The human body is getting colder. Since the nineteenth century, normal body temperatures have dropped by a fraction of a degree, according to a provocative study of more than 677,000 thermometer measurements taken from people in the United States since 18601.
People’s bodies are now, on average, cooler than the textbook figure of 37 °C, having fallen by a few hundredths of a degree per decade, estimates a team led by Julie Parsonnet, an infectious-disease epidemiologist at Stanford University in California. “If you ask a room of physicians, ‘What’s the normal temperature?’, they’ll tell you it’s 37,” says Parsonnet. She suspects that falling rates of chronic infections explain our cooler bodies.
The 37 °C figure for normal body temperature was first determined in 1851 by Carl Reinhold August Wunderlich, a German physician who took millions of measurements from some 25,000 people and reported a range of 36.2 to 37.5 °C. “It became just the standard. It was adopted in textbooks, and it was just what people believed,” Parsonnet says.
Nobody rigorously challenged Wunderlich’s figure until 1992, when a team at the University of Maryland in Baltimore tested 148 people participating in a vaccine trial and found that their temperatures averaged 36.8 °C2. A 2017 study of more than 35,000 people in the United Kingdom found an average of 36.6 °C3.
The 1992 study’s lead author, infectious-disease physician Philip Mackowiak, suspected that the rudimentary thermometers available to Wunderlich explained the discrepancy. He later tested one of Wunderlich’s thermometers — in the collection at the Mütter Museum in Philadelphia, Pennsylvania — and found that its reading was too high by more than 1 °C. Mackowiak concluded that measurement error was behind Wunderlich’s 37 °C average.
But Parsonnet says her team’s data suggest body temperatures really are cooling. The team looked at three data sets. In the earliest one, a database of 83,900 temperatures collected between 1862 and 1930 from veterans of the American Civil War, the researchers found that people born earlier tended to have higher temperatures than those born in later years, even when body temperatures were measured in the same period — presumably, with the same technology.
This suggests that improvements in thermometer technology were not behind the trend, Parsonnet says. “If it’s just the thermometers changing, the year that the temperature was taken should make the difference.”
Using the Civil War data, along with hundreds of thousands of measurements collected in the 1970s and between 2007 and 2017, Parsonnet’s team modelled changes in body temperature. Women born in the first decade of the nineteenth century had temperatures 0.32 °C higher than those of women born in the late 1990s; for men, the difference was 0.59 °C. Overall, temperatures dropped at a rate of 0.03 °C per decade, Parsonnet’s team reports in eLife (see ‘Chilling effect’).
Parsonnet thinks that lower rates of infection are probably the best explanation for the falling temperatures. Inflammatory immune responses to long-term infections such as tuberculosis and gum disease can elevate body temperature, she notes.
“If you looked at the great majority of people back in the nineteenth century, I’m sure literally all of them had a chronic inflammatory condition,” she says. “They lived to be 40 years old or less. They all had terrible dentition.” A small 2008 study of healthy volunteers in Pakistan, where tuberculosis is still relatively common, reported average body temperatures of 36.9 °C4.
That explanation is “intriguing and plausible”, says Jill Waalen, an epidemiologist at the Scripps Research Translational Institute in La Jolla, California, who reviewed the paper for eLife. None of the temperature measurements the researchers relied on spanned the period beginning in the 1940s, when antibiotics were introduced. Waalen says that a marked drop in body temperatures during this period would support the theory that infections explained the cooling trend.
Makowiak, though, isn’t convinced that body temperatures are falling. “I’m concerned because there are so many variables that are unaccounted for,” he says. For instance, the Civil War data do not indicate whether temperatures were taken orally or in the armpit (which can differ in the same person), nor the time of day they were collected (bodies tend to warm through the day).
“There’s no biological explanation that I find convincing,” Makowiak adds. “We’re talking about 200 years, which in the evolution of life is just a blink of the eye.”
But human physiology has changed in other ways, so it should be no surprise if our bodies are a bit cooler, says Parsonnet. “We’ve also grow taller, we’ve grown fatter. We’ve changed since the 1850s. Temperature is just another marker of that change.”
Nature 577, 306 (2020)
Protsiv, M., Ley, C., Lankester, J., Hastie, T. & Parsonnet, J. eLife 9, e49555 (2020).
Mackowiak, P. A., Wasserman, S. S. & Levine, M. M. J. Am. Med. Assoc. 268, 1578–1580 (1992).
Obermeyer, Z., Samra, J. K. & Mullainathan, S. Br. Med. J. 359, j5468 (2017).
Adhi, M. et al. J. Pakistan Med. Assoc. 58, 580–584 (2008).