To the Editor.—
I agree with Wikstrand et al1 that the results of the Metoprolol Atherosclerosis Prevention in Hypertensives Study have "important implications for clinical practice and public health policy." Consequently, we have to consider the 48% difference in total mortality between the metoprolol- and the diuretic-treated groups with care. It is a pity that these results are not stratified according to other coronary heart disease risk factors besides smoking. An unequal distribution of these risk factors between the two mortality groups could put a different face on the outcome of the study and may be of more importance for clinical practice than the magnitude of the decrease in mortality alone. In other words, who do we have to treat with a β-blocker, all hypertensive patients or a particular risk group?