Independent and interactive effects of apolipoprotein E phenotype and cardiorespiratory fitness on plasma lipids

PURPOSE: To examine whether the association Apolipoprotein E (APo E) phenot ype with plasma lipids is influenced by physical fitness level. Also, to ex plore the interactive and independent relative contributions of Apo E pheno type, fitness (or physical activity), and other modifiable factors to varia tion in plasma low density lipoprotein (LDL-C) and high density lipoprotein (HDL-C) levels at baseline and over a seven-year follow-up. METHODS: Physical fitness (duration of a graded treadmill test), ADo E phen otype, plasma LLL-C and HDL-C, and covariates were measured at baseline and seven years later in a bi racial cohort of young adults, aged 18-30 years at baseline in 1985-86 (N = 3629), from the Coronary Artery Risk Developmen t In Young Adults (CARDIA) study. RESULTS: Fitness did not influence the associations of Apo E and LDL-C or H DL-C. The independent effects of several modifiable variables (changes in K eys' score, smoking, oral contraceptive use, education, body weight, alcoho l intake, and fitness), when combined, contributed considerably more than A po E to the variance in LDL-C changes (6.74% or 8.71% for combined modifiab le variables vs. 1.27% or 0.90% for ApoE, in women or men, respectively) an d HDL-C changes (13.11% or 12.66% fur combined modifiable variables vs. 0.1 2% or 0.02% for ApoE, in women or men, respectively). The pattern of findin gs was similar when self reported physical activity was substituted for fit ness. CONCLUSIONS: Changes in modifiable factors, including fitness, may be stron ger correlates of changes in LDL-C and HDL-C over time than the immutable f actor, Apo E phenotype. Ann Epidemiol 2001;11:94-103. (C) 2001 Elsevier Sci ence Inc. All rights reserved.