A Systematic Review, Meta-Analysis and Meta-Regression on the Effects of Carbohydrates on Sleep.
Level 1 - systematic review of randomized trials
Systematic review and meta-analysis of intervention trials
PubMed 33919698 · doi:10.3390/nu13041283
What was done
Authors conducted a systematic review and random-effects meta-analysis and meta-regression across PubMed, SCOPUS, and Cochrane Library (through October 2020) to evaluate the effects of carbohydrate quantity, quality, and periodization on sleep. The meta-analysis included 11 articles representing 27 nutrition trials, yielding 16 comparison data sets (11 for sleep quantity, 5 for sleep quality).
What was found
Compared to high carbohydrate intake (HCI), low carbohydrate intake (LCI) increased N3 sleep duration (effect size [ES] = 0.37; 95% CI: 0.18 to 0.56; p < 0.001) and proportion (ES = 0.51; 95% CI: 0.33 to 0.69; p < 0.001). Conversely, HCI prolonged REM sleep duration (ES = -0.38; 95% CI: 0.05 to -8.05; p < 0.001) and proportion (ES = -0.46; 95% CI: -0.83 to -0.01; p < 0.001) compared to LCI. Carbohydrate quality did not affect sleep stages. In meta-regression, changes in carbohydrate percentage of daily energy intake significantly explained variance in sleep onset latency (R² = 25.87, p = 0.018), while glycemic load explained variance in sleep onset latency (R² = 50.8, p = 0.048), sleep efficiency (R² = 89.2, p < 0.001), and wake after sleep onset (R² = 64.9, p = 0.018).
Why it matters
This meta-analysis clarifies that carbohydrate quantity modifies sleep architecture by shifting the balance between deep slow-wave (N3) and REM sleep, whereas carbohydrate quality primarily impacts sleep initiation and continuity metrics.
Limits
The analysis is restricted to 11 articles, and the abstract does not report individual participant counts, demographics, or dietary adherence. Long-term interventions were not evaluated, limiting findings to short-term carbohydrate alterations.
Cited by
- supports Scientific research shows that dietary intake of starches improves sleep.