University of Minnesota

Western Collaborative Group Study

Study Category: The Cohort Studies (1947-1972)
Year Begun: 1960
Principal Investigator(s): R.H. Rosenman was P-I for the incidence study (1960-1969). D.R. Ragland and R.J.Brand were the P-I's for the mortality study (1982-1983).


The Western Collaborative Group Study (WCGS) was designed to test the hypothesis that the so-called Type A behavior pattern (TABP) – “characterized particularly by excessive drive, aggressiveness, and ambition, frequently in association with a relatively greater preoccupation with competitive activity, vocational deadlines, and similar pressures” – is a cause of CHD. Two additional goals, developed later in the study, were (1) to investigate the comparability of formulas developed in WCGS and in the Framingham Study (FS) for prediction of CHD risk, and (2) to determine how addition of TABP to an existing multivariate prediction formula affects ability to select subjects for intervention programs.

Participants and Methods of Procedure:

3524 men aged 39-59 and employed in the San Francisco Bay or Los Angeles areas were enrolled in 1960 and 1961. In addition to determinations of behavior pattern, the initial examination included medical and parental history, socioeconomic factors, exercise, diet, smoking, alcohol consumption, diet, serum lipid and lipoprotein studies, blood coagulation studies, and cardiovascular examination. Men continuing in the study were re-examined annually in order to obtain an interim cardiovascular history and ECG. Endpoints were a history of classical angina pectoris without apparent myocardial infarction (MI), symptomatic MI, and unrecognized MI. Follow-up for CHD incidence was terminated in 1969. Other investigators conducted the follow-up for mortality in 1982 and 1983.


TABP was positively associated with incidence of CHD at 2 years, 4.5 years, and 8.5 years of follow-up.  [Comment:  Unexplained changes in baseline data occurred in the final 8.5-year report.   In comparison to the reports at 2 and 4.5 years of follow-up, the final report shows 28  more men excluded from the population at risk because of prior CHD at baseline (141 instead of 113), 28 fewer men in the population at risk (3154 instead of 3182), 5 more men with the Type A behavior pattern pattern (1589 instead of 1584), and 33 fewer Type B men (1565 instead of 1598).  The report does not mention these changes, but does state that final ratings of behavior pattern were made “without knowledge of intake-history or measurements” in order to avoid possible bias.  However, the changes in baseline data  exist in the published papers, and  raise serious  questions about the integrity of the baseline database.  How did it happen that the final report showed 5 more Type A men at baseline despite having 28 fewer men overall?]

TABP was not associated with 22-year risk of CHD death. 4 Analysis of deaths during 4 intervals — 1/60-12/69, 1/70-8/74, 9/74-2/79, and 3/79-12/83 — showed that TABP was not significantly associated with CHD death in the first, third, and fourth intervals, but was inversely associated with CHD death in the second interval, i.e., Type A men had lower risk of CHD death than Type B men during the interval that began immediately after follow-up for nonfatal events ended. Among men who survived an initial CHD event for at least 24 hours, risk of CHD death was significantly lower in Type A than Type B men. Comment: How to account for these discrepant results? Were they due to some unknown factors that modified a true association or to some bias introduced in the conduct of the study? However one chooses to answer the question, these results cast substantial doubt on the robustness of the TABP hypothesis.

Results for other factors such as age, serum cholesterol, systolic blood pressure, relative body weight, hematocrit, cigarette smoking, and ECG abnormalities were similar to those obtained in the FS. CHD risk scores calculated from measurements of these factors according to the FS formula and the WCGS formula were correlated >0.80, and the frequency of CHD events observed in WCGS was close to the frequency expected based on the FS prediction formula after adjustment for length of follow-up. Comment: The results confirmed the importance of the major modifiable risk factors and provided strong support for the idea that replication and consistency of CVD risk factor studies is the main argument for generalizability of the risk score.

In men age 39-49 at baseline, the WCGS formula for prediction of CHD risk based on the factors listed above predicted that 45.6 of 145 cases would occur in the upper decile of risk. When TABP was added to the formula, the expected number of cases in the upper decile increased slightly to 47.5. Similar results were observed for men age 50-59 at baseline.8 Authors’ comment: “It appears that similar results can be expected for any of the traditional risk factors when assessed as a last addition to a list of other risk factors. This observation raises important questions about the number and strength of additional independent risk factors that might be needed to substantially increase the effectiveness and efficiency of selection of subjects for intervention programs. It further suggests that the strength of a particular risk factor may not be as important from the point of view of intervention as the ability to safely and conveniently achieve even a moderate risk reduction in a large number of persons.” (HB)


Rosenman RH, et al. Coronary Heart Disease In The Western Collaborative Group Study: A Follow-Up Experience Of Two Years. JAMA 1966; 195:130-136.

Rosenman RH, et al. A Predictive Study Of Coronary Heart Disease: The Western Collaborative Group Study. JAMA 1964: 189:15-26.

Rosenman RH, et al. Coronary Heart Disease In The Western Collaborative Group Study: A Follow-Up Experience Of 4½ Years. J Chron Dis 1970; 23:173-190.

Rosenman, RH, et al. Final Follow-Up Of 8½ Years In The Western Collaborative Group Study. JAMA 1975 233: 872-877.

Ragland DR, Brand RJ. Type A Behavior And Mortality From Coronary Heart Disease.
N Engl J Med 1988; 318:65-69.

Brand RJ, et al. Multivariate Prediction Of Coronary Heart Disease In The Western Collaborative Group Study Compared To The Findings Of The Framingham Study. Circulation 1976: 53; 348-355.