Evolving heart
In 1948, 5,209 residents of a medium-sized New England town signed up for what would become the longest-running, systematic medical study in the world. The Framingham Heart Study, as it was called, was the first to show that smoking, obesity, and high cholesterol all increased people’s chances of developing heart disease. Six decades on, it’s also the first multigenerational human study to reveal that some of these same traits are actively undergoing natural selection.
The news should come as a surprise to many physicians. Ever since Charles Darwin, a prevailing attitude among medical practitioners has been that evolution does not operate in humans because modern medicine and culture have greatly leveled the playing field by homogenizing survival rates. The same sentiment has also been echoed by some leading evolutionary biologists, most famously the late Stephen Jay Gould.
Not so, says Yale University’s Stephen Stearns, who specializes in life history evolution. Survival rates have indeed evened out, particularly among children, yet human birth rates remain highly variable. Some people simply have more children than others. And if there’s variation in lifetime reproductive success, and if some heritable trait is associated with that variation, then natural selection must be acting. To demonstrate natural selection in humans, however, requires a multitude of data collected over successive generations. Enter Framingham.
In 2005, Raju Govindaraju, a medical geneticist at Boston University and the former director of the Framingham Heart Study Genetics Laboratory, approached Stearns with an idea: analyze the Framingham data from an evolutionary perspective. Govindaraju had read many of Stearns’s seminal evolution papers as a plant biology graduate student in the 1970s, and was sitting on a goldmine of Framingham data, so he gave Stearns a call. “That was one of the best phone calls I ever had,” Govindaraju says.
“The thing that immediately struck me,” recalls Stearns, “was that, gosh, we can actually study selection operating on a contemporary human population and thereby make clear to everybody that natural selection is operating on humans.”
Stearns hired a postdoc, Sean Byars, and together with Govindaraju and Douglas Ewbank, a University of Pennsylvania demographer, set to work analyzing a handful of medically relevant traits for their effects on women’s lifetime reproductive rates. They measured the statistical associations between the traits and family size in the first two generations of Framingham women to estimate the strength of natural selection and the potential genetic response to selection. Early results show that women with lower cholesterol, lower blood pressure, lower blood glucose, and women who conceive earlier in life and reached menopause at a later age, all had more offspring. As a result, a model based on the data (that also controlled for social factors that influence fertility) predicts that levels of all these genetically based traits will change over the next generation. “People, myself included, may have written off evolution in humans,” says Ewbank. “But it’s still there. It’s still happening.” But the same response won’t continue indefinitely. Obviously, says Stearns, “if selection continued to reduce cholesterol, we couldn’t build a brain.”
The researchers also plan to study several other traits of medical interest, including high-density lipoproteins, triglycerides, and bilirubin levels. The researchers have not yet looked at the genetics behind the trend, but neither did Darwin when he was studying natural selection, which can be demonstrated solely by showing statistically that a trait can be heritable and lead to larger family sizes.
“The findings are startling,” says Govindaraju. “This will set the stage for really a new way of thinking about modern populations,” and allow researchers to make short-term predictions about humanity’s future evolution. Peter Ellison, a Harvard evolutionary biologist who was not involved in the study, agrees. “We can now think about human phenotypes much more dynamically than we normally do,” he says.
Stearns and his colleagues, who plan to publish their findings later this year in a special supplement of the Proceedings of the National Academy of Sciences, are now looking at male reproduction and combining their phenotypic analyses with genome-wide single-nucleotide polymorphism data from Framingham participants to search for genetic signatures of the tradeoffs between survival and reproduction. “We’ll try to push Framingham as far as we can,” says Stearns.