We regret that tables in this article by Tunstall-Pedoe et al (20 September 1997;315:722-9) contained several undetected errors because, instead of transferring the data from a wordprocessor file, as the author was mistakenly informed, the tables were rekeyed. The correct data are given below. The tables appear with the correct figures on our website, and reprints of the article containing the corrected tables are available from the author (h.tunstallpedoe@dundee.ac.uk).
The row numbers refer to rows of data in the body of the table, ignoring subheadings to columns. Table 2 Row 19—For men the multiplicative constant 95% confidence interval for physical inactivity in work for CHD deaths should have read 1.11 to 1.73 (not 1.11 to 1.34). Table 3 Row 1—For women in class 3 for previous coronary heart disease the percentage should have read 8.2 (not 8.2F). Row 3—For men in class 3 for previous coronary heart disease the figure for CHD deaths should have read 3.83 (not 3.783). Row 11—For men in class 3 for serum cotinine the figure for CHD deaths should have read 0.98 (not 0.89). Row 13—For women in class 2 for alcohol the percentage should have read 14.6 (not 414.6). Table 4 Row 1—For men the value for the first centile of height should have read 1.57 (not 1.57F). Row 15—For women in the fifth fifth of systolic blood pressure the figure for CHD deaths should have read 13.01 (not 13.1). Table 5 Row 8 —For men in the fourth fifth of HDL cholesterol the figure for all deaths should have read 0.71 (not 0.81). Row 13—For women the value of the 20th centile (first fifth) of blood glucose should have read 4.22 (not 2.44). Table 6 Row 2—For women in the third fifth of urinary sodium the figure for all CHD should have read 0.97 (not 01.97). Row 3—For men in the fifth fifth of urinary sodium the figure for CHD deaths should have read 0.92 (not 10.92). Row 8—For women the multiplicative constant 95% confidence interval for urinary potassium excretion for all deaths should have read 0.71 to 0.92 (not 0.81 to 0.92). Row 14—For men in the fifth fifth of carotenoid intake the figure for all CHD should have read 0.70 (not 01.70). Row 15—For women in the fourth fifth of caretenoid intake the figure for CHD deaths should have read 0.78 (not 0.87).