Coffee, myocardial infarction and coronary death

 

Message

Moderate coffee drinking (up to 5 cups a day) is not associated with much increased risk of heart attack or coronary death. Moderate increased risk associated with drinking 10 or more cups a day cannot be ruled out.

Reference

S Greenland. A meta-analysis of coffee, myocardial infarction, and coronary death. Epidemiology 1993 4: 366-374.

Search

The literature was searched using MEDLINE between 1986 and mid 1992. Bibliographies of identified papers were then examined. Studies were excluded if there was insufficient data to perform the meta-analysis or the associations investigated were not suitable (e.g. only angina prevalence was examined).

Eight case control studies and fourteen cohort studies were identified. No information is given on participants, i.e. number in each study, gender, age or other lifestyle habits. Participants in the cohort studies were followed from two to 26 years and cases included coronary disease, coronary death and/or myocardial infarction. Cases in eight of the nine case control studies were myocardial infarction.

The method of assessing coffee consumption, type of coffee consumed, and caffeine content is not reported. Consumption of coffee is given as a percentage of participants consuming more than a number of cups a day (e.g. 22% drinking more than 7cups per day). All cohort study results were based on total coffee consumption (caffeinated and decaffeinated). Three case-control studies were based on caffeinated coffee only.

Results

In the eight case-control studies, those drinking 5 cups a day had an increased risk of myocardial infarction of 40% compared with non-drinkers (geometric mean rate ratio 1.42, 95% confidence interval 1.30 to 1.55).

In the fourteen cohort studies, those drinking 5 cups a day had an increased risk of myocardial infarction or coronary death of 18% compared with non-drinkers (geometric mean rate ratio 1.18, 95% confidence interval 1.03 to 1.34).

Comment

There are several problems with this meta-analysis. It isn't the easiest read or the simplest description of what's going on. There is insufficient information concerning the participants and assessment of coffee consumption. Case control studies are subject to bias (e.g. errors in remembering coffee consumption before illness) and therefore not associated with high quality results.

Results varied between individual cohort studies, perhaps a reflection of the different outcomes used between them (i.e. coronary disease and/or death), as well as their size, varying between 36 and 940 cases. Furthermore we are not told whether studies had adjusted for risk factors, e.g. smoking, age, etc. (so assume at the very least that some had not), which would attenuate the results. Overall, the best we can say is that moderate coffee drinking is not associated with much increased risk of heart attack or coronary death. Moderate increased risk associated with drinking 10 or more cups a day cannot be ruled out.