OBJECTIVE: Studies investigating the association of coffee consumption and serum cholesterol levels report conflicting results. In an attempt to resolve this controversy, we reviewed the literature to answer the question: Is there a true positive association between coffee consumption and serum cholesterol levels?

DATA SOURCES: A Medline database search dating back to 1965 was utilized. Key words used in the search were coffee, caffeine and cholesterol. Cholesterol was expanded to include lipoproteins and LDL-, HDL- and VLDL-cholesterol. All articles that presented cholesterol data in association with coffee consumption were examined for references missed by Medline. Recently published articles were located by a hand search through Current Contents and the latest monthly editions of Index Medicus.

STUDY SELECTION: Three reviewers made the decision to include all publications that met the following criteria: a) reported original experimental results; b) reported total serum cholesterol levels; and c) were published in peer-reviewed journals.

DATA EXTRACTION: Two to four articles were read and analyzed each week in chronological order. Independent data extraction was performed by three reviewers, who then met as a group once a week to cross-check the analyses.

DATA SYNTHESIS: A trend, representing the association between coffee consumption and serum cholesterol, was calculated for each study. The trend was based on the percent difference in cholesterol values between subjects drinking four or more cups of coffee per day in comparison to those drinking zero or less than one cup of coffee each day. In order to compare studies that reported different cup sizes and different levels of intake, weighted mean cholesterol levels were calculated. In studies discussing the data in terms of correlations, trends were established according to the r values provided by the authors.

CONCLUSIONS: The majority of studies demonstrated a positive trend in at least one subpopulation of their subjects, indicating that serum cholesterol levels increase with increasing coffee consumption. Stronger trends were seen among subjects drinking boiled coffee than in those drinking filtered, decaffeinated or instant coffee. However, most studies were not randomized clinical trials, and results can be countered by a number of biases prevalent in the studies, indicating the need for additional well-designed investigations to resolve remaining issues.

This abstract is reproduced with the permission of the publisher. Article only available in print.