AGE-SPECIFIC CORRELATION ANALYSIS OF LONGITUDINAL BLOOD PRESSURE DATA

Abstract

Rosner, B. (Channing Laboratory, Harvard Medical School, Boston, MA 02115), C. H. Hennekens, E. H. Kass and W. E. Miall. Age-specific correlation analysis of longitudinal blood pressure data. Am J Epidemiol 106: 306–313, 1977. To evaluate the predictive value of blood pressures for future levels, longitudinal measurements were analyzed among Welsh subjects from age 5 to 74 at entry. The measurements were taken on 863 Individuals from the Vale of Glamorgan in 1956, 1960, 1964, and 1971 and on 734 individuals from the Rhondda Fach in 1954, 1958, 1964, and 1971. The tracking correlation, defined as the correlation between blood pressure readings on the same person taken at two different times, Is used as a descriptive measure of the magnitude of the association. The tracking correlations range from 0.25 to 0.6–0.7 with most of the increases occurring before age 20. The tracking correlation depends on initial age, sex, and time interval between measurements. The age-specific tracking correlations decrease as time interval between measurements increases. For a given time interval, the age-specific correlations are slightly higher for females than for males. These findings are very similar for each of the two regions studied.