Schultes · PLoS medicine 2007 · Controlled experimental crossover physiology study · n=32

Defective awakening response to nocturnal hypoglycemia in patients with type 1 diabetes mellitus.

Cited 100 times in the scientific literature.

Level 3 - non-randomized controlled study

Controlled experimental physiology study comparing two cohorts (T1DM vs. healthy controls) across two test conditions.

PubMed 17326710 · doi:10.1371/journal.pmed.0040069 · record verified 2026-08-28

What was done

In a controlled physiological study, 16 patients with type 1 diabetes mellitus (T1DM) and 16 healthy control participants matched for gender, age, and BMI were evaluated over two separate nights. During the experimental night, insulin was infused over 1 hour once polysomnography confirmed stage 2 sleep, producing a linear fall in plasma glucose to a nadir of 2.2 mmol/l. During the control night, euglycemia was maintained. Polysomnography was used to detect awakening, and plasma epinephrine was measured to evaluate hormonal counterregulation.

What was found

During the hypoglycemia condition, only 1 of 16 T1DM patients awakened compared to 10 of 16 healthy controls (p = 0.001). During euglycemic control nights, 0 participants in either group awakened during the corresponding window. Plasma epinephrine increased by at least 100% in all participants who awakened, as well as in 5 participants who did not awaken (3 T1DM, 2 controls; p < 0.001). In all participants who woke up, the epinephrine rise preceded polysomnographic awakening by a mean (± SEM) of 7.5 ± 1.6 minutes.

Why it matters

The findings show that normal awakening triggered by acute nocturnal hypoglycemia is severely impaired in T1DM, identifying a failure in central nervous system activation that puts patients at heightened risk for prolonged, undetected, and potentially fatal hypoglycemia during sleep.

Limits

The sample size is small (n = 16 per group). The study evaluated an acute, experimentally induced 1-hour drop in glucose initiated specifically during stage 2 sleep, which may not capture spontaneous real-world nocturnal glycemic trajectories or hypoglycemia arising from slow-wave or REM sleep. Clinical histories, duration of diabetes, and prior hypoglycemia awareness status were not detailed in the abstract.

Cited by