Psychological Factors and Vulnerability to Psychiatric Morbidity after Myocardial Infarction

A sample of 97 males suffering from myocardial infarction was studied in order to determine the influence of psychologi­ cal variables and emotional states on psychopathology dis­ played by patients in the course of their cardiological recovery. Discriminant analysis revealed that depression in the coro­ nary unit and nonuse of problem-solving strategies were the most sensitive variables to correctly classify psychiatric and nonpsychiatric cases (76.6% of the total sample). Multiple regression analysis of the psychological variables and emotion­ al states showed that severity of psychopathology was directly related to early depressive reaction and use of avoidant strate­ gies. The prevalence of psychiatric disorders was 55.5%, with depression (RDC) being the most frequent diagnosis (59.4%), followed by ‘irritable dysphoria’ (27%) and anxiety disorders (RDC; 21.6%).


Introduction
Myocardial infarction (MI) is a life-threat ening event, in which patients must cope with the immediate traumatic situation, and deal with the long-term threat to their health and well-being after leaving hospital. Rehabilita tion programs have demonstrated the impor-tance of psychological factors in compliance and recovery [1] and there are retrospective and prospective epidemiological studies which underline the significant relations among initial emotional responses and car diological and psychological outcome [2][3][4]. On the other hand, several authors have re ported a psychiatric morbidity in the acute Prof phase of MI ranging from 39 to 66% [2,5] and there is evidence that this prevalence of psy chopathology remains unchanged 1 year after leaving the coronary unit [2,4], Lloyd and Cawley [6] identified two groups of patients with psychiatric symptoms after MI: one with previous psychopathological disorders and high scores in neuroticism, in whom psychiat ric disorders were persistent, and the other without previous psychopathology, who re cover from their psychiatric symptoms within 1 year. In consequence, in the last decade, the role of psychological factors in determining the long-term adjustment of coronary patients has been a matter of considerable interest [7], Among personality factors, neuroticism has been related to a poor emotional adjustment [8] and a higher risk of psychiatric disorders after MI [4], while psychoticism seemed to be a reliable indicator of good psychological ad justment [9]. Moreover, several studies have demonstrated that the use of denial mecha nisms after MI is predictive of a good psycho logical and social outcome [10][11][12].
The present study attempts to determine the influence of coping strategies, personality variables and emotional states in the early phase of MI on later psychopathology dis played by patients. Among personality vari ables, two dimensions of the Gray [13] per sonality theory are studied, in order to relate results to a biologically based conceptual framework.

Sample Selection
A series of 110 consecutive males under 65 years suffering from a first MI and admitted to the coronary unit of the Clinic Hospital of Barcelona were inter viewed for inclusion in the study; illiterate subjects and patients suffering from concomitant diseases were ex cluded. Patients selected were informed of the nature and conditions of the research program.

Instruments
Personality Variables Susceptibility-to-Punishment Scale (SPS) [14], This is a 36-item scale including situations and behavioral habits to identify subjects selectively responsive to anxiety and fear stimuli, according to the Gray [13] theory of anxiety. Subjects must answer each item with 'yes' or 'no' (c.g., 'Do you generally avoid giving your opinion about topics you know nothing about?', 'Are you a shy person?'. 'Do you tend to keep in the back ground during fights?').
Susceptibility-to-Reward Scale (SRS). Developed by Muntancr and Torrubia [15], also from the Gray personality theory, this assesses the tendency to selec tively respond to stimuli suggesting emotional well being, reward and consummatory behavior (c.g.. 'Whenever possible, do you tend to demonstrate your skills?' 'Do you like novelties?'). These two question naires (SPS and SRS) have been developed from sever al studies of their relationship with other well-known personality measures.
Eysenck Personality Questionnaire (EPQ) [16], This was validated in Spain and derived from the Eysenck [17] personality theory, which identifies ncurotieism (N), extroversion (E) and psychoticism (P) as independent, biologically based personality dimen sions. The questionnaire also has a control scale (I.) to evaluate sincerity in answering.
Bortner Questionnaire (BQ) [18]. Validated in Spain. This is a 14-itcm self-administered scale to assess type A behavior, which is considered as an inde pendent coronary risk factor [19], The subject must indicate on a horizontal line where he falls on the dimension on each item. [20]. This is derived from Lazarus' theory of stress and devel oped to identify thoughts and strategies used by sub jects to cope with stressful events. The questionnaire contains eight subscales (confronting, distancing, selfcontrol. seeking social support, accepting responsibili ty, escape-avoidance, problem solving and positive reappraisal).
Beck Depression Inventory (BDI) [22]. A self-report index of mood, derived from cognitive theories of depression.
Clinical Interview Schedule (CIS) [23]. This instru ment records psychiatric symptoms and psychopatho logical states, which can be scored on a scale of severity 188 García/Valdés/Jódar/Ricsco/de Flores Psychiatric Morbidity after Myocardial Infarction from 0 to 4. From the 22 items contained in the inter view it is possible to derive an index of clinical severi ty The interview was administered by two psychia trists, who were specifically trained in the use of the instrument.
Social-Functioning Schedule (SF) [25], This is a structured interview to assess social functioning in nonpsychotic patients by exploring 12 different aspects of daily life. The interviewer must evaluate social diffi culties through visual-analogue scales, and there is also a familial version to obtain additional information from relatives.

