in order of their contribution to the multiple regression equation. Hypnotic medication, anticoagulants and myocardial infarction are the variables contributing most to tranquilizer use before CABG. Including nitroglycerin, antacid use, anger scale, educational level and alcohol intake as independent variables in the equation before CABG, about 18% of the total tranquilizer use variance can be explained. In the analysis of post CABG tranquilizer use using pre CABG variables, a different picture is seen. Hypnotic and antacid medication remain important variables, but myocardial infarction is not included. The pre CABG medication

R’

0.18

-

-

0.06267

Constant

0.09237 0.07867 0.07681

-0.09337

0.02157 6.1272E-’ 0.02643 -

-6.1977E-’

Education Alcohol Anger Smoking Angina Anomie Cardiac med.

Occlusion

0.19615 0.16661 0.12259 0.10071

NS

0.0118 0.0337 0.0379 NS NS NS NS

0.0079

0.0001 0.0001 0.0009 0.0068

0.22893 0.16014 0.13514 0.09940

E RtGRESSION

Anticoag MI Antacid Nitroglycerin

MUKfIPI

Hypnotic

VII.--STEPWISE

Prc CABG (pre variables) B P P 0.26763 0.17741 0.0001

TABLE

SHOWING VARIABLES

ASSOCIATED

0.04

0.08247

0.04810

-0.04728

0.07861

-0.09317

0.07749

0.01289

-4.3259E-’

0.06004

NS

NS 0.0786 NS 0.0433 NS NS NS

0.0214

NS NS 0.0569 NS

Post CABG (by pre variables) B P P 0.09419 0.08985 0.0280

ANALYSlS

3.148698-’

0.08503 -0.0120010 0.05984

0.01853

-5.6337E-”

0.10339 0.06542

Post CABG B

USE

0.06

0.11347 0.12231 0.11347 0.08749

-0.12179

-

0.07644 0.09562

P

(by post variables)

WITH TRANQUILIZER

0.01973 0.01973 NS NS 0.0099 0.0879 0.0045

0.0079

0.0937 0.0402 NS NS

P NS

o2 ;

8 i: 2 G

s p.

; c: _. w ;1

2

variables quinidine, analgesics and digitalis, as smoking, became more important for post CABG tranquilizer use. Together these variables explain 4% of the variance. In the analysis of tranquilizer use after CABG using post variables, angina pectoris became a major contributor to tranquilizer use. Anticoagulants, cardiac medication, myocardial infarction and anomie were also important contributors to tranquilizer use. These variables together could explain about 6% of the tranquilizer use. Slightly more tranquilizer use after CABG can be explained using post CABG variables than using pre CABG variables. DISCUSSION

The proportion of patients using tranquilizers after CABG is substantially less than that before CABG. In a statistical study of this type, caution is required in making clinical interpretations, but our findings suggest that patients having more severe coronary disease are more likely to take tranquilizers before and after operation. After CABG, particularly, those patients who fail to find relief from chest pain would be more likely to use tranquilizers. Neurotic issues are of secondary importance to the severity of heart disease symptoms when the effect on tranquilizer use is considered. Before CABG a number of variables, including myocardial infarction history and use of a number of cardiac and other drugs, make major contributions to tranquilizer use. Following CABG, angina pectoris, additional cardiac medication and anomie became the variables most consistently associated with tranquilizer use. The association of multiple medication use with tranquilizers before and after operation suggests that chronicity and severity of heart disease is generally one factor in tranquilizer use that does not change much with operation. These findings suggest that severity of coronary disease is important to tranquilizer use before CABG. After CABG, the picture is less clear. Different coronary disease symptoms play a significant role in tranquilizer use. Myocardial infarction is more important before CABG, angina pectoris after CABG. The specific positive contribution of angina pectoris and chest pain to tranquilizer use after coronary artery bypass is an intriguing new finding. Because angina pectoris contributes to tranquilizer use after, but not before, CABG, patients who do not achieve the expected relief of angina pectoris from CABG may be at additional risk for continued tranquilizer use. Almost all patients who use tranquilizers after CABG have used them before CABG. Therefore, pain after CABG would appear to enhance continuation of tranquilizer use rather than create new tranquilizer use. The effect of neuroticism appears secondary to the effect of coronary disease in tranquilizer use because neurotic traits drop out as a significant multiple regression variable after CABG. These findings suggest that the underlying physical disease or physical disability is more important than psychological reactions to tranquilizer use in coronary disease. This relationship could also be explained as related to the general pain and discomfort experienced by patients which is of either physical or emotional origin. Such an explanation is suggested by the strong association between tranquilizer use antacids, nitrates). Thus, patients who and pain medication use (analgesics,

