nabu.science
MethodologyAboutContact

 

  • Snapshot
  • Key Claims
  • Strengths & Limitations
  • Commentary
  • From the Authors
  • Author Note / Field Feedback
  • Dimension breakdown
  • Impact Profile
Free · About 5 minutes
Report #NBU-1400F021·Framework v5.9·Pipeline v0.4·Evaluated 08 Oct 2026·OECD: 3.2 Clinical medicine·Subfield: Critical Care and Intensive Care Medicine·Basis: Full text·Type: Observational cohort study

CSF neurofilament light chain concentration in patients with delirium following hip fracture: a multicenter prospective study.

Titlestad et al.Titlestad, Watne, Blennow et al.·2026·BMC geriatrics·10.1186/s12877-026-07630-4 ↗

Primary claim

Our findings show that CSF-NfL concentrations were associated with delirium in patients without pre-existing dementia, suggesting possible undiagnosed dementia or, less likely, delirium-related neuronal injury.

Supported · an adjusted observational subgroup signal remains vulnerable to residual confounding from undetected dementia
In brief

A large prospective study clarifies that CSF-NfL associations largely attenuate after adjustment, but residual dementia misclassification limits the subgroup inference.

ReliabilityCaveats4 of 4 checks completeRead the 3 entries ↓
Fragile — the claim's weakest requirement is that the effect exists at all. Experimental. Every point below is traceable to a quoted passage, but this forecast has not yet been validated against known replication outcomes.
Abstract

BACKGROUND: Studies suggest delirium is associated with neuronal injury, which may further raise mortality risk. Neuronal injury can be assessed by measuring neurofilament light chain (NFL) concentrations in cerebrospinal fluid (CSF). This study aimed to investigate the association of CSF-NfL with delirium and mortality in patients with hip fracture.

METHODS: The study comprised two prospective cohorts of 548 hip fracture patients with per-operative CSF samples. CSF-NfL concentrations were measured using a commercial ELISA. Delirium was assessed daily from admission until the fifth postoperative day and survival censored at one year. The multivariable analyses (Logistic and Cox regression) were adjusted for age, sex, glomerular filtration rate, dementia status, comorbidity, and Activities of Daily Living. Additionally, Cox regression was adjusted for delirium.

RESULTS: In total, 259 (52%) patients developed delirium. In univariate analysis, CSF-NfL was higher among patients with delirium (2116 pg/ml versus 1366 pg/ml, odds ratio (OR) 2.21, (95% confidence interval (CI) [1.75,2,78], P < 0.001). In adjusted analysis, CSF-NfL was not significant (OR 1.29, [0.92,1.81], P = 0.128) and only remained significantly associated with delirium in the subgroup of patients without dementia (OR 1.84, [1.17, 2.89], P = 0.007). In unadjusted analysis of mortality, CSF-NfL was significantly associated with death at one year (hazard ratio (HR) 1.60, [1.37, 1.87], P < 0.001) but not in adjusted analysis (HR 1.03 [0.84, 1.26], P = 0.736).

CONCLUSIONS: Our findings show that CSF-NfL concentrations were associated with delirium in patients without pre-existing dementia, suggesting possible undiagnosed dementia or, less likely, delirium-related neuronal injury. The CSF-NfL-associated mortality hazard was non-significant after adjustment, mainly for delirium. Thus, the clinical context must be considered when studying CSF-NfL and delirium.

Key Claims

Primary finding · Supported

Our findings show that CSF-NfL concentrations were associated with delirium in patients without pre-existing dementia, suggesting possible undiagnosed dementia or, less likely, delirium-related neuronal injury.

Fragile — the claim's weakest requirement is that the effect exists at all. Experimental. Every point below is traceable to a quoted passage, but this forecast has not yet been validated against known replication outcomes.

an adjusted observational subgroup signal remains vulnerable to residual confounding from undetected dementia

Abstract, Conclusions
Quote verified
Secondary · Supported

In adjusted analysis, CSF-NfL was not significant (OR 1.29, [0.92,1.81], P = 0.128).

the reported adjusted estimate crosses the null, though its interval permits a modest association

Abstract, Results; Table 2
Quote verified
Secondary · Supported

CSF-NfL was not independently associated with one-year mortality after hip fracture.

the adjusted Cox estimate is null, although covariate temporality and residual confounding limit interpretation

Discussion, Conclusion
Quote verified

Derived from the full evaluation — not a separate score.

