Averaging sleep spindle occurrence in dogs predicts learning performance better than single measures

Area of Science:

• Neuroscience
• Comparative Cognition
• Sleep Science

Background:

• A positive association between sleep spindle occurrence and learning success is documented in humans and other species.
• However, the test-retest reliability of this effect is debated, with some large studies failing to confirm it.
• Discrepancies in automatic sleep spindle detection methods contribute to inconsistent findings.

Purpose of the Study:

• To investigate the relationship between sleep spindle density and learning success in dogs.
• To determine if using consistent detection methods across different tasks improves the reliability of observed associations.
• To explore whether trait density, an averaged measure, is a more robust predictor of cognitive performance.

Main Methods:

• Utilized a consistent automatic sleep spindle detection method across various learning tasks in dogs.
• Measured sleep spindle density (spindles per minute) and learning success (post-sleep recall).
• Averaged measurements of spindle density and learning success to reduce potential measurement error.

Main Results:

• A positive association between sleep spindle density and learning success was observed in dogs when using a uniform detection method.
• Averaging across multiple measurements of density and learning enhanced the visibility and reliability of this association.
• Trait density, estimated through averaged occurrence, emerged as a more dependable predictor of cognitive performance compared to single-measure estimates.

Conclusions:

• Consistent methodology in sleep spindle detection is crucial for reliably observing its link to learning success.
• Averaging sleep spindle density and learning success measures can mitigate measurement error and strengthen the observed effect.
• Trait sleep spindle density appears to be a more stable and reliable indicator of cognitive performance in dogs than transient measures.