An authors' error appeared in the paper of Dr Engström and others (2000;54:104–7). In table 3, the number of never smokers without cardiac events should be 4137 (instead of 4537). Some percentages were not correctly rounded. Of the women who suffered cardiac events, 33 (32%) (instead of 33 (31%)) were single, 13 (12%) (instead of 13 (13%)) had diabetes, 3 (3.4%) (instead of 3 (2.4%)) had college/university education. The number of individuals with missing information was in part given in the methods and in part in the table, which caused some confusion. The numbers in the tables and the methods are, with exception of the never smokers, correct.