Unraveling the associations and causalities between glucose metabolism and multiple sleep traits

Minhan Yi, Quanming Fei, Ziliang Chen, Wangcheng Zhao, Kun Liu, Shijie Jian, Bin Liu, He Meng, Xiaoli Su, Yuan ZhangFrontiers in Endocrinology 2023 · DOI

This study, published in Frontiers in Endocrinology, provides a comprehensive look at the "chicken-or-the-egg" dilemma between sleep disorders and metabolic health. By combining a massive meta-analysis of over 840,000 participants with Mendelian Randomization, the researchers demonstrate that the link between poor sleep and dysregulated glucose is not just correlative, but bidirectional and causal.

TL;DR

Researchers have confirmed a causal feedback loop where high post-meal glucose levels drive insomnia and daytime sleepiness, while short sleep duration directly worsens long-term blood sugar levels (HbA1c). This study moves beyond simple association, using genetic data to prove that treating one condition likely requires managing the other.

What Problem Does This Paper Address?

It has long been observed that patients with Type 2 Diabetes (T2DM) suffer from poor sleep, and conversely, that chronic "short sleepers" are at higher risk for metabolic syndrome. However, observational studies are notoriously plagued by confounding variables—like diet, sedentary lifestyle, or socioeconomic status—making it difficult to determine if sleep issues cause metabolic problems or vice versa.

The authors sought to bridge this gap by clarifying the causal directions between four glycemic traits (fasting glucose, insulin, HbA1c, and 2-hour post-meal glucose) and three sleep traits (insomnia, excessive daytime sleepiness, and sleep duration).

Methods

The study employed a dual-pronged approach:

  1. Quantitative Meta-Analysis: A systematic review of 116 existing studies to establish broad associations.
  2. Two-Sample Mendelian Randomization (MR): This is the paper's core strength. MR uses genetic variants (Single Nucleotide Polymorphisms, or SNPs) associated with a trait as "instrumental variables." Because these variants are randomly assigned at birth, they act as a natural randomized controlled trial, bypassing the confounding factors found in traditional observational research.

The primary MR method used was Inverse Variance Weighting (IVW), supplemented by sensitivity analyses like MR-Egger and Weighted Median to ensure the results weren't being skewed by "pleiotropy" (where one gene affects multiple unrelated traits).

Datasets

  • Meta-analysis: Pooled data from 116 studies involving more than 840,000 participants.
  • Glycemic GWAS: Summary statistics from a meta-analysis of 281,416 non-diabetic participants across various ancestries.
  • Sleep GWAS: Data from the UK Biobank, covering approximately 450,000 individuals of European ancestry for traits like insomnia, daytime sleepiness (EDS), and sleep duration.

Key Insights and Ideas

  • The 2hGlu Factor: Interestingly, 2-hour post-challenge glucose (2hGlu) showed a stronger causal link to insomnia and daytime sleepiness than fasting blood glucose. This suggests that how the body handles a "glucose spike" after eating is more disruptive to sleep architecture than baseline sugar levels.
  • The Vicious Cycle of Insulin: The study found that fasting insulin (FINS) is associated with short sleep duration. Reversely, short sleep duration was found to causally influence HbA1c levels. This suggests a self-reinforcing loop where high insulin leads to poor sleep, which then drives up long-term average blood sugar.
  • The "U-Shaped" Relationship: The meta-analysis confirmed that both very short (<7h) and very long (>9h) sleep durations are associated with abnormal glucose metabolism, though the causal drivers for "long sleeping" remain less clear than those for "short sleeping."

Key Results

  • Increased Risks: Participants with abnormal glucose metabolism had a 65% higher risk of insomnia (OR 1.65) and a 37% higher risk of excessive daytime sleepiness (OR 1.37).
  • Causal Impact of Glucose: MR analysis showed that every 1-unit increase in genetically predicted 2hGlu increased the risk of EDS by 2.2% and insomnia by 2.0%.
  • Sleep's Impact on HbA1c: Short sleep duration was causally linked to an increase in HbA1c levels (β = 0.131), providing a clear genetic link between sleep deprivation and worsening glycemic control.

Limitations and Future Work

  • Ancestry Bias: The GWAS datasets were predominantly from populations of European ancestry, which may limit how well these causal links apply to other ethnic groups.
  • Subjective Reporting: Most sleep data in the large-scale GWAS (like the UK Biobank) relied on self-reported questionnaires rather than objective measures like polysomnography or actigraphy.
  • Future Directions: The authors suggest that clinical strategies for T2DM should explicitly include sleep hygiene, and sleep clinics should consider routine metabolic screening. Future research should focus on whether pharmacological interventions for glucose (like GLP-1 agonists) directly improve sleep quality via these causal pathways.

Summary updated

Abstract

Our results suggested mutual risk and causal associations between the sleep traits and glycemic traits, shedding new light on clinical strategies for preventing sleep disorders and regulating glucose metabolism. Future studies targeting these associations may hold a promising prospect for public health.
BibTeX
@inproceedings{yi2023unraveling,
  author = {Minhan Yi and Quanming Fei and Ziliang Chen and Wangcheng Zhao and Kun Liu and Shijie Jian and Bin Liu and He Meng and Xiaoli Su and Yuan Zhang},
  title = {{Unraveling the associations and causalities between glucose metabolism and multiple sleep traits}},
  booktitle = {Frontiers in Endocrinology},
  year = {2023},
  doi = {10.3389/fendo.2023.1227372},
}
Figure