In a New York Times column in July 2012, the filmmaker Errol Morris took a few paragraphs to ponder the likelihood of death by asteroid.

“NASA issued reassuring public statements [after an asteroid flew close to Earth in 2011], but I’m not so sure,” he wrote. “It’s about the size of an aircraft carrier. Okay. That seems pretty big to me. Do you mean I shouldn’t worry about being hit by a meteor the size of an aircraft carrier?”

He then asked two questions of his readers: Did they agree, as the physicist David Deutsch put it, that—at least asteroid-wise—“we live in an era of unprecedented safety”? And how confident were they in their choice?

The column, which included a space at the end for readers to submit their answers, was titled, “Are You an Optimist or a Pessimist?” But in a follow-up article in the Times the next month, Morris made a confession: “My quiz wasn’t really a test of the optimism or pessimism of the reader,” he wrote. “It was a test of the effect of typefaces on truth.” Each time someone opened the original column, he revealed, the paragraph quoting Deutsch would appear in one of six random fonts, and the differences between them were statistically significant. People who had seen Deutsch’s statement of reassurance in Baskerville were the most likely to believe it and the most convinced of their choice; at the other end of the spectrum, those who had seen Comic Sans were both less likely to believe and less sure of their answers.

“Can we separate the form of writing from its content?” Morris continued, asking a question he had already started to answer.

New research shows that the answer may hold for healthcare as well as apocalyptic flying space debris. In a small study published last week in the journal Patient Education and Counseling, researchers from the University of Manchester and Leeds Beckett University in the U.K. gave a group of 71 volunteers (half pregnant women, half midwives) a sheet outlining a post-delivery maternal-health program. Participants were given the information in one of four formats: a simple heading and easy-to-read font (what the researchers called “double fluent”), a complex heading and difficult-to-read font (“double awkward”), simple title and difficult font, and complex title and easy font.

The four study conditions: a) easy title and font; b) difficult title and font; c) easy title in difficult font;
d) difficult title in easy font (Patient Education and Counseling)

The volunteers were then asked to rate the program based on how appealing it was, how easy it was to understand, how risky it seemed, and how difficult it would be to follow.

Among the pregnant women, those whose sheets were printed in the easier font, regardless of title, perceived the program as easier to understand than those whose sheets were printed in a more difficult typeface. Those with the “double awkward” papers perceived it to be the most complicated of all. (The format of the information had no significant effect for the midwives, though. The researchers didn’t have any firm conclusions as to why, but hypothesized that their prior knowledge of the subject area may have made them less susceptible to differences in its presentation.)

Despite its small sample size, the paper is one more in a collection of studies pointing toward the same idea: When it comes to creating health information, it may be wise to consider style along with substance. A 2008 study published in the journal Psychological Science, for example, found that people were generally more willing to incorporate an exercise routine into their daily schedules when they read the instructions in a clearer font. In a 2007 study in the journal Annals of Pharmacotherapy, researchers from Harvard Medical School found that patients had a better understanding of how to take their prescription drugs, as well as benefits and possible side effects, when the labels included “larger fonts, lists, headers, and white space And other studies have evaluated the readability of everything from instructions for asthma management to information for cancer patients to guidelines for oral health in kids, arguing that the visuals of health information directly affect how well people adhere to what they read.

Health agencies, too, have tapped in to the potential power of font: The Centers for Disease Control and Prevention’s guide for health materials, called “Simply Put,” includes a section on choosing the right typeface. The Centers for Medicare and Medicaid Services has a similar guide, and so does the U.S. National Library of Medicine. All three contain similar, fairly intuitive advice: Use simple fonts (Times New Roman, Arial, Helvetica) that aren’t too big or too small (usually 12-point, though 14-point might be better for material geared towards seniors); serif letters, which the brain can process more quickly, should generally be used for body text; and avoid italics and underlining, which can be overwhelming to the eye.

But for all the details meticulously laid out by the CDC and others, the authors of the Psychological Science study summed it up best. “If it’s hard to read, it’s hard to do,” they wrote—whether that “it” is taking a pill, remembering to exercise, or trusting NASA to keep us safe from an asteroid that’s flying a little bit too close for comfort.