These conference abstracts from the 2026 Sleep meeting predominantly present early-stage biomarker and phenotypic analyses in isolated RBD (iRBD), with only one published review article included. Most findings are cross-sectional and lack longitudinal phenoconversion data—the critical missing piece needed to establish clinical utility. Administrative-code-only RBD ascertainment inflates misclassification risk in the largest study, while biomarker studies remain hypothesis-generating pending control group replication and outcome validation.
- Strengths: Very large national cohort (N=34,485) provides substantial statistical power for mortality comparisons; adjustment for key confounders (age, sex, race/ethnicity, BMI, CCI) reduces residual confounding partially; long observation window (1999–2020) enables meaningful mortality ascertainment.
- Weaknesses: ICD coding for RBD is known to be unreliable and lacks polysomnographic confirmation, substantially inflating misclassification risk and producing heterogeneous phenotypes; PTSD diagnosis via ICD code in a VA population introduces ascertainment bias; mutual exclusivity of groups does not preclude phenotypic overlap; cause of death is not reported, so mortality differences could reflect competing comorbidities rather than RBD-specific pathology; polysomnographic validation is absent, likely including many false positives, particularly in the PTSD group where nightmares can mimic RBD.
- Risk of bias: High — ICD-code-only case ascertainment without PSG confirmation substantially inflates misclassification risk, especially for the PTSD-associated group where nightmare disorder is a common confounder.
- Statistical adequacy: Appropriate — sample sizes are large, regression models are appropriately adjusted, and the primary comparisons are well-powered; however, the validity of the underlying phenotyping undermines the meaning of statistical precision.
- Directly supported: PTSD-associated and PD-associated RBD groups differ substantially in age distribution and comorbidity burden; adjusted mortality odds differ significantly by group in this administrative dataset.
- Inferential: That these groups represent biologically distinct RBD subtypes warranting stratified surveillance is plausible but not established by administrative coding alone.
- Overreach: The conclusion that findings "support stratified approaches to surveillance for synucleinopathy" implies clinical actionability that administrative code-based phenotyping cannot justify without PSG-confirmed cohort replication.
- Strengths: Prospective, multicenter design with blinded pathological assessment reduces single-site bias and observer bias; PSG-confirmed iRBD diagnosis in the majority of subjects is a major methodological advantage over administrative-code studies; anatomically targeted multi-site biopsy protocol enables regional deposition mapping.
- Weaknesses: Sample size of 80 (with only 49 at 12-month interim) is small for a biomarker validation study and underpowered to detect longitudinal change or subgroup differences; the 12-month change in composite score did not reach statistical significance, yet the authors describe a "36% increase"—this selective emphasis on relative change is potentially misleading; no healthy control or disease-control arm limits sensitivity and specificity estimation; validated questionnaire was accepted as PSG equivalent for some subjects, introducing diagnostic heterogeneity; phenoconversion endpoint—the study's most clinically important question—is not yet answered; current results are purely cross-sectional biomarker data.
- Risk of bias: Moderate — blinded assessment and multicenter design reduce performance and detection bias, but small interim sample, lack of controls, and incomplete longitudinal data introduce substantial imprecision.
- Statistical adequacy: Underpowered — the 12-month longitudinal comparison in 49 patients was not significant (p>0.05) and the study is clearly underpowered to answer its second and third aims at this interim stage.
- Directly supported: Cutaneous P-SYN can be detected in 75% of PSG/questionnaire-confirmed iRBD patients at baseline; the biopsy procedure appears safe.
- Inferential: A trend toward increasing P-SYN deposition over 12 months is biologically plausible given the progressive nature of prodromal synucleinopathy.
- Overreach: The authors describe skin biopsy as "a safe and highly sensitive method"—sensitivity cannot be established without a specificity-defining control group; additionally, claiming a "36% increase" in P-SYN as a meaningful finding when the change is non-significant (p>0.05) overclaims the data.
- Strengths: Sleep EEG as a non-invasive measure of neurological state is a conceptually sound research direction in PD.
- Weaknesses: No abstract, methods, results, or sample size information is available—this paper cannot be critically appraised; verified status is false, requiring heightened skepticism; without any verifiable content, no factual claims can be assessed; the concept of "neurological aging" as an endpoint is itself poorly operationalized without knowing the study's specific definitions and measurements.