Procedure
All selected patients were interviewed by two psy chiatrists in the coronary unit, 3-4 days after being hospitalized, and were informed of the nature of the study. During this first contact, when informed con sent was obtained, all personality questionnairesexcluding the WCQ -were administered after record ing medical and general data according to the protocol. Moreover, social functioning preceding the MI (SF1; patients and spouses) and emotional state in the coro nary unit (STAI1, BD I1) were also evaluated.
One month later, patients were interviewed to de termine social functioning (SF2), and two trained psy chiatrists administered the CIS and established psy chiatric diagnoses according to RDC. To evaluate the interrater reliability (separate interviews for each rat er), a k correlation analysis was calculated from 20 interviews to determine the degree of classificatory agreement (k = 0.86). In addition, the psychometric instruments to evaluate emotional states (STAI2, BDI2) were also administered. Finally, the WCQ was administered, relating to the coping strategies used after leaving the hospital.
One year later, data related to cardiological and occupational evolution were recorded and patients were interviewed to assess the evolution of their men tal state during this period. Patients were asked about their emotional state (type of mood, presence of anxi ety or depression), their physical state (sleep pattern, degree of activity, level of energy), their mental state (attention, mental performance), their occupational activity and their need of psychopharmaceuticals or psychiatric consultation.

Analysis o f Data
Statistical analyses were computed using paramet ric and nonparametric measures (t of Student-Fischer, U of Mann-Whitney, y} and Pearson coefficient, and discriminant analysis was performed according to the Wilks' X method). Data were processed by an Apple Macintosh using the Statvicw 512 program and multi variate analysis was carried out in the Calculation Cen ter of the University of Barcelona, using the BMDP program.

