Study Reveals Negative Feedback Can Impair Learning in Patients With Sleep Disorders


A recent study on narcolepsy type 1 and central disorders of hypersomnolence showed that patients exhibited decreased learning abilities in response to negative feedback, suggesting a link between altered decision-making and vigilance disruptions.

Mélanie Strauss, MD, PhD, hospital professor at Hôpital Universitaire de Bruxelles, in Belgium

Mélanie Strauss, MD, PhD

In a newly published study in Sleep Medicine, patients with narcolepsy type 1 (NT1) and other central disorders of hypersomnolence (CDH) showed a decrease in learning abilities when receiving negative feedback compared with positive feedback in a reinforcement learning task. These findings suggest observed alterations in decision-making abilities may be associated with vigilance disruptions and emphasize the importance of restoring adequate vigilance to potentially minimize cognitive impairments in patients with CDH.1

Among participants with NT1 (n = 27), other CDH (n = 34), and healthy participants (n = 34), results showed normal and similar positive reinforcement learning. Authors noted that this pattern was not suggestive of dopamine deficiency, however, both participants with NT1 and other CDH had decreased learning abilities to avoid losses in the reinforcement learning task. Thus, the investigators observed that the decreased negative reinforcement learning in participants with CDH was associated with the alteration of vigilance.

Top Clinical Takeaways

  • The altered reinforcement learning profile extends beyond narcolepsy type 1, encompassing other central disorders of hypersomnolence, emphasizing a shared cognitive challenge.
  • Cognitive deficits and impaired decision-making observed in narcolepsy and CDH highlight the importance of monitoring alertness in assessing cognitive functions in these populations.
  • The study calls for further research to explore the potential benefits of stimulant medication in patients with sleep disorders, potentially addressing cognitive impairments.

“The results of this study show that the altered reinforcement learning profile found in participants with NT1 is not specific to this population but is shared by other CDH. It is then rather unlikely that hypocretin deficiency may play a unique or specific role in this behavior,” lead author Mélanie Strauss, MD, PhD, hospital professor at Hôpital Universitaire de Bruxelles, in Belgium, and colleagues wrote.1

READ MORE: Daytime Continuous Polysomnography Provides Reliable Diagnosis of Pediatric Narcolepsy Type 1

In this study, investigators recruited participants from the Hotel Dieu Hospital and the Pitié-Salpêtrière Hospital in Paris, France, and healthy participants. The participants with NT1, other CDH, and healthy volunteers were matched for age and sex. In the 40-minute duration of the reinforcement learning task, participants were asked to choose between 2 abstract symbols presented on a computer screen to win as much money as possible.2 The researchers compared clinical characteristics and overall gain between groups using Kruskal-Wallis ANOVAs (KW-ANOVA) for continuous variables, and 2-by-2 comparisons using Wilcoxon ranksum (unpaired) tests and had a comparison of proportions using a Pearson's Chi-squared test.

In both NT1 and CDH groups, the learning progression was better in the gain than in the loss condition. This was observed in the post hoc Wilcoxon signed-rank tests (NT1: gain, 6.56 [0.27] v loss, 5.59 [0.19]; P = .0061; CDH: gain, 6.83 [0.261] v loss, 6.06 [0.18]; P = .0074,). Also, these findings were similar across conditions in healthy controls (gain, 6.28 [0.33] v loss, 6.32 [0.21]; P = .45). The difference between participants with NT1 and CDH compared with the controls was driven by a poorer learning performance in the loss condition in patients with NT1 and CDH as observed in the KW-ANOVA analysis (between groups in the loss condition, P = .031; in the gain condition, P = .49).

All told, limitations of the study included the lack of CSF hypocretin-1 measurement in the participants, noting it would be interesting to see how hypocretin levels influenced alertness and performance. Another limitation noted by the authors was the small sample size of participants with NT2 (n = 20) and particularly IH (n = 14), which prevented them from conducting a more in-depth analysis of their specific results.

“This study shows that learning from punishment is altered in individuals with CDH. This profile is not specific to individuals with NT1 and is associated with sleepiness. This study provides new insights into the nature of cognitive deficits and especially decision making seen in individuals with narcolepsy and CDH and emphasizes the importance of monitoring alertness when assessing cognitive functions in these populations. Further research is needed to investigate the potential benefits of stimulant medication for these individuals,” Strauss et al noted.1

1. Strauss M, Griffon L, Elbaz M, Arnulf I, Chennaoui M, Léger D. Altered reinforcement learning in Narcolepsy type I and other central disorders of hypersomnolence. Sleep Med. Published online November 4, 2023. doi:10.1016/j.sleep.2023.11.001
2. Pessiglione M, Seymour B, Flandin G, Dolan RJ, Frith CD. Dopamine-dependent prediction errors underpin reward-seeking behaviour in humans. Nature. 2006;442(7106):1042-1045. doi:10.1038/nature05051
Related Videos
Jessica Ailani, MD
Frederic Schaper, MD, PhD
Jaime Imitol, MD
Jason M. Davies, MD, PhD
Carolyn Bernstein, MD
Prashanth Rajarajan, MD, PhD
Mandy Alhajj, DO, James Dolbow, DO & Neel Fotedar, MD
Riley Bove, MD
© 2024 MJH Life Sciences

All rights reserved.