Type A behaviour and ischaemic heart disease.

It is a common medical belief that psychosocial factors, in some form or another, contribute to the risk of developing ischaemic heart disease (IHD). Anecdotal evidence comes from many of the great figures of medical history. William Harvey described a coronary patient whose illness was the result of 'anger and indignation which he yet communicated to no one' (Leibowitz, 1970). John Hunter's angina was brought on by 'agitation of the mind' and he was reported to have died from an attack provoked by a particularly irritating hospital board meeting in 1793 (Home, 1794). In 1910, Osier described the typical coronary patient as 'the keen and ambitious man, the indicator of whose engine is always at full speed ahead' a description very like that which has latterly been termed Type A behaviour. Since Osier, there have accumulated many clinical and epidemiological reports relating psychosocial factors to ischaemic heart disease (e.g. see the reviews by Jenkins, 1971, 1976). These data are of varying quality and do not provide consistent evidence of a role in aetiology. Perhaps the strongest evidence for a causal link between psychosocial factors and ischaemic heart disease comes from the studies of Type A behaviour (Dembroski, 1977). We would do well, therefore, to examine critically the evidence on Type A behaviour with a number of questions in mind. Has a causal link been established ? What is the magnitude of the association ? To what extent do variations in the frequency of Type A behaviour account for variations in the occurrence of IHD ? First, what is Type A behaviour ? One might have hoped for a definition which we do not have. It is described as an ' action-emotion' complex (Rosenman, 1977) that is characterized by impatience, a sense of time-urgency, competitiveness, striving for achievement, aggressiveness, hyper-alertness, restlessness, explosive speech and abruptness of gesture. The techniques used for measurement do not clarify the issue very much. In their initial cross-sectional studies, Friedman and Rosenman used clinical judgements to classify people and demonstrated an association between Type A behaviour and ischaemic heart disease in both men and women (Friedman & Rosenman, 1959; Rosenman et al. 1961). Subsequently, in a longitudinal study in California, they used a more formal version of clinical judgement. Type A behaviour was assessed by a structured interview. In classifying people, importance was placed not only on the content of the answers to the questions but on the individual's mannerisms and behaviour during the interview. This technique had the curious property that exactly 50 % of the population at entry into their longitudinal study were classified as Type A, there being no reason a priori why this should be so. It is not clear which characteristics of Type A individuals should be considered as representing the crucial part of the behaviour pattern and which are merely associated characteristics. Nevertheless, this global classification into Type A or Type B was remarkably revealing. Over an 8 | year follow-up period in the California study, Type A men had twice the incidence of ischaemic heart disease as Type B men (Rosenman et al. 1975). For these studies to be replicated, reproducible measures of Type A behaviour had to be developed. Jenkins developed an elaborate questionnaire designed for self-administration the Jenkins Activity Survey (JAS) which was correlated with the structured interview (73 % agreement) (Jenkins et al. 1967). Although the JAS assessment of Type A behaviour has been shown to be associated with IHD in cross-sectional studies (Jenkins, 1976), in Rosenman & Friedman's longitudinal study the JAS was less strongly associated with IHD incidence than was the structured interview (Jenkins et al. 1974). It seems that the JAS is missing some crucial component of the Type A behaviour pattern either because of the way the questions were asked, or the way the questionnaire was scored, or possibly