If you've ever read anything about the colonies of bacteria that live on and inside you, you'll no doubt have come across the neat little 'fact' that microbial cells outnumber human cells in your body by a ratio of around 10:1.
You'll find it in scientific papers, magazine articles, TED talks, and popular science books, and while it does a good job at illustrating just how crucial bacteria are to the existence of human beings, it isn't actually true.
In 2016, a review of more than four decades of research into the human microbiome found that there is zero scientific evidence to back this oft-cited factoid up.
Instead, the ratio looks to be about 1.3-to-1, with the average human playing host to around 100 trillion microbes, give or take. But even that isn't the whole story.
To figure out the true number, a team of biologists led by Ron Milo from the Weizmann Institute of Science set out to review all the available literature on the microbe populations that live inside us.
They found that for a man between 20 and 30 years old, with a weight of about 70 kg (154 pounds) and a height of 170 cm (about 5'7) - they call him the 'reference man' - there would be about 39 trillion bacterial cells living among 30 trillion human cells.
This gives us a ratio of about 1.3:1 - almost equal parts human to microbe.
So where did the 10:1 ratio come from, and why did Milo and team have to be so specific about the particulars of a 'reference man', instead of just coming up with a ratio for the plain old average human?
The origin of the 10:1 ratio has been traced to a paper published in 1970 by American microbiologist Thomas D. Luckey, who estimated that there are 100 billion microbes in a gram of human intestinal fluid or faeces.
Because there are about 1,000 grams of these materials in the average adult, he said, that equals a total of 100 trillion microbes. (Bear in mind, that every single one of these figures has no backing in scientific evidence whatsoever, as Ed Yong has noted over at The Atlantic.)
Seven years later, well-known microbiologist Dwayne Savage took this vague estimate, mashed it up with the fact that there are about 10 trillion human cells in the average human, and came up with the 10:1 ratio.
Everyone, from fellow scientists to the public alike, took that factoid and ran with it, and it wasn't until 2014 that someone made an effort to fact-check it.
Judah L. Rosner, a molecular biologist and geneticist from the National Institutes of Health, wrote a letter to Microbe Magazine insisting that more recent estimates for human cell numbers were nowhere near 10 trillion.
In fact, coming up with a cell count for the average human would be basically impossible, as Ed Yong explains:
"More recent estimates, he noted, put the total number of human cells at anywhere from 15 trillion to 724 trillion, and the number of gut microbes at anywhere between 30 trillion and 400 trillion. Which gives a ratio that can best be expressed as ¯\_(ツ)_/¯."
This is where the reference man comes in - narrowing down the sex, age, weight, and height of a hypothetical human made it a whole lot easier for the reviewers to figure out what the average number of human cells would be.
For their reference man, it's about 30 trillion, the researchers estimate.
On top of that, Milo and his team also discovered that microbe cell counts in the colon - where Luckey got his original figure from - have regularly been overestimated in the scientific literature.
"[W]hen previous studies made their estimates, they used the density of bacteria per gram of 'wet content' of the colon, times the volume of the entire alimentary canal," Lindsey Kratochwill has noted at Popular Science.
"But, these researchers argue, the bacteria density of the colon is much higher than the rest of the tract, so assuming that the entire alimentary canal is as bacteria-filled as the colon is would be overkill."
Accounting for this, and the fact that we have a far higher concentration of bacteria in our guts than in other organs and body parts (which means you can't take a sample from the colon and say it's representative of the entire body a la Luckey), Milo's team came up with an updated and more scientifically accurate estimate of 39 trillion microbial cells, based on the available evidence.
But even that 1.3:1 ratio - 39 trillion microbial cells to 30 trillion cells - isn't really something we should be citing in our textbooks and scientific papers going forward, as Ed Yong argues.
"These new estimates might be the best we currently have, but the studies and figures that Milo amassed come with their own biases and uncertainties," he says.
"My preference would be to avoid mentioning any ratio at all - you don't need to it convey the importance of the microbiome."
The review is published in PLOS One.
A version of this story was originally published in 2016.