In the last two decades, public health experts have urged Americans to make their hearts healthier. Blood pressure guidelines have been updated, the risks of smoking and diabetes have been emphasized, and dramatically more people have taken the cholesterol-lowering drugs called statins.

But for all of these steps towards prevention, how big has the payoff been?

Researchers posed this question in the latest issue of the Journal of the American Medical Association. To answer it, they used data from five large observational studies and followed thousands of Americans ages 40-79 who didn’t have cardiovascular disease. They matched people up by age, race and sex from the earlier era (1980-1990) and a later one (1996 through 2011). The pairs—14,009 in all—were tracked for up to 12 years to see who developed cardiovascular disease. The researchers then estimated how much of that was due to blood pressure, diabetes, smoking and high cholesterol.

They found that over time, the incidence of coronary heart disease—which includes events like heart attack and intense chest pain—dropped about 20%. “That's a positive message that all the efforts have proven to bear fruit,” says Michael Pencina, professor of biostatistics and informatics at Duke University and one of the authors of the study.

When the researchers looked at how much cholesterol, blood pressure and smoking influenced coronary heart disease, they found these risk factors mattered just as much in the later years as much as they had in the earlier era.

That each of these risk factors still matters is proof that we can be doing more to address them, Pencina says. “Prevention in the young is one frontier that we need to be exploring more actively.” Even though elevated risk factors in younger adults are known to lead to greater risk later on, most of the guidelines only begin for people at age 40, he says.

Diabetes was the lone risk factor that seemed to have a weakened impact on coronary disease over time. Researchers aren't quite sure why, but Pencina thinks that it may be because we’re better at detecting diabetes now than in the past. “If you had diabetes more than 20 years ago, it was really obvious that you had it—it must have been a really bad case,” he speculates. Now, it's possible that we’re picking up more people who have the disease, but a milder version of it.

“Twenty percent is good, but it’s nowhere near where we could or would like to get,” Pencina says. “There is still a lot of room for improvement."

 

This article originally appeared on Time.com.