IJE Advance Access originally published online on September 9, 2008
International Journal of Epidemiology 2008 37(6):1406-1407; doi:10.1093/ije/dyn187
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Commentary: Type A behaviour and heart disease: no less inscrutable in Japan
* Department of Psychology, University of Cardiff, Cardiff CF10 3XQ, UK.
E-mail: gallacher{at}Cardiff.ac.uk
Accepted 12 August 2008
Ikeda et al. present the first large-scale prospective evidence on the association of Type A behaviour with heart disease from an Asian population sample.1 Follow-up of 86 361 Japanese over 10 years found no association in women and an inverse association in men. Men with the lowest Type A scores had a 32% greater risk of newly diagnosed heart disease and a 35% greater risk of myocardial infarction. Greater risk of similar magnitude was also shown for all other indices of heart disease although these did not achieve formal levels of statistical significance. Interest in this study lies in its cultural context. The main limitation is the assessment of Type A behaviour.
Type A assessment was by four items, each assessing a component of the behaviour pattern comprising competitiveness, speed and impatience, aggressiveness and lastly, irritability. These items were scored using a three point Likert scale and the scores summed to provide an overall score for Type A behaviour. The scores were then divided into four groups which, given the limitations of the scale's distribution, roughly represented quartiles. The rationale for collapsing the Type A scale is not given, although observation of the results suggests it was due to a non-linear association of Type A with some outcomes. The minimalist assessment of ‘Type A’ means that the study tells us little about Type A and heart disease, however, the study raises important issues in understanding associations of psychosocial constructs with disease outcomes.
At its core, Type A behaviour is a critique of American culture.2,3 By medicalizing the critique in terms of heart disease, operationalizing it in terms of behaviour and adopting a gentle writing style, Friedman made the critique acceptable. Nevertheless, it is primarily a challenge to a culture being increasingly predicated upon competition and material acquisition at the expense of more mature personal qualities. On this basis, the foundational issue for cross-cultural Type A research is one of meaning, with the primary question being does the critique transfer between cultures? Evidence suggests not necessarily in Japan. Although case–control studies lend some support for the Type A hypothesis,4,5 a cross-sectional study6 and now a prospective study1 report anomalous results. As Ikeda et al. conclude, Type A seems to be culturally contingent. Japan, with its distinctive culture and ‘salary-man’ emphasis is likely to have its own constellation of constructs that are associated with health outcomes. The concept of ‘Karoshi’, for example, meaning death through long working hours, is likely to be a more culturally meaningful health related focus than Type A behaviour.7,8
The epidemiologic imperative of population sampling has frequently resulted in abbreviated psychosocial measurement, justified often on the basis of face validity or a few associations of dubious relevance. The measurement of Type A behaviour in the Ikeda study, is a good example, but is far from unique. This approach is due to an uncritical acceptance of operational measurement in which the ‘true’ score of the test is defined as the actual score, as observed when a standard procedure is followed; error being assumed to be distributed at random around the point estimate. Amongst psychologists this approach is known as classic test theory.9 This approach is reasonable for the measurement of observed variables such as reaction time as strategies can be implemented to reduce or adjust for measurement error. It is inadequate for latent variables such as intelligence where there is no satisfactory operational definition of what is being measured. Fluid intelligence, known as ‘g’ amongst psychometricians, is indicated by various cognitive tests, but test scores and ‘g’ are not coterminous and any single test is considered to be a biased estimate of ‘g’; bias being due, for example, to test familiarity or test anxiety.10 Type A behaviour comprises three latent variables (competitiveness, time urgency and hostility) which are not easily measured. Having survived the Type A interview at the hands of Nancy Fleischmann (the ‘gold standard’ diagnostician of WCGP11 and RCPP12 fame) in which layers of verbal and behavioural subterfuge were uncovered, I know that no questionnaire has come close to ‘outing’ me as did Nancy. If epidemiology is to test psychosocial hypotheses rigorously, it can no longer appeal to face validity and larger numbers. Approaches to measurement, such as item response theory in which factor analytic methods are used to identify the latent variables underlying test scores, are required for definitive studies to be conducted.