Tranquilizers

and coronary

481

bypass

experience pain, whether angina pectoris or an atypical pain syndrome, could be but the underlying common variable expected to appear relatively ‘neurotic’ accounting for tranquilizer use would be the pain experience. This simple explanation would not be expected to account for all tranquilizer use, as some patients with non painful coronary disease may still have severe emotional reactions to their condition which can contribute to tranquilizer use. Two variables suggesting emotional distress that may have obscured neurotic traits, anger and anomy, were related to tranquilizer use pre CABG and post CABG, respectively. These findings are consistent with an interpretation that tranquilizer use in coronary disease is reversible and more closely related to the individual’s coronary disease picture than to personality variables. These findings suggest that symptomatic heart disease is an important contributor to tranquilizer use before and after CABG and is independent of emotional factors. The clinical value of tranquilizing medication to coronary disease cannot be established by these data, but these findings suggest that treatment of painful heart disease symptoms with effective cardioactive drugs could reduce some reliance on tranquilizers recently. It has not been established, for example, whether increased use of recently available inotropic anti angina1 drugs might reduce tranquilizer use. Answers to these questions will require further research. Acknowledgements-We are indebted to members providing some of the data used in this study.

of the

Milwaukee

Cardiovascular

Registry

for

REFERENCES I. 2.

3. 4. 5.

6. 7. 8.

9. IO Il.

WHEX~LY D. Stress and the Heart, 2nd Edn, pp. 39-63. New York: Raven Press, 1981. LD, BARB~RIAK JJ, ANDERSON AJ. Determinants of tranquilizer use in patients presenting for coronary arteriography. Paper presented at the 15th European Conference on Psychosomatic Research, 9-14 September 1984, London, England. YOUNG LD, BARBORIAKJJ, ANDERSON AJ. HOFFMAN RG. Attitudinal and behavioral correlates of coronary heart disease. J psychosom Res 1980; 24: 311-318. BERK~IAN LS, BRESLOW C. Health and Ways of Living. New York, Oxford, 1983. LECANN AF, YOUNG LD, BARB~RIAK JJ, ANDERSON AJ, GOLDSTEIN MD. Comparison of the Milwaukee neurotic trait screening scale with the symptom check list. Percept Motor Skills 1987; 85%8SY. JENKINS CD. Psychological and social percussions of coronary disease. New Engl J Med 1971; 284: 246255, 307-317. YOUNG LD. BARBORIAK JJ. Reliability of a brief scale of assessment of coronary prone behavior and standard measures of type A behavior. Percept Motor Skills 1982; 5.5: 1039-1042. YOUNG LD, BARBORIAK JJ, ANDERSON AJ. Coronary pain behavior, stress and alcohol consumption: the relation to coronary artery disease. In Stress nnd Alcohol Use (Edited by P~HARECKY LA, BRICK J). New York: Elsevier. 1983. EVER~I-T DS. The Analysis of Contingency Tables. New York: John Wiley, 1977. NIE NH, Hu CH, JENKINS TG, S~EINBRENNER K, BERT DH. Statisticul Package for the Social Sciences, 2nd Edn. New York: McGraw-Hill. 197.5. COHEN J, COHEN P. Applied Multiple RegressionlCorrelation Analysis for the Beha\,ioral Sciences. Hillsdale, NJ: Lawrence Erlbaum, 1975. YOUNG