Strengths · Limitations

Strengths

#contribution

Large prospective multicentre CSF cohort

The 548-patient study is substantially larger than the prior CSF-NfL studies it discusses and includes participants from five hospitals. This scale permits adjusted delirium, mortality, and dementia-stratified analyses.

↳ Methods, Study Participants; Introduction, prior CSF-NfL studies

#methodological rigour

Explicit multivariable confounding strategy

The analysis prespecifies age, sex, renal function, dementia, ASA status, ADL, and cohort, then reports crude and adjusted estimates and change-in-estimate analyses. A complete-case comparison is also reported.

↳ Methods, Statistics; Tables 2–3; Results, Impact of Clinical Factors

#positioning

Limitations linked to interpretation

The Discussion explains how imperfect IQCODE sensitivity could produce residual confounding in the no-dementia subgroup and how once-daily assessment could miss delirium episodes. It also explicitly declines causal inference from the observational design.

↳ Discussion, subgroup interpretation and limitations

Limitations

#methodological rigour

Selection mechanisms remain insufficiently analysed

Eligibility was limited to patients receiving spinal anaesthesia, and the first cohort excluded patients considered too frail to survive surgery. The implications of these restrictions and differing cohort protocols are not examined through participant flow or selection analysis.

↳ Methods, Study Participants

#methodological rigour

Subgroup inference depends on dementia classification

The positive association is confined to patients classified as not having dementia, yet IQCODE sensitivity is imperfect and the paper acknowledges possible undiagnosed or subclinical neurodegeneration. This makes residual confounding a plausible explanation for the subgroup signal.

↳ Methods, Clinical Data; Discussion, dementia-stratified interpretation

#reporting

Reporting inconsistencies require correction

The Figure 1 narrative swaps dementia-group sample sizes, Table 3 contradicts the statement that all clinical factors remained significant, and the Abstract ambiguously associates 259 cases with the 548-person cohort. These defects do not reverse the main adjusted estimates but reduce confidence in presentation accuracy.

↳ Abstract; Results, Figure 1 discussion; Table 1; Table 3

Does this match your read? Add your assessment ↓

Nabu Commentary

The large prospective sample and explicit multivariable strategy support the central finding that crude CSF-NfL associations attenuate after adjustment. Methodological Rigour is moderated by selection into spinal anaesthesia, differing cohort protocols, simple imputation, and limited analysis of how those features affect inference. The argument remains followable, but the swapped Figure 1 group labels and the Table 3 significance contradiction reduce clarity. The Discussion handles residual dementia misclassification and competing literature constructively, while direct clinical impact remains limited by invasive sampling and predominantly null adjusted results.

Evaluate another paper

Free · About 5 minutes · Research preview

Author Note / Field Feedback

Nabu’s assessment, alongside the field’s view.

Are you an author of this paper?

Dimension breakdown

Evidence Score

Out of 5
012345
3.6±0.2

Contribution
3.7

The study meaningfully enlarges the small CSF-NfL delirium evidence base and adds adjusted, dementia-stratified, and mortality analyses. Its advance remains bounded by predominantly null adjusted results, cohort overlap, and an exploratory subgroup association.

Introduction, prior CSF-NfL studies; Results, Tables 2–3 and dementia-stratified analysis

Overall, CSF-NfL was not associated with delirium or one-year mortality

Methodological Rigour
3.5

Prospective ascertainment, blinded laboratory analysis, an a priori covariate set, cohort adjustment, and complete-case comparison support competent execution. Selection restrictions, heterogeneous cohort protocols, simple imputation, and once-daily assessment introduce meaningful uncertainty that is not fully analysed.

Methods, Study Participants, Delirium Screening, Sampling, and Statistics

All patients who... underwent surgery in spinal anesthesia, were eligible for inclusion.

Reporting
3.7

The objectives, analyses, and principal adjusted results are generally easy to follow and mostly use associational language. Reporting is reduced by swapped dementia-group labels, a Table 3 narrative contradiction, ambiguous denominator framing, and an over-specific attribution of mortality attenuation to delirium.

Abstract; Results, Figure 1 discussion; Results, one-year mortality; Table 3

all clinical factors remained significant

Positioning
3.9

The Discussion compares the findings with relevant CSF- and blood-NfL studies and connects dementia misclassification, single sampling, and missed delirium episodes to interpretation. The treatment of cohort heterogeneity, assay differences, and the Fong comparison is less complete.

Discussion, paragraphs comparing prior delirium and mortality studies; Discussion, limitations

Residual confounding may also have been present in the group classified as not having dementia.