- Risk of bias: High — paper is unverified (verified: false) and no abstract or methods are available; risk of bias cannot be assessed and the paper's existence as a peer-reviewed publication cannot be confirmed.
- Statistical adequacy: Underpowered — cannot be assessed; no data provided.
- Directly supported: Nothing can be confirmed as supported; no data are available.
- Inferential: The concept that sleep EEG may reflect neurodegeneration in PD is supported by existing literature external to this paper, but this paper's specific claims cannot be evaluated.
- Overreach: Cannot be determined without data; however, the framing of sleep EEG as a quantifier of "neurological aging" risks conceptual overreach if not precisely defined.
- Strengths: Large, well-characterized iRBD cohort from a multicenter consortium with multimodal data (neuropsychology, DAT imaging, MRI) provides breadth for exploratory analysis; Bonferroni correction for multiple comparisons limits type I error inflation across the many outcome tests; bedpartner-based mismatch is a novel and clinically accessible marker that requires no additional testing.
- Weaknesses: Cross-sectional design precludes causal inference; mismatch direction could be a consequence rather than a predictor of cognitive status; SCOPA-Sleep was designed for Parkinson's disease, not specifically iRBD, and its reliability and validity in this prodromal population have not been independently established; bedpartner availability introduces selection bias—patients without bedpartners are excluded entirely; the DAT and apathy null findings receive little interpretive attention, risking reporting bias; cortical thickness effects are in small, specific auditory/language regions and may not survive correction for total number of cortical regions tested.
- Risk of bias: Moderate — large consortium cohort and Bonferroni correction are strengths, but cross-sectional design, bedpartner selection bias, and multiple testing across neuroimaging regions without fully transparent region-of-interest selection raise concern.
- Statistical adequacy: Appropriate for primary cognitive comparisons given sample size; cortical thickness regional analyses may be underpowered for the number of regions implicitly tested and the small effect sizes in brain structure outcomes.
- Directly supported: Night-time and daytime underestimators of sleep problems score lower on validated cognitive measures and show reduced cortical thickness in auditory-language regions compared to overestimators in this cross-sectional analysis.
- Inferential: That mismatch may reflect anosognosia or impaired interoception linked to auditory cortex thinning is biologically plausible but speculative without longitudinal or mechanistic data.
- Overreach: The conclusion that PT-BP mismatch "can be used as a clinical marker" implies prospective predictive validity and clinical actionability that a retrospective cross-sectional analysis cannot establish.
- Strengths: Video-PSG confirmation for all RBD diagnoses is the gold standard and a clear methodological strength; exclusion of established synucleinopathies (PD, DLB, MSA) removes the most obvious confounders for the non-synucleinopathy RBD question.
- Weaknesses: Total n=84 with only 12 tauopathy-associated patients is severely underpowered; subgroup comparisons involving n=12 have very low statistical power and high type I and type II error rates; single-center retrospective design with hospitalized patients introduces major selection bias; no specific effect sizes, p-values, or confidence intervals are reported in the available abstract, making quantitative evaluation impossible; antidepressant-associated RBD is classified as a distinct group, but antidepressants are commonly used in PTSD, depression comorbid with synucleinopathies, and tauopathies—group boundaries may be biologically artificial; no longitudinal follow-up.
- Risk of bias: High — single-center hospitalized sample with severely unequal and small subgroups, retrospective design, and no reporting of effect sizes or confidence intervals prevents meaningful quantitative interpretation.
- Statistical adequacy: Underpowered — comparisons involving the tauopathy group (n=12) are critically underpowered; non-parametric tests can detect only large differences at this sample size.
- Directly supported: Tauopathy-associated RBD appears to have higher rates of overt cognitive impairment and cerebral atrophy than iRBD in this highly selected hospitalized sample.
- Inferential: That these three subtypes represent biologically distinct entities with different pathomechanisms is plausible and consistent with emerging RBD nosology literature.
- Overreach: The claim that findings have "important implications for diagnosis, risk assessment, and clinical management" is premature given n=12 in the critical tauopathy subgroup and the absence of quantitative effect size reporting.
