As the Medical Director of the Kansas Business Group on Health I’m sometimes asked to weigh in on hot topics—but rarely hotter than today’s topic—that might affect employers or employees. This is a reprint of a blog post from KBGH:

If I were to ask a grade-schooler what a normal blood pressure is, I doubt he could tell me 120/80 mmHg. Nor could he tell me that a normal pulse is somewhere between 60 and 100 beats per minute, nor could he tell me that a normal body mass index is less than 25.0 kg/m2. But if I asked that same kid what a normal body temperature is, I bet he would blurt out “ninety-eight point six” before I could even finish the question. It’s printed into our brains, like “1776” or “three strikes.” That number—98.6—comes from an 1851 study by Carl Reinhold Wunderlich, a German physician who studied over a million temperature readings in over 25,000 patients to settle on 37° Celsius (98.6° Farenheit) as the “normal” body temperature, and 38°C (100.4°F) as a fever.

But the thermometer was still a fairly primitive technology in 1851. Daniel Farenheit had only invented his version in 1717, and by the time of Wunderlich’s study thermometers were still mostly found in academic centers. It took a solid twenty minutes to get a reliable reading. Not only that, but Wunderlich was studying a population that was simply sicker on average than modern people are. His patients, even those who reported good health, were more likely to have elevated body temperatures due to inflammation, bad teeth, or smoldering infections with tuberculosis or other horrifying diseases. And nobody had air conditioning.

So in the modern era, when thermometers are ubiquitous and their technology mundane, maybe it should not surprise us that the dogma of 98.6 is now in doubt.

As infectious disease physician Richard T. Ellison III writes in the New England Journal of Medicine Journal Watch (paywall), more recent studies show an average temperature of 97.9°F (36.6°C). But even that number is not rock-solid. Our body temperatures appear to change with age. To reach this conclusion, researchers looked at almost 700,000 temperature measurements from three sources: Union Army Civil War veterans from 1862 to 1930, National Health and Nutrition Examination Survey I (NHANES I) subjects from 1971–1975, and the Stanford Translational Research Integrated Database (STRIDE) cohort from 2007–2017. They found that across all three studies, body temperature progressively decreased with age, equivalent to -0.003°C to -0.0043°C (-0.005°F to -0.0077°F) per year. This means that a 10-year old with a body temperature of 97.9°F could be expected to have a body temperature of only 97.6°F—a full degree Farenheit lower than our old standard!—by the time she is seventy.

Not only did the new analysis of these old studies show that body temperature declines with age, it showed that body temperatures have declined with time. That is, temperatures recorded in the 1970s were lower than in the 1860s through 1930s, and temperatures recorded from 2007 through 2017 were lower than those in the 1970s. Again, this is probably because of reduced inflammation as infectious diseases became less common and because of the gradual adoption of air conditioning. If your mind isn’t already swimming, the study also found that body temperatures change with the time of day, the body weight of the patient, and even with race and sex.

Does this mean that your doctor’s office is lying to you when they tell you your body temperature? Are the infrared thermometers being used in airports to detect people infected with corona virus a fiction? Your answer depends on what we are trying to detect with the elevated temperature. Our old friend 98.6°F does now appear to be a figment of our collective nostalgia. But the Centers for Disease Control and the World Health Organization, who both are most interested in detecting potentially sick people early in their disease, still stick to a fever definition of 38°C (100.4°F). Some have argued that this number should be lowered to 99.9°F or even 99.5°F to increase the sensitivity of finding sick people. The Infectious Disease Society of America, whose emphasis is more on the care of ill patients in critical care and hospital settings, uses the same definition of 38°C (100.4°F), though, albeit with the caveat that the temperature of 100.4°F must be maintained for at least an hour, or the patient must have a single value over 101.0°F.

All this is to say that nothing in medicine is simple. Humans are complex creatures living in a dynamic environment, and as much as we like to define simple, neat cutoffs for “abnormal,” sometimes it just isn’t possible. So the next time your doctor’s office tells you that your body temperature is “normal,” take all this into account. Is it normal in the Wunderlich sense, in that it’s near 98.6°F? Or is it normal in the context of the A/C in her office, your age, your sex, your race, your living in the 2020s, and the state of your teeth?