Reliability

Caveats4 of 4 checks


Results checkNoted
  • Our findings show that CSF-NfL concentrations were associated with delirium in patients without pre-existing dementia, suggesting possible undiagnosed dementia or, less likely, delirium-related neuronal injury: "patients with dementia without delirium (median 1930 pg/ml, IQR = 1722, n = 68 versus 1895 pg/ml, IQR = 1717, n = 191, respectively)" (Results, delirium according to dementia status) vs "Dementiag, h, n (%)30 (13)191 (74)" (Table 1, dementia row; columns No delirium and Delirium). Table 1 assigns 30 dementia patients to no delirium and 191 to delirium, whereas the subgroup narrative assigns these counts in reverse; correcting those counts does not change the separately reported adjusted non-dementia association. → Conclusion holds
  • CSF-NfL was not independently associated with one-year mortality after hip fracture: "non-dementia group HR 0.89, [0.56, 1.25] P = 0.372" (Results, one-year mortality according to dementia status). Assuming a log-Wald 95% interval, its geometric midpoint is sqrt(0.56 × 1.25) = 0.837 and its reconstructed standard error is ln(1.25/0.56)/3.92 = 0.205, which combined with HR 0.89 gives a two-sided approximate P = 0.569 rather than 0.372. → Conclusion holds
Reporting and transparencyNoted

The Abstract needs an explicit delirium-analysis denominator and a corrected confidence-interval decimal; the Methods give incompatible descriptions of ADL imputation.

Conflict of interest: not stated · Ethics approval: not stated · Data availability: not stated · Code availability: not stated

Reference IntegrityNoted

68 references in manuscript 66 of 66 checkable references found in an index 2 have no canonical index record — counted, but not index-checkable the reference list may be incompletely parsed; counts are indicative 1 reference confirmed by manual review

References not verified automatically are resolved manually within 24 hours.

Post-Publication RecordClean

No retraction or expression of concern found in Crossref (includes Retraction Watch).

Last checked: 17 Oct 2026

Sources: Crossref (incl. Retraction Watch) ✓

Impact Profile

Impact areaHealth · Neurological · Other neurological disorders
Type of advanceEmpirical
Route to useSuggested
AdvanceExtends
BeneficiariesCatch-all cause
CitationsToo early
Policy · patentNot yet traced
ReachedNot reachedNot assessable or unverified
ReadinessValidated
Foundational
Controlled
Piloted
Validated
Implementable
FoundationalImplementable

The association was tested in two independent prospective cohorts across multiple hospitals using real-world hip fracture patients with daily delirium assessment.

From the paper
“we can improve the interpretation of CSF-NfL data in clinical and research contexts”

The discussion names a possible real-world use of its own findings, namely improving interpretation of CSF-NfL data, but does not specify who would act on them.

Pathway

Researchers and clinicians studying delirium would use these confounder-adjusted findings to interpret NfL biomarker data and design future blood-based studies of neurodegeneration.

Barrier

CSF requires invasive lumbar puncture, and the association is largely explained by age and undiagnosed cognitive impairment, so it is not usable as a delirium biomarker.

Evaluated 13 Nov 2026 · GBD 2023 · glm_v1_3_3 · impact_profile_v1_3_3

Report an issue →

AI-generated, human-governed. Something look off? Contact us to request a review.

Keep reading

Related evaluations

Other papers Nabu has evaluated blind, starting with the same field.

  1. Apolipoprotein E Mimetic Peptide CN-105 and Postoperative Delirium in Older Patients: The Phase 2 MARBLE Randomized Clinical Trial

    Timko, Cooter Wright, Smith, et al. · 2026 · Basic medicine

    Primary claim

    CN-105 (vs placebo) administration was feasible and did not increase AEs.

    Supported
    Evidence Score
    3.7±0.2
    Reliability
    Caveats
  2. Blood-Brain Barrier Permeability in Cases of Post-operative Delirium Is Associated with Central Nervous System Phosphatidylcholine Imbalances

    Jung, Pan, Sweeney, et al. · 2026 · Clinical medicine

    Primary claim

    In persons prone to postoperative delirium, the equilibrium of certain species of PC across the BBB correlates with BBB permeability.

    Tentative
    Evidence Score
    2.8±0.2
    Reliability
    Caveats
  3. On-premise medical AI agents for reliable clinical decision-making

    2026 · Clinical medicine

    Primary claim

    Diagnostic behavioral consistency provided the strongest discrimination of correctness (area under the curve (AUC) = 0.860) and remained informative under stress testing (AUC = 0.875).

    Supported
    Evidence Score
    3.7±0.2
    Reliability
    Caveats

How the rubric works →