The causal significance of associations between Type A and CHD remains unresolved. The analytic model that has driven the debate has been to establish Type A as an independent risk factor. This approach, a hangover from ‘black box’ epidemiology, has generated more heat than light, with critics declaring that independent associations demonstrate residual confounding and believers arguing that a new risk factor has been established. A more reasonable approach is to place any psychosocial risk factor within a known pathway of biomolecular and biomechanical causality. Adopting this approach is to show that the psychosocial factor is not independent, but an integral part of an increasingly detailed bigger picture. The use of directed acyclic graphs (DAGs) is useful in this process.13 DAGs make explicit the causal pathway being tested as well as the assumptions being made (correctly or otherwise). Anyone who met ‘Mike’, as Meyer Friedman was known to his colleagues, was aware that he found the trend towards competition and materialism destructive. He was not too specific on the mechanisms, however, but proposed that plaque disturbance and arterial sludging were likely triggers to critical events.2,3 This is hardly controversial with several studies linking coronary incidents to a variety of potential triggers.14–17 The difficulty from a psychosocial perspective has been identifying susceptibility to chronic emotional disturbance. A more challenging issue is whether psychosocial factors contribute to plaque formation. For progress to be made, hypotheses involving psychosocial factors need to be mechanism specific rather than black box. We should no longer be interested in the independence of risk factors, but in the interdependence between risk factors.
In conclusion, Ikeda et al. have provided a valuable service. They have taken the operational black box approach to its practical limit and found it wanting. There is now opportunity to ask more specific and integrated research questions, and in doing so, to assess psychosocial constructs more thoughtfully and within their cultural context. In an era of increasingly large and cost efficient study infrastructures,18 these goals are achievable.
Conflict of interest: None declared.
 |
References |
|---|
 Top References |
|---|
1 Ikeda A, Iso H, Kawachi I, Inoue M, Tsugane S, for the JPHC study group. Type A brehavior and risk of coronary heart disease: the JPHC study. Int J Epidemiol (2008) 37:1395–405.
2 Friedman M, Rosenman RH. Type A Behavior and Your Heart. (1974) New York: Knopf.
3 Friedman M, Ulmer D. Treating Type A Behavior and Your Heart. (1984) New York: Knopf.
4 Kayaba K, Yazawa Y, Natsume T, et al. The relevance of psychosocial factors in acute ischemic heart disease. A case-control study of a Japanese population. Jpn Circ J (1990) 54:464–71.[Medline]
5 Yoshimasu K. Relation of type A behavior pattern and job-related psychosocial factors to nonfatal myocardial infarction: a case-control study of Japanese male workers and women. Psychosom Med (2001) 63:797–804.
6 Hayano J, Kimura K, Hosaka T, et al. Coronary disease-prone behavior among Japanese men: job-centered lifestyle and social dominance. Type A behavior pattern conference. Am Heart J (1997) 134:1029–36.[CrossRef][Web of Science][Medline]
7 Fukuoka Y, Dracup K, Froelicher ES, et al. Do Japanese workers who experience an acute myocardial infarction believe their prolonged working hours are a cause? Int J Cardiol (2005) 100:29–35.[CrossRef][Web of Science][Medline]
8 Uehata T. Long working hours and occupational stress-related cardiovascular attacks among middle-aged workers in Japan. J Hum Ergol (Tokyo) (1991) 20:147–53.[Medline]
9 Lord FM, Novick MR. Statistical theoroes of mental test scores. (1968) Reading, MA: Addison-Wesley.
10 Reeve CL, Heggestad ED, Lievens F. Modelling the impact of test anxiety and test familiarity on the criterion related validity of cognitive ability tests. Intelligence (2008) (in press).
11 Rosenman RH, Brand RJ, Jenkins D, Friedman M, Straus R, Wurm M. Coronary heart disease in Western Collaborative Group Study. Final follow-up experience of 8 1/2 years. JAMA (1975) 233:872–77.
12 Friedman M, Thoresen CE, Gill JJ, et al. Alteration of type A behavior and its effect on cardiac recurrences in post myocardial infarction patients: summary results of the recurrent coronary prevention project. Am Heart J (1986) 112:653–65.[CrossRef][Web of Science][Medline]
13 Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiology (1999) 10:37–48.[CrossRef][Web of Science][Medline]
14 Carroll D, Ebrahim S, Tilling K, Macleod J, Smith GD. Admissions for myocardial infarction and World Cup football: database survey. Br Med J (2002) 325:1439–42.
15 Mittleman MA, Maclure M, Tofler GH, Sherwood JB, Goldberg RJ, Muller JE. Triggering of acute myocardial infarction by heavy physical exertion. Protection against triggering by regular exertion. Determinants of Myocardial Infarction Onset Study Investigators. N Engl J Med (1993) 329:1677–83.
16 Mittleman MA, Maclure M, Sherwood JB, et al. Triggering of acute myocardial infarction onset by episodes of anger. Determinants of myocardial infarction onset study investigators. Circulation (1995) 92:1720–25.
17 Gallacher JE, Sweetnam PM, Yarnell JW, Elwood PC, Stansfeld SA. Is type A behavior really a trigger for coronary heart disease events? Psychosom Med (2003) 65:339–46.
18 Gallacher JE. The case for large scale fungible cohorts. Eur J Public Health (2007) 17:548–49.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||