Arnoriaga-Rodríguez · The Journal of clinical endocrinology and metabolism 2023 · cross-sectional case-control study · n=118

Gut Microbiota Composition and Functionality Are Associated With REM Sleep Duration and Continuous Glucose Levels.

Cited 18 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional case-control observational study

PubMed 37159524 · doi:10.1210/clinem/dgad258 · record verified 2026-08-30

What was done

This cross-sectional case-control study evaluated 118 middle-aged healthy volunteers (39.1-54.8 years; 60 with obesity) recruited from a tertiary hospital. Participants underwent 10 days of continuous glucose monitoring (Dexcom G6) to evaluate glucose variability (SD, CV, IQR) and time in ranges (% TIR, % TIR2 at 126-139 mg/dL, and % TIR3 at 140-199 mg/dL). REM sleep duration was tracked over the same 10 days using wrist actigraphy (Fitbit Charge 3). Gut microbiota composition and functionality were analyzed using shotgun metagenomic sequencing.

What was found

Participants with obesity demonstrated increased glycemic variability alongside higher % TIR2 and % TIR3. REM sleep duration was independently and negatively associated with % TIR3 (β = -.339; P < .001) and glucose variability standard deviation (β = -.350; P < .001). Metagenomic profiling showed that taxa from the Christensenellaceae family were positively associated with REM sleep and negatively associated with continuous glucose monitor levels, whereas taxa from the Enterobacteriaceae family and bacterial pathways involved in iron metabolism showed the opposite associations.

Why it matters

These findings suggest that reduced REM sleep duration and higher glucose variability covary with specific taxonomic and functional gut microbiota signatures in adults with and without obesity.

Limits

The cross-sectional design prevents drawing causal conclusions about whether gut dysbiosis, poor REM sleep, or glucose dysregulation drive the observed associations. REM sleep was estimated via consumer-grade wrist actigraphy rather than formal polysomnography. The sample size was modest (n = 118) and drawn from a single tertiary hospital.

Cited by