Type A is one of the clearest examples in health psychology of an idea that arrived with a striking result, spread far beyond what that result supported, and then partly unravelled — with a documented commercial interest sitting in the middle of the story. The arc is worth following in order, because each stage explains why the label still circulates decades after the claim behind it stopped holding.
The 1950s: Clinical Observation
Friedman and Rosenman were practising cardiologists who believed the standard risk factors of the day did not account for what they were seeing. They began describing a behaviour pattern among their patients — hurried, competitive, easily angered — and published the association with cardiovascular findings in JAMA in 1959.
What made this credible rather than anecdotal was that they built an assessment for it. The structured interview provoked the pattern rather than asking about it, and interviewer-rated behaviour is a stronger form of data than self-report, which matters for what happened later.
The 1960s and 1970s: The Cohort Result
The Western Collaborative Group Study enrolled a large cohort of middle-aged men from 1960 and followed them prospectively. Its mid-1970s report found Type A men developing coronary heart disease at roughly twice the rate of Type B men — a prospective design, a hard endpoint, a large effect.
That was enough for a 1981 review panel to treat Type A as an independent cardiac risk factor, and enough for the concept to escape into general circulation. The popular version simplified as it travelled, dropping the hostility component and keeping the busy, driven caricature.
The 1980s: Failure to Replicate
Then the effect stopped appearing. The Multiple Risk Factor Intervention Trial found no association between Type A and coronary events in its high-risk sample, and several other cohorts through the decade produced null results. A pattern emerged in which the strongest findings came from studies using the structured interview and the weakest from those using self-report questionnaires like the Jenkins Activity Survey.
This is the ordinary shape of a construct that was measuring something narrower than it claimed. The instruments were not interchangeable, which meant the literature was not really studying one thing, and the global pattern was too coarse to carry the effect.
The Funding Record
Running alongside this is a documented commercial history. Internal tobacco-industry documents, analysed by Petticrew, Lee and McKee and published in 2012, show that tobacco companies funded Type A research over an extended period. The interest is not hard to reconstruct: an industry facing the case that smoking causes heart disease benefits from a competing explanation located in personality.
The honest reading is precise. This does not show that the researchers falsified anything or that the WCGS result was invented. It shows that a research programme with a commercially useful conclusion had a heavily interested sponsor, which is a reason to weigh independent replication more heavily — and independent replication is exactly what did not come.
The 1990s Onward: What Was Salvaged
Rather than discard the work, researchers decomposed it, and the component analyses converged on hostility. Miller and colleagues' 1996 meta-analysis found hostility measures associated with physical health outcomes, and cynical mistrust and frequently activated anger — not time urgency, not achievement striving — carried whatever signal remained.
There was also a genuine intervention result. Friedman's Recurrent Coronary Prevention Project found that counselling directed at Type A behaviour reduced recurrence among heart attack survivors, which suggests the behaviour is modifiable even if the original risk claim was overstated.
What the Story Is Actually About
Type A is a case study in how a construct can be simultaneously overstated and not worthless. The two-box split was too coarse, the popular version dropped the part that mattered, the replication record is poor, and the funding history warrants scrutiny — and yet a real, narrower finding about anger and health came out of it.
That is the version worth carrying: a behavioural description with an interesting history and one durable component. See Type A personality and stress for how it plays out day to day, and the Type A / Type B test for your own pattern.