(April 2001)

The relation of hostility and anger to coronary heart disease and heart attacks has been debated for decades. At first, anger was part of the Type A aggressive, extroverted personality that was supposed to have increased risk of heart attack. The Type A personality eventually turned out not to be a predictable risk for heart attack. Then interest focused on a component of the Type A personality - anger and hostility. The evidence has increased, but is still not entirely persuasive. A new and impressive study in the May 2, 2000 issue of the journal Circulation provides additional supportive evidence.

Janice E. Williams, PhD and her colleagues from the University of North Carolina School of Public Health followed 12,986 black and white men and women who were given a ten-item anger scale and followed for six years. Only 7.7 percent of the study group had high scores on the anger scale, an additional 55 percent showed moderate anger, and the rest were considered to have low anger.

Those who score highest on the anger scale had almost a doubled risk of heart attacks compared to those with the lowest anger scores. When the study group was divided into those who had normal blood pressure and those who had high blood pressure, the findings were surprising. Those with normal blood pressure and the highest anger scores had almost a threefold increased risk of heart attack; but, those with elevated blood pressure showed no increased risk.

The results applied equally to blacks and whites. Those who showed moderate anger on the ten-item scale showed no increased risk of heart attack or severe coronary heart disease compared to those who scored lowest on the anger scale.

The authors concluded "anger proneness as a personality trait may place normotensive (normal blood pressure) middle-aged men and women at significant risk for coronary heart disease, morbidity, and death".

Critique: This study has some real strengths - a large group of people studied with a substantial follow-up period and a large number of major events (heart attacks) to analyze. But, there are also problems. First, the three groups - high, medium, and low anger - were not similar in many other respects. Those with the most anger smoked more, were less educated, and had significantly greater waist to hip ratios. Although the authors adjusted for this in the analyses, the fact is that these were very different groups from the start. Second is the finding that having more anger did not increase the risk of heart attack for those with high blood pressure; that really does not make much sense.

Third, there was no real dose response; moderate anger resulted in no more risk than the low anger group; only the small percentage with the highest anger scores showed increased risk. The lack of any increase at all in the moderate anger group compared to the low anger group mandates caution in interpreting the study.

Additionally, the authors used a single fairly simplistic anger score. There is nothing wrong with that, but this, too, makes it necessary to be cautious about drawing dogmatic conclusions.

Some studies have shown that the danger of anger and hostility in regard to increased risk of heart attack depends on whether the anger and hostility are inwardly directed or directed outwards. There is still a very good possibility that excessive anger is bad for your health - we just need more evidence.

Williame, J.E., et al. Anger proneness predicts coronary heart disease risk. Circulation. Vol 101. (May 2) Pgs 2034-2039. 2000.