A Prospective Study of Blood Pressure and Serum Creatinine

Abstract

Objectives. —To describe associations of past and current blood pressure levels with serum creatinine levels and hypercreatinemia in the general population. Population. —1399 middle-aged residents of Washington County, Maryland, who had their blood pressure measured during a cancer screening campaign in 1974 (the "Clue" Study) and had their blood pressure and serum creatinine level measured in an atherosclerosis risk factors study from 1986 through 1989 (the Atherosclerosis Risk in Communities [ARIC] Study). Design. —Nonconcurrent prospective study. The outcome variables were serum creatinine level and hypercreatinemia (serum creatinine >115 μmol/L in men, >97 μmol/L in women) measured from 1986 through 1989. Main predictors were 1986-1989 blood pressure values (cross-sectional association) and 1974 blood pressure values (longitudinal association). Gender-adjusted associations were assessed and compared by linear and logistic regression. Results. —Both serum creatinine and hypercreatinemia were better predicted by past than by current blood pressure values. Creatinine values exhibited a gradual and statistically significant association with blood pressure levels measured in 1974, even across "normal" values of blood pressure and creatinine. The association with 1986-1989 blood pressure measurements was weaker and nonsignificant. The odds of hypercreatinemia in 1986-1989 were increased 1.5-fold to twofold, with a 20 mm Hg increment in 1974 blood pressure values, but the odds remained constant across 1986-1989 blood pressure values. Conclusions. —Blood pressure and creatinine level are associated in the general population. The observed association was stronger when a number of years had elapsed between the assessments of blood pressure and creatinine level. These findings are consistent with the hypothesis that blood pressure elevations, even below the hypertensive range, may induce early renal damage. (JAMA. 1993;269:488-493)