To the Editor.—
The recent article by Eckardt et al (1981;246:2707) has a serious flaw in the selection of controls. A control group can be defined as identical to the test group in all respects except for absence of the one factor being studied, in this case, alcoholism. In their study, at least five factors other than alcohol use differed between the two groups: age, smoking history, geographic location, hospital status (inpatient or outpatient), and the laboratory performing the analyses. While the authors adjusted their results for differences in age and smoking, and while there are no readily available data on geographic location affecting laboratory results, the other two factors may have been responsible for some of the differences observed.The concentration of some chemical constituents changes markedly with changes in posture. Statland et al1 showed higher total protein and creatinine levels in healthy persons when sampled after standing
Dufour DR, Irving J. Selection of Controls in Testing for Alcoholism. JAMA. 1982;247(18):2496–2497. doi:10.1001/jama.1982.03320430020010
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: