[Skip to Content]
[Skip to Content Landing]
February 12, 1992

The Effect of Filtered-Coffee Consumption on Plasma Lipid LevelsResults of a Randomized Clinical Trial

Author Affiliations

From the Departments of Medicine (Drs Fried, Levine, Kwiterovich, and Pearson and Mss Wilder and Moy) and Epidemiology (Drs Pearson and Diamond), The Johns Hopkins Medical Institutions, Baltimore, Md. Dr Fried is now with Kaiser Permanente, Washington, DC, and Dr Pearson is now with the Mary Imogene Bassett Research Institute of Columbia University, Cooperstown, NY.

JAMA. 1992;267(6):811-815. doi:10.1001/jama.1992.03480060057030

Objective.  —To determine the effect of filtered-coffee consumption on plasma lipoprotein cholesterol levels in healthy men.

Design.  —Randomized controlled trial with an 8-week washout period followed by an 8-week intervention period during which men were randomly assigned to drink 720 mL/d of caffeinated coffee, 360 mL/d of caffeinated coffee, 720 mL/d of decaffeinated coffee, or no coffee.

Setting.  —Outpatient clinical research center in a university medical center.

Participants.  —One hundred healthy male volunteers.

Outcome Measure.  —Changes in plasma lipoprotein cholesterol levels during the intervention period.

Results.  —Men who consumed 720 mL of caffeinated coffee daily had mean increases in plasma levels of total cholesterol (0.24 mmol/L, P=.001), low-density lipoprotein cholesterol (0.17 mmol/L, P=.04), and high-density lipoprotein cholesterol (0.08 mmol/L, P=.03). No significant changes in these plasma lipoprotein levels occurred in the other groups. Compared with the group who drank no coffee, the group who drank 720 mL/d of caffeinated coffee had increases in plasma levels of total cholesterol (0.25 mmol/L, P =.02), low-density lipoprotein cholesterol (0.15 mmol/L, P=.17), and high-density lipoprotein cholesterol (0.09 mmol/L, P=.12) after adjustment for changes in diet.

Conclusion.  —Consumption of 720 mL/d of filtered, caffeinated coffee leads to a statistically significant increase in the plasma level of total cholesterol, which appears to be due to increases of both low-density lipoprotein and high-density lipoprotein cholesterol levels.(JAMA. 1992;267:811-815)