Researchers analyzed data from 23,708 U.S. adults over 30 years of age who had participated from 2005 to 2016 in the National Health and Nutrition Examination Survey. The survey collects 24-hour dietary recalls from each participant – detailing not just what, but when, all food was consumed.

Respondents were categorized according to their HbA1c level, a measure of glucose control, into four groups: nondiabetes, prediabetes, controlled diabetes and poorly controlled diabetes.

Among the whole survey sample, snacks accounted for between 19.5% and 22.4% of total energy intake – while contributing very little nutritional quality.

In descending order of proportion, snacks consisted of convenience foods high in carbohydrates and fats, sweets, alcoholic beverages, non-alcoholic drinks that include sugar-sweetened beverages, protein, milk and dairy, fruits, grains and, lagging far behind, vegetables.

Noting that capturing 24 hours of food consumption doesn’t necessary reflect how people usually eat, “it gives us a really good snapshot of a large number of people,” Taylor said. “And that can help us understand what’s going on, where nutritional gaps might be and the education we can provide.”

Finding that people with diabetes had healthier snacking habits was an indicator that dietary education is beneficial to people with the disease. But it’s information that just about everyone can use, Taylor said – and it’s about more than just cutting back on sugar and carbs. 

“We need to go from just less added sugar to healthier snacking patterns,” he said. “We’ve gotten to a point of demonizing individual foods, but we have to look at the total picture. Removing added sugars won’t automatically make the vitamin C, vitamin D, phosphorus and iron better. And if we take out refined grains, we lose nutrients that come with fortification.