Purdue HHS researchers analyze best times to eat, exercise in a day

Written by: Tim Brouk, tbrouk@purdue.edu

“Breakfast should be your biggest meal of the day.”

“You should exercise first thing in the morning.”

“Don’t eat before bed.”

You’ve heard these nuggets of advice and more from amateur nutritionists and fitness gurus, but do they offer any scientific truth?

Heather Eicher-Miller headshot

Associate Professor Heather-Eicher-Miller

Purdue University College of Health and Human Sciences researchers Heather Eicher-Miller, associate professor in the Department of Nutrition Science, and Libby Richards, associate professor in the School of Nursing, published findings from cluster analyses on temporal nutrition intake as well as when and how hard to exercise in a 24-hour span.

The scientists utilized health data from about 1,000 Americans ages 20-65 from the Centers for Disease Control and Prevention’s National Health and Nutrition Examination Survey (NHANES) to find trends in when and how U.S. adults eat and move. They found a link between the diet and exercise time-based patterns and both waist circumference and body mass index (BMI), which is calculated by weight divided by the square of height. In particular, the group that ate similar amounts of energy at evenly spaced times throughout the day and exercised in the morning or evening had smaller waist circumferences and better BMIs.

While much more research must be conducted before concluding when it’s most optimal to eat or exercise, Eicher-Miller and Richards found the cluster of patients that consumed three calorically balanced meals spaced out evenly through the day had more ideal BMI scores. After evaluating the movement data, the Purdue researchers discovered those closest to the healthiest BMI scores of 18.5-24.9 participated in higher or more activity in the morning or evening. The purposeful exercise could consist of a brisk walk or more.

Libby Richards

Associate Professor Libby Richards

“The activity pattern was likely deliberate physical activity, more than a leisurely walk across campus,” Richards said. “Most likely, these folks are deliberately exercising, doing something to enhance their health. You have to deliberately step it up.”

On the nutrition side, the patients with balanced, moderate meals spaced throughout the day clustered to the healthiest level — away from patients who eat more of their calories in the morning and less through the rest of the day, and those who eat most in the early and late evening.

“I actually was a little surprised,” Eicher-Miller said. “I’ve heard that said before, too, that you should eat your biggest meal for breakfast. In our data set, there was a group that ate more in the morning. I did expect that would be the group with the best BMI, but that wasn’t the case.”

The researchers received three-year funding from the National Institutes of Health (NIH).

Research history

Eicher-Miller initiated the 24-hour temporal patterning project as an extension from previous research early in her career, which explored how diet quality is linked to the timing of eating. The fitness aspect was added to expand the work’s reach.

“We started to think bigger: What additional health behaviors could we include?” Eicher-Miller said.

For most of the work, Eicher-Miller and Richards were joined by Purdue Department of Statistics Associate Professor Anindya Bhadra; Saul Gelfand and Edward J. Delp, professors in Purdue’s Department of Electrical and Computer Engineering; and Erin Hennessy, a nutrition science assistant professor from Tufts University.

Since 2020, Eicher-Miller, Richards and their team have published seven papers from the initiative with a trio of articles currently under review.

Participant data

Eicher-Miller and Richards favor the national survey data of the NHANES as it was collected during physical examinations by healthcare professionals. With trusted data comes better analysis and results.

“We actually know how much physical activity they have because they wore (accelerometer) monitors instead of just asking them,” Richards explained. “There’s also bloodwork associated with it, so we really know if someone has high cholesterol and so forth. The survey is nationally representative. Participants are strategically selected so the data can be made to be representative of the demographics of the U.S.”

Richards noted the data set cut off at age 65 because most older Americans have different nutritional intake and exercise compared to those 40 years younger. Pregnant and nursing women were also excluded.

The team used the information on the number of calories participants ate at times throughout the day along with the amount of exercise they did at times throughout the day to create groups of people with similar patterns in the 24-hour window. For exercise, Richards found the cluster with healthiest BMI were most active between 9-11 a.m. or 6-9 p.m., while for diet, the healthiest group kept meals below 1,200 calories in the morning, afternoon and evening.

The researchers created four groups for diet and exercise to keep the sample sizes large because the larger the sample, the more robust the results. More patterns would mean smaller samples and therefore weaker results.

Much more to come

The researchers foresee the data set offering more insight into human health. Eicher-Miller and Richards expect to include other human behaviors to the mix, such as sleeping. Sedentary behaviors could also be examined. They would also like to expand the research to include children.

“We expect to see different patterns of behaviors in kids,” said Richards, who revealed more NIH funding is being targeted. “We could look at how their behavioral patterns link to important health outcomes.”

Then, the researchers expect to expand the 24-hour window to additional days over a year to see if the benefits of eating and exercising at certain times remain true.

“There’s environmental and personal data we can also tie into this,” Eicher-Miller said. “The more we include, the more we see additional things we don’t know and that we want to try to pursue and push farther.”

Richards added, “We do have access to a data set now that has people’s data at least at two time points. We’d like to see if the patterns we find hold over the long-term. Seeing if the patterns hold longitudinally will really strengthen the claims that we are able to make.”

But before you move your workout to later in the day or pass on that heavy Hoosier breakfast, the researchers stated more work must be done before they can professionally recommend behavioral changes in adults’ diet and physical activity habits.

“This kind of study, since it’s just cross-sectional, it’s not something we want to give advice for someone, for an individual,” Eicher-Miller said. “We still need more information, and we will try to build on it. If we really wanted to have this as solid, evidence-based guidance, then we need to have some clinical trials where we actually assign people randomly to go on certain time diet patterns and then find the same patterns linked with those outcomes. It’s still in an investigatory stage.”