CGM Accuracy Drops During Exercise in Para Cyclists

What Was Studied and Why It Matters

Continuous glucose monitors (CGMs) are small sensors worn on the body that track blood sugar levels around the clock — no finger pricks required. They're increasingly used not just by people with diabetes, but also by athletes who want to fine-tune their nutrition and energy management.

But here's an important question: how accurate are these devices when your body is working hard? Researchers in Europe set out to answer this by studying 13 Para cyclists — athletes with physical disabilities who compete in adaptive cycling events, including hand bikers and traditional cyclists.

What the Researchers Did

Each athlete wore an Abbott CGM sensor for two weeks during their normal training routine. They logged their meals and workouts, and completed one standardized training session in a controlled setting. Researchers compared CGM readings against traditional finger-prick blood sugar tests — taken at rest and during exercise — to measure how accurate the CGM actually was.

What They Found

Blood Sugar Levels Were Generally Healthy

Good news first: these Para cyclists showed healthy blood sugar patterns overall. Their average blood sugar over 24 hours was a normal 5.7 mmol/L, and they spent 91% of the time in a healthy blood sugar range. Spikes above the healthy range happened mainly right after meals and during exercise — both expected occurrences.

Nighttime Lows Are a Real Concern for Some Athletes

Athletes with spinal cord injuries were more likely to experience low blood sugar at night (called nocturnal hypoglycemia). This is a meaningful safety concern, since low blood sugar during sleep can go unnoticed and become dangerous.

CGM Accuracy Fell Sharply During Exercise

This is the study's most important finding for anyone using a CGM during physical activity. At rest, the CGM was reasonably accurate — off by about 12% on average. But during exercise, that error jumped to 34%. For hand bikers — whose arm muscles are closest to where the sensor is typically worn — the error reached a striking 41%.

• CGM accuracy at rest: ~12% error margin (acceptable)
• CGM accuracy during exercise: ~34% error margin (significantly reduced)
• Hand bikers during exercise: ~41% error margin (markedly unreliable)

What This Means for Real Patients and Athletes

If you use a CGM during workouts — whether you have diabetes or are using it for performance nutrition — don't rely solely on CGM readings when you're actively exercising. The sensor's placement near working muscles can distort readings. A traditional finger-prick test remains the gold standard when accuracy is critical, such as before adjusting insulin or making decisions about food intake mid-workout.

For those managing diabetes alongside an active lifestyle, having reliable supplies on hand matters. MDS Diabetes carries a range of CGM accessories, lancets, and glucose testing supplies to support both everyday monitoring and active routines.

Athletes with spinal cord injuries should speak with their care team about monitoring strategies specifically during nighttime hours.

Bottom Line

CGM sensors are a valuable tool for athletes, but their accuracy drops significantly during exercise — especially when worn near highly active muscles. Para cyclists with spinal cord injuries face an added risk of nighttime low blood sugar. Always confirm CGM readings with a finger-prick test before making important diabetes management decisions during physical activity.