As I have said in earlier posts, law enforcement investigation techniques depend largely upon the premise that all humans are physiologically identical (see "Convicting the Average DUI Suspect"). Without that presumption, field sobriety and breath alcohol tests would not be possible. Unfortunately, the premise is always false.
I’ve written about many examples of physiological differences — from person to person and within one person from moment to moment — which will directly alter breath or blood alcohol testing (see, for example, "Diabetes and the Counterfeit DUI", "GERD, Acid Reflux and False Breathalyzer Results", "The Effect of Anemia on Breath Tests" and "Do Breathalyzers Discriminate Against Women?"). Yet another example of variability is body temperature. Put simply, an individual’s body temperature will have a direct effect on the results of a breath test.
The effects of changes in body temeprature from the norm of 98.6 degrees on breath testing has been discussed in an article entitled "Body Temperature and the Breathalyzer Boobytrap", 721 Michigan Bar Journal (September 1982). If because of illness, for example, the body temperature is elevated by only 1 degree Centrigrade (1.8 degrees Fahrenheit), the 1:2100 breath-to-blood partition ratio used by breathalyzers will be altered so as to produce a 7 percent higher test result. Higher body temperatures will, of course, result in even greater errors.
Dr. Michael Hlastala, Professor of Physiology, Biophysics and Medicine at the University of Washington, confirms this. In an article entitled "Physiological Errors Associated with Alcohol Breath Testing", 9(6) The Champion 18 (1985), he notes that even the average body temperature of a normal, healthy person "may vary by as much as 1 degree Centigrade above or below the normal mean value of 37 degrees Centigrade — or 1.8 degrees from the mean value of 98.6 degrees Fahrenheit". Not only can the normal mean body temperature of an individual vary from that of other persons, but the "temperature of any individual may vary from time to time during the day by as much as 1 degree Centigrade".
Result? The partition ratio for alcohol in blood is altered — meaning, according to Professor Hlastala, a 6.3 percent error for every 1 degree Centigrade increase or decrease from the presumed normal body temperature. Yet another example of how breathalyzers are not actually testing you. Instead, they are testing an "average" person who does not exist. But you are the one who gets convicted.