Predictors of falling cholesterol levels in older adults: the Cardiovascular Health Study

Ann Epidemiol. 2004 May;14(5):325-31. doi: 10.1016/j.annepidem.2003.09.006.

Abstract

Purpose: To estimate 4-year change in serum total cholesterol levels in a population-based sample of older adults and identify independent predictors of cholesterol decline.

Methods: Prospective study of 2837 adults aged 65 years and older with serum cholesterol measured in 1992-1993 and 1996-1997.

Results: Mean serum cholesterol levels declined 6.3 mg/dl between the two examinations. Declines were greater in white (-7.3 mg/dl) than black (-1.4 mg/dl) participants and in those in good/excellent health (-0.9 mg/dl) vs. fair/poor health (-3.1 mg/dl; both p < 0.01). Factors associated with greater decline on multivariate analysis included age, male gender, and higher white cell count, albumin, and baseline cholesterol. Cholesterol levels declined 2.0 mg/dl per 6 year increment in baseline age and 6.8 mg/dl more in men than women after adjustment for other factors. C-reactive protein levels were unrelated to cholesterol change.

Conclusion: Declining cholesterol levels were associated with male gender, advanced age, weight loss, and white blood cell count but not with C-reactive protein levels. The role of declining cholesterol synthesis, due to as yet undefined age-related changes or to cytokine-mediated reductions related to illness, should be examined to help clarify the mechanisms of the sometimes marked declines in cholesterol levels observed at advanced ages.

Publication types

  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • Age Factors
  • Aged
  • Black or African American / statistics & numerical data
  • Cardiovascular Diseases / epidemiology*
  • Cardiovascular Diseases / ethnology
  • Cardiovascular Diseases / physiopathology
  • Cholesterol / adverse effects
  • Cholesterol / blood*
  • Female
  • Forecasting
  • Health Status
  • Humans
  • Male
  • Medicare
  • Prospective Studies
  • Risk Factors
  • Sex Distribution
  • Sex Factors
  • United States / epidemiology
  • White People / statistics & numerical data

Substances

  • Cholesterol