رفتار نوع A، ارتباط اجتماعی و مرگ کرونر TYPE A BEHAVIOUR, SOCIAL CONTACT ANDCORONARYDEATH

INTRODUCTION Many studies have been published on the association between Type A behaviour (TAB) and the non-fatal manifestations of coronary heart disease (CHD): myocardial infarction (MI) and angina. The findings are mixed and controversial, but are generally interpreted as showing that at least some aspects of the pattern, notably hostility and hard-driving competitiveness, increase non-fatal CHD risk in healthy populations (e.g. Booth-Kewley and Friedman, 1987; Matthews, 1988; Houston et al., 1992). In contrast, the available evidence on the relationship between TAB and coronary death (CD) is sparse but striking, in that it suggests either no relationship, or the possibility of an inverse association. In a prospective study of 257 men with a history of CHD, Ragland and Brand (1988a) found that the CD rate for Type As, assessed using the Structured Interview (SI), was 58% of that found in men without Type A characteristics (Type Bs). Two other studies of MI survivors, whose behaviour pattern was assessed with the Jenkins Activity Survey, also found this inverse association, though in neither case was it statistically significant (Case et al., 1985; Shekelle, Gale and Norusis, 1985). Ragland and Brand (1988b) have also reported a prospective analysis of coronary deaths in the Western Collaborative Group Study cohort, where they found no overall difference between Type As and Bs, classified using the SI 22 years earlier. Intriguingly, they did find a significantly increased risk for Type Bs during the interval 11-15 years after assessment. Finally, Orth-Gomer and Unden (1990) have reported no difference in coronary mortality rates for SI-classfied As and Bs over a 10 year period. Interpretation of these findings is complicated by the possibility of change in TAB over time, either in response to the experience of CHD (Ragland and Brand, 1988a), and/or as a function of age (Sparacino, 1979). Some clarification may be gained by focusing on deaths from first coronary events, as Gallacher (1989) has suggested. Ragland and Brand (1988a) did provide an analysis of deaths within 24 hours of a first coronary event, and reported no A/B difference. However, their sample contained only 26 cases. Similarly, Orth-Gomer and Unden’s (1990) sample produced only 7 coronary deaths in men who were initially free of CHD. Some clarification may also be achieved by assessing Type A behaviour shortly before coronary death, rather than only 10 or 22 years earlier. Clearly, large-scale prospective studies which include repeated assessment of TAB are the ideal strategy to achieve these aims, but they are prohibitively expensive given the relative infrequency of coronary death, The case-control design, despite its well-known limitations, provides a way of acquiring large samples of coronary death cases with no CHD history, and estimates of the characteristics they exhibited shortly before death (Appels, 1992). One objective of the present study was to assess the relationship between TAB and coronary death using such a design. For reasons given in the methods section, the definition and measure of TAB used in the present study were those developed in the Framingham study (Haynes et al., 1978). The specific TAB elements assessed are shown below in Table 4. Risk of coronary death has also been found to be prospectively associated with limited social contact, integration or support (House, Landis and Umberson, 1988; Schwarzer and Leppin, 1989). Virtually all of this research has been pursued independently of that on TAB, despite theoretical and empirical grounds for combining them. Good theoretical arguments for studying TAB in its social and ecological context have been articulated (Smith and Anderson, 1986; Margolis et al., 1983), though little heeded. Empirically, associations between TAB and social activity (e.g. Spicer, Jackson and Scragg, 1993) raise the possibility of confounding and mediation processes. Moreover, Orth-Gomer and Lunden (1990) have provided evidence of an interaction effect whereby social isolation increases risk of coronary death in Type As, but not in Type Bs. The general objective of the present study was to examine the independent and joint effects of TAB and social contact on the risk of coronary death. In the light of the available evidence we hypothesised that social contact would be inversely related to risk of coronary death, but left the direction of the TAB and interaction effect unspecified. We were particularly interested in comparing the results of the present analysis with those from our case-control study of non-fatal MI in men, where we found that higher risk was associated with Type A status, but not with low social contact (Spicer, Jackson and Scragg, 1993).