Aims: To take appropriate account of duration of follow-up in estimating age-specific associations of 'usual' blood pressure and cholesterol with death from coronary heart disease.
Methods and results: Blood pressure and cholesterol were measured in 18 841 men at entry to the Whitehall study, and coronary heart disease mortality was recorded during a 26-year period. Biennial re-measurements of these risk factors in the Framingham study were used to obtain period-specific corrections for 'regression dilution'. For coronary heart disease deaths at ages 40-64, 65-74 and 75+ years in the Whitehall study, the mean times since baseline were 9, 15, and 21 years, respectively. In uncorrected analyses of coronary heart disease risk in each age range, a 10 mmHg lower systolic blood pressure at baseline was associated with proportional risk reductions of only 19%, 14% and 10%, respectively (P<0.001 for trend with age). After period-specific correction for regression dilution, a 10 mmHg lower usual systolic blood pressure about 5 years before coronary heart disease death was associated with risk reductions of 25%, 23% and 21%, respectively (trend P=0.1). Similarly, a 1 mmol . l(-1)lower blood cholesterol was associated with proportional reductions in coronary heart disease risk of 21%, 17% and 11% (trend P<0.001) in uncorrected analyses, by contrast with proportional reductions of 27%, 26% and 21% (trend P=0.5) after period-specific correction.
Conclusions: After making appropriate allowance for the longer interval between baseline measurements and death at older ages, reductions in the risk of coronary heart disease death associated with differences in usual levels of blood pressure or cholesterol at some particular fixed time prior to death were about as great in old age as in middle age.
Copyright 2001 The European Society of Cardiology.