For more than a century, 98.6 degrees Fahrenheit – 37 degrees Celsius – has been used as a landmark of human health. We've suspected for a while now that the number needs adjusting, but a new study shows it's not for the reasons we thought.
In spite of the cumbersome tools the German physician Carl Reinhold August Wunderlich used to establish an average healthy body temperature in 1851, he probably got it right. Remarkably, we may have just gotten colder.
"Our temperature's not what people think it is," says medical researcher Julie Parsonnet from Stanford University in the US.
"What everybody grew up learning, which is that our normal temperature is 98.6, is wrong."
It's a standard that has weathered the ages surprisingly well. Wunderlich took literally millions of measurements from tens of thousands of patients.
But the instrument he used to establish ranges of healthy and unhealthy temperatures (not to mention daily fluctuations, effects of age, and so on) was just over 22 centimetres long – cutting edge for its time, but crude by modern means.
"In the 19th century, thermometry was just beginning," Parsonnet notes.
Since then, a handful of studies have been critical of Wunderlich's measurement, prompting calls to drop it by a fraction of a degree.
Parsonnet and her colleagues were curious about whether the cause of the contrasting measurements was actually improved technology, or if it accurately reflected changes in our physiology.
To find out, the researchers dug through the medical records of nearly 24,000 Union Army veterans following the US Civil War to work out just how hot we ran around a century ago.
These numbers were then compared to around 15,000 records from an early 1970s national health survey and 150,000 records from a Stanford clinical data platform representing the early 2000s. In total, the team had details on more than half a million individual temperature measurements.
Sure enough, there was a clear, significant difference over time. Temperatures among those living at the end of the 19th century were slightly warmer. Men born in the 2000s, for example, were 0.59 degrees Celsius cooler than those born in the early 1800s, representing a steady decline of 0.03 degrees Celsius per decade.
The drop was similar for women, with a drop of 0.32 degrees Celsius since the 1890s.
With the difference confirmed, the team turned their attention to differences within each group, assuming they would each have had their temperatures measured with instruments of roughly similar precision.
Over the decades covered by the thousands of Civil War veteran records, they saw a gradual decline that matched the trend based on their original comparisons.
Since it's unlikely that thermometer technology or methods evolved at a rate that would explain this steady drop, we're free to ask whether something about our own bodies – or our environment – caused temperatures to slip south.
"Physiologically, we're just different from what we were in the past," says Parsonnet.
"The environment that we're living in has changed, including the temperature in our homes, our contact with microorganisms and the food that we have access to. All these things mean that although we think of human beings as if we're monomorphic and have been the same for all of human evolution, we're not the same. We're actually changing physiologically."
Improvements in health and nutrition could be a fruitful place to search for an explanation. Our increasing body masses would push metabolisms into warmer categories, but inflammation is linked closely with variations in body temperature, and a decline in chronic infections just might explain why we're a little less feverish.
Past changes could also provide some insight into where we're going in the future, as we head into a world of intense social and environmental change.
A warmer future just might see a return to a slightly warmer body temperature, especially if it comes with new diseases and declining health.
Wunderlich's magic number might still need an adjustment, but the good doctor would be pleased to know it wasn't a result of his dedicated hours of watching his foot-long thermometer.
This research was published in eLife.