Reply to Fan <i>et al.</i> : Assessment of the Neuroprotective Effect of Continuous Positive Airway Pressure in Obstructive Sleep Apnea: Can Static Metrics Map Dynamic Pathology?
Bibliographic record
Abstract
From the Authors: We thank Fan and colleagues for their discerning comments on our recent randomized controlled trial (1). Obstructive sleep apnea (OSA) is associated with neurocognitive dysfunction, including memory impairment (2). Our study showed that continuous positive airway pressure (CPAP) could improve neuroimaging biomarkers and emphasizes the importance of early intervention (1). We agree with Fan and colleagues that future prospective studies with larger sample sizes should be performed to validate the neuroprotective mechanisms of CPAP. With regard to some other issues that Fan and colleagues raised, we provide a response here. First, the intranetwork functional connectivity (FC) of default mode network (DMN) assessed by functional magnetic resonance imaging is one of the key secondary outcomes in our study. We found that there were significant differences in the FC of DMN between the CPAP group and the best supportive care group at 6 months after treatment. We acknowledged that positron emission tomography imaging is also an important tool in assessing brain function and activity. Combined assessment with positron emission tomography and functional magnetic resonance imaging might give more important information; this should be considered in future studies. Heterogeneity in response to CPAP treatment might exist among different phenotypic subtypes of OSA (3). In our study, we excluded subjects with obesity hypoventilation syndrome (see online supplement for (1); therefore, the factor (OSA with or without obesity hypoventilation syndrome) could not be analyzed. Second, we only got time of usage and residual apnea–hypopnea index from the CPAP machine, whereas the nocturnal oxygen saturation fluctuations or sleep efficiency could not be obtained. Future studies may benefit from using pulse oximeters to dynamically record oxygen saturation during CPAP treatment. We agree that the potential interference of comorbidities and pharmacological interventions should not be ignored; therefore, we performed additional subgroup analyses by comorbidities (hypertension, coronary artery disease, diabetes mellitus, and hyperlipidemia). There were no statistically significant differences in the key secondary outcomes (FC of DMN and cortical thickness) between the CPAP + BSC group and the BSC group, and no significant interactions between the comorbidities and the interventions were observed (for interactions, all P > 0.05) (Figures 1 and 2). Effects of CPAP treatment on functional connectivity (FC) of default mode network (DMN) by comorbidity subgroups at 6 months. We used linear mixed models for repeated measures of the DMN to assess between-groups difference at 6 months after enrollment. Outcome analyses are reported as least-squares means and 95% CIs, including the mean differences between groups. BSC = best supportive care; CIs = confidence intervals; CPAP = continuous positive airway pressure. Effects of CPAP treatment on cortical thickness by comorbidity subgroups at 6 months. We used linear mixed models for repeated measures of the cortical thickness to assess between-groups difference at 6 months after enrollment. Outcome analyses are reported as least-squares means and 95% CIs, including the mean differences between groups. BSC = best supportive care; CIs = confidence intervals; CPAP = continuous positive airway pressure. Finally, we agree with Fan and colleagues that the Montreal Cognitive Assessment is not sensitive for detecting subclinical impairment, and we have stated this in our study (1). We completely agree that computerized cognitive tests and neuroinflammatory markers are important for future randomized controlled trials, which will enable us to elucidate potential neuroprotective mechanisms of CPAP treatment. Author Contributions: All authors contributed to the writing and review of the manuscript and approved the final copy of the manuscript. Artificial Intelligence Disclaimer: No artificial intelligence tools were used in writing this manuscript. Originally Published in Press as DOI: 10.1164/rccm.202505-1243LE on July 30, 2025 Author disclosures are available with the text of this letter.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.011 | 0.081 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.002 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.003 | 0.003 |
| Scholarly communication | 0.004 | 0.005 |
| Open science | 0.004 | 0.002 |
| Research integrity | 0.030 | 0.043 |
| Insufficient payload (model declined to judge) | 0.006 | 0.006 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".