- Strengths: Use of PPMI—a well-characterized, prospective, multisite dataset with standardized assessment protocols—provides higher data quality than most single-center studies; FDR correction for multiple comparisons is appropriate for exploratory multi-endpoint analyses; MDS-UPDRS Part III is a validated, standardized motor outcome measure.
- Weaknesses: Cross-sectional design prevents determination of whether lymphopenia precedes, accompanies, or results from the neurodegenerative process; NLR is a non-specific inflammatory marker affected by infection, medications, metabolic syndrome, sleep disruption itself, and many other conditions; the motor association effect size is not reported in the abstract; 116 iRBD participants is a modest sample for a biomarker association study; PPMI iRBD participants may not be representative of community iRBD patients given referral and enrollment biases.
- Risk of bias: Moderate — PPMI data quality is good, but cross-sectional design, non-specific biomarker, and potential uncontrolled confounders (medications, metabolic factors) create meaningful residual confounding.
- Statistical adequacy: Appropriate for the primary group comparison given n=400 total; FDR correction is properly applied, though the within-iRBD motor association analysis with n=116 is modestly powered for multiple neuropsychological endpoints.
- Directly supported: Lymphocyte counts are lower and NLR is higher in PPMI-enrolled iRBD participants versus age-matched controls; within iRBD, higher NLR is associated with worse motor scores after FDR correction.
- Inferential: That this reflects prodromal immune dysregulation related to synucleinopathy pathophysiology is biologically plausible given emerging evidence of neuroinflammation in PD, but alternative explanations (reverse causation, confounding) are not excluded.
- Overreach: The conclusion that "immune dysregulation characterized by elevated NLR is detectable in iRBD and is related to subtle motor dysfunction before the emergence of overt PD" implies a mechanistic and temporal relationship that cross-sectional data cannot establish.
1. Jones M et al. (VA RBD cohort) — Despite ICD-code limitations, this is the largest study in the set (N=34,485) with a hard mortality endpoint and nationally representative VA data, providing population-level descriptive evidence of RBD clinical heterogeneity that motivates PSG-validated follow-up work.
2. Levine T et al. (Syn-Sleep Study) — The only prospective, blinded, multicenter biomarker study in this set with PSG-confirmed iRBD and a mechanistically grounded endpoint (cutaneous P-SYN); its completion with phenoconversion data would represent a genuine advance in prodromal biomarker science.
3. Ha J et al. (NAPS Consortium PT-BP mismatch) — The NAPS Consortium PT-BP mismatch paper uses the best-characterized iRBD cohort in the set with multimodal outcomes and a clinically accessible observational variable; the cognitive and cortical thickness associations are the most scientifically concrete positive findings reported across these papers, pending longitudinal validation.
These six papers collectively address RBD as a heterogeneous prodromal syndrome from administrative epidemiology, peripheral and cutaneous biomarkers, neuroimaging, and sleep-behavior phenotyping perspectives. A consistent theme is phenotypic diversity within the RBD umbrella: comorbid PTSD, tauopathy, antidepressant exposure, and Parkinson's disease each appear to confer distinct clinical signatures, though none of the studies has sufficient rigor to translate this heterogeneity into stratified clinical pathways. The biomarker papers (cutaneous P-SYN and NLR) target mechanistically distinct aspects of prodromal synucleinopathy—peripheral protein deposition and immune dysregulation, respectively—and both are cross-sectionally promising but critically lack longitudinal phenoconversion data that would establish predictive utility.
The NAPS Consortium PT-BP mismatch paper is the most methodologically grounded of the group in terms of cohort characterization, but its cross-sectional design limits causal interpretation of what may be a cognitively informative behavioral observation. A recurring and serious limitation across all papers is the absence of longitudinal phenoconversion outcomes: knowing that a biomarker or clinical feature is present in iRBD is insufficient without knowing whether it predicts which patients convert, to what synucleinopathy subtype, and on what timeline. The unverified EEG biomarker paper contributes nothing evaluable to this synthesis.
Our research team was unable to independently verify this citation: Lanir-Azaria S, "[Multidimensional Sleep EEG Biomarkers and Neurological Aging in Parkinson's Disease]," SLEEPJ, 2026-05-15. Treat claims based on it with appropriate skepticism.
All other cited sources