Results
From the 110 consecutively interviewed pa tients, 9(8.1 %) died, and 4(3.6%) refused to be included in the study, so the final sample con sisted of 97 patients (88.1 % of the initial popu lation), with a mean age of 50.21 ± 7.61 years. Seventy-six (69%) were married, 44 (40%) came from a rural environment, and 13 ( Table 2 shows the psychiatric morbidity and emotional states of patients 1 month after suffering from the MI. This sample was com posed of 69% of the patients included in the first evaluation, and constitutes 60% of the initially interviewed total population. Pa tients who did not attend this second evalua tion said they lived too far from the hospital (60%) or had no time or opportunity to attend (10%). Eight patients (26.6%) were out of reach and one (3.3%) died 10 days after being discharged from the hospital. The diagnosis of 'irritable dysphoria', not included in the RDC, describes a group of patients character ized by presenting chronic irritability, which interferes with their occupational activity and complicates their social and familial relations. This state of irritability, already present be fore the MI, was a salient feature of these patients who, on the other hand, showed nei ther clinical manifestations of anxiety nor met all RDC requirements for minor depression (patients did not show depressive mood, as criterion A states).
In order to identify the variables that de termine the appearance of psychiatric disor ders, discriminant analysis among the tw groups (psychiatric cases and noncases) was performed, using the most significant vari ables of bivariant analysis EPQ-N, STAIT-T l, BQ, escape-avoidance, problem solving. STAI-S1, STAI-S2, BDI1, BDI2, and SF1).  Values are mean ± SD. 191 fearful stimuli, as Gray postulated for neurot ic subjects [13]. Moreover, Ml patients scored high in trait anxiety, although not reaching the level found in other coronary samples, and particularly in the study of Sykes et al. [7], in which STAI-T scores were notably higher (36.09 ± 9.45). Nevertheless, emotional hy perreactivity was the common finding ob tained by the different psychometric tools to assess personality variables. Type A scores were similar to those found in other American [26] and Spanish studies on coronary patients [27,28] using the BQ and, as would be expected, type A prevalence was noticeably higher than that obtained in the general popu lation (25.8 versus 14.4%). Ml patients showed a much higher preva lence of psychopathology than found in other Spanish studies of the general population (19%) [29] and of patients suffering from bronchial asthma (18%) [30], systemic lupus (34%) [31 ], neoplastic diseases (35%) [32] and coronary heart disease (44%) [5] using the same instrument (CIS). Depression and anxi ety were the most frequent disorders (59.4% of the total), as in other previous studies [2,3], but the rate in the two diagnoses was quite dif ferent (depression appeared to be twice as fre quent as anxiety). Such an epidemiological discrepancy may be explained by at least three factors. First, we used RDC to establish psy chiatric categories, and minor depression may be overrepresented by being the recom mended diagnosis when depressive and anxi ety symptoms are interlinked. Second, studies on psychiatric morbidity tend to use psycho metric tools to assess psychopathological states rather than using diagnostic criteria to establish independent psychiatric categories. In the studies in which RDC were used, the prevalence of depressive disorders was very similar to that found in our sample [33], Third, 27% of patients were classified as suf fering from irritable dysphoria', which is not a RDC psychiatry category, but a convention al diagnosis to describe a chronic emotional state tending to easy irritability, potentially included among anxiety disorders. Although these reasons may account for the relatively higher prevalence of depression in our coro nary patients, the fact is that their STAI-S scores in the coronary unit were markedly lower than reported by Sykes et al. [7] in a similar sample (34.39 ± 10.73). On the other hand, there were more psychiatric anteced ents in coronary patients diagnosed as psy chiatric cases, but the relevant percentage of patients with psychiatric symptoms without past psychiatric history (27.2%) outlines the convenience of searching for other indicators of psychiatric vulnerability.
MI patients presenting psychiatric disor ders 1 month following the MI were more anxious and neurotic than noncases, and scored significantly higher in type A behavior. The sensitivity of this type A scale for dis criminating among psychiatric cases and non cases probably depends on the fact that BQ is an assessment tool which also records neurot ic manifestations of personality [9], With re spect to the coping styles, the most significant differences among the two groups were repre sented by the escape-avoidance strategy (more prevalent among psychiatric cases) and the problem-solving strategy (predominant among noncases), therefore confirming the general statement that coping effectiveness in patients suffering from life-threatening ill nesses appears to be negatively linked to fre quent use of avoidance [34], These findings do not confirm previous studies reporting the protective role of denial mechanisms in coro nary patients [10,12], which may be consid ered as a coping strategy conceptually close to avoidance and distancing. On the other hand, coronary patients who asked for psychiatric attention during the year following the MI were significantly more fearful, anxious, neu 192 García/Valdés/Jódar/Riesco/de Flores Psychiatric Morbidity after Myocardial Infarction rotic and introverted, which is in agreement with the predictions of the Eysenck [17] and Gray [13] personality theories on psychiatric vulnerability. In fact, these patients were sig nificantly more anxious and depressed both in the coronary unit and at home, and showed poorer social functioning before the Ml and 1 month later. Since mood state has not been evaluated before MI, we do not know if the poorer social functioning of coronary patients classified as psychiatric cases can be attribut ed to the presence of previous affective disor ders but, in any case, it would be related to the significantly higher occupational stress re ported by these patients (who also scored sig nificantly higher in type A behavior). The finding that a depressive reaction in the coronary unit and the nonuse of problem solving strategies are the two psychological variables most useful to discriminate psy chiatric vulnerability corroborates the gener alized observation that an early emotional reaction after MI predicts long-term psychopathological evolution [2,3,11], and that depressive symptoms and emotional distress are positively related to the use of avoidant strategies and negatively related to the use of problem-solving strategies [35]. These results are important for anticipating the cardiological and psychopathological outcome, since there is evidence that depressive disorders are the best predictor of cardiac events during the year fol lowing catheterization [36], and that high BDI scores constitute a significant risk factor for death or cardiac arrest in MI patients [26]. Moreover, among MI patients there is a sub group with high scores in trait anxiety and poor prognosis, for whom early discharge would be contraindicated [37] and, in agreement with our results, there is also proof that between onehalf and two-thirds of these psychiatric pa tients remain distressed at 1 year follow-up, if they do not receive specific psychiatric treat ment [38]. Therefore, our results may contrib ute to the early identification of patients at high psychopathological risk by means of the two discriminant variables (depression in the coro nary unit and nonuse of problem-solving strat egies), so that specific treatments can be pre scribed to prevent the chronic evolution of psy chiatric disorders and their negative effect on cardiological recovery.