Our first research paper is out. Here’s what we found.

For the past few years, more than 62,000 Levels members chose to share their CGM data, food and activity logs, and lab results for research. Today we’re publishing the first paper from that data, a preprint by physician-scientists Paul Ginart, MD, PhD, and Michael Loesche, MD, PhD, along with a new home for this work: Levels Research.

What makes this dataset unusual is who’s in it. Nearly 97% of people who shared labs were in the normal or prediabetes range, a group medical research rarely studies because it mostly looks at people who are already sick. If you were one of those members, thank you. Here’s what the researchers found. 

Five Key Takeaways

1. Your average glucose is the number to watch.

Of everything a CGM measures, it was most closely tied to insulin resistance: in the top quarter of average glucose, 36 in 100 people crossed a common insulin-resistance cutoff, compared with 8 in 100 in the bottom quarter. It was also the steadiest number from week to week, and about a week of CGM wear gave a reliable read.

2. Judge a glucose spike by how long it lasts, not how high it goes.

Among people with the same average glucose, taller or more frequent spikes weren’t linked to more insulin resistance. Wider ones were: when glucose stayed up longer, fasting insulin, triglycerides, and inflammation tended to run higher.

3. A normal A1c isn’t the whole picture.

Nearly a quarter of people with a normal HbA1c had an average glucose of 110 mg/dL or higher, and in that group high fasting insulin was about twice as common. A CGM may be most useful for younger adults, since A1c tends to creep up with age even when glucose doesn’t.

4. Give a food about five tries before you judge it.

The same food produced noticeably different responses from one meal to the next, so any single meal is really a snapshot of that day. Averaging around five meals gave a consistent picture of your typical response.

5. Carb sensitivity is real, but it’s not the same as insulin resistance.

Some people’s glucose rose about seven times as much per gram of carbs as others’, and each person’s response was fairly consistent. But steep and gentle responders had about the same insulin levels, so a steep carb response on its own isn’t a red flag.

One important note: Every result here is an association measured at a single point in time. None shows cause and effect, and none tracks long-term health. The research page walks through what each finding means, what it doesn’t, and the questions this study can’t answer yet.

Looking to understand more about your metabolic health?
Explore our health news and advice site, Metabolic Insights

228 Park Ave South, PMB 63877, New York, NY 10003

Unsubscribe  |  Help Center  |  Terms of Service  |  Privacy Policy