# Pathways to resilience: relationships between cognitive reserve, psychological debt, and Alzheimer's disease biomarkers.

- **DOI:** [10.1186/s13195-026-02054-z](https://doi.org/10.1186/s13195-026-02054-z)
- **Rating:** Sound (3.2 out of 5)
- **Verdict:** An integrated SEM confirms reserve–cognition associations, but cross-sectional identification and marginal model fit limit mechanistic interpretation.
- **Framework version:** v5.9
- **Pipeline version:** v0.4
- **Evaluation date:** 2026-10-08
- **Evaluation page:** https://www.nabu.science/paper/10.1186%2Fs13195-026-02054-z

## Key claims

1. higher cognitive reserve was associated with better cognitive function (p = .005), independent of AD-related biomarkers
   - Support: Supported
   - Type: Primary finding
   - Basis: cross-sectional age-adjusted SEM supports the association, but cannot identify a resilience mechanism
   - Location: Abstract; Results, Structural Equation Modeling; Table 2
2. AD-related biomarkers of CSF Aβ42 (p = .021), splenial WMH (p = .044), and hippocampal neurodegeneration (p = .007) were each independently associated with lower cognitive function
   - Support: Supported
   - Type: Secondary
   - Basis: joint age-adjusted paths support direction, though several reported intervals and p-values are inconsistent
   - Location: Abstract; Results, Structural Equation Modeling; Table 2
3. higher cognitive reserve was correlated with lower psychological debt
   - Support: Supported
   - Type: Secondary
   - Basis: the reported covariance supports a correlational claim within the selected cross-sectional sample
   - Location: Results, Structural Equation Modeling; Table 2
4. neither construct showed a significant association with the AD-related biomarkers (p ≥ .177)
   - Support: Tentative
   - Type: Secondary
   - Basis: cross-sectional null paths lack equivalence testing or a reported power analysis
   - Location: Abstract; Results, Structural Equation Modeling; Table 2
5. The relationships between cognitive reserve and psychological debt, and between CSF Aβ42 and splenial WMH were stronger in APOE ε4 non-carriers than in carriers.
   - Support: Tentative
   - Type: Secondary
   - Basis: within-group significance and a global test do not establish both path-specific between-group differences
   - Location: Abstract; Results, APOE ε4 Status; Fig. 2

## Strengths

- **Integrated latent multi-biomarker model** The SEM jointly models cognitive reserve, psychological debt, three biomarkers, and PACC5-based cognitive function rather than analyzing each factor in isolation.
  - Dimension: contribution
  - Location: Methods, Structural Equation Modeling; Fig. 1
- **Transparent robustness and subgroup analyses** The paper reports robust estimation, an Aβ42/40 sensitivity analysis, an SCD/MCI analysis, and measurement and structural invariance procedures across sex and APOE ε4 groups.
  - Dimension: methodological rigour
  - Location: Methods, Sensitivity Analysis and Multigroup Analysis; Results
- **Limitations linked to interpretation** The discussion explicitly connects cross-sectional design to causal uncertainty and sub-threshold CFI/TLI to marginal model fit, while also noting self-report and biomarker-coverage limitations.
  - Dimension: positioning
  - Location: Strengths and Limitations

## Limitations

- **Reported statistics are not fully coherent** Several confidence intervals imply materially smaller p-values than reported, and the SCD/MCI sensitivity analysis repeats the full-sample coefficient and interval while giving a different p-value.
  - Dimension: methodological rigour
  - Location: Results, Structural Equation Modeling; Tables 2–3; Sensitivity Analyses
- **Limited identification of reserve pathway** The primary model is cross-sectional and adjusts regression paths only for age, leaving diagnosis, site, sex, and premorbid-ability confounding insufficiently addressed for mechanistic interpretation.
  - Dimension: methodological rigour
  - Location: Methods, Structural Equation Modeling; Strengths and Limitations
- **APOE moderation wording exceeds tests** The paper reports a global APOE likelihood-ratio test and significant covariances within non-carriers, but not path-specific between-group tests supporting the claim that those relationships were stronger.
  - Dimension: reporting
  - Location: Abstract; Results, APOE ε4 Status; Fig. 2

## Reliability checks

Caveats (4 of 4 checks).

- **Results check:** Noted
  - higher cognitive reserve was associated with better cognitive function (p = .005), independent of AD-related biomarkers: "Higher cognitive reserve was significantly associated with better cognitive function (β = 0.27, p = 0.005; 95% CI: 0.14, 0.41)." (Results, Structural equation modeling). Assuming a symmetric normal-Wald interval for the reported standardized coefficient, SE ≈ (0.41 − 0.14)/3.92 = 0.0689 gives z ≈ 3.92 and two-sided p ≈ 0.000089, rather than 0.005. → Conclusion holds
  - AD-related biomarkers of CSF Aβ42 (p = .021), splenial WMH (p = .044), and hippocampal neurodegeneration (p = .007) were each independently associated with lower cognitive function: "CSF Aβ42 (inverted; β = −0.18, p = 0.021; 95% CI: −0.30, −0.06)" (Results, Structural equation modeling). Assuming a symmetric normal-Wald interval for the reported standardized coefficient, SE ≈ (−0.06 − (−0.30))/3.92 = 0.0612 gives |z| ≈ 2.94 and two-sided p ≈ 0.0033, rather than 0.021. → Conclusion holds
  - AD-related biomarkers of CSF Aβ42 (p = .021), splenial WMH (p = .044), and hippocampal neurodegeneration (p = .007) were each independently associated with lower cognitive function: "hippocampal neurodegeneration (β = −0.28, p = 0.007; 95% CI: −0.41, −0.15)" (Results, Structural equation modeling). Assuming a symmetric normal-Wald interval for the reported standardized coefficient, SE ≈ (−0.15 − (−0.41))/3.92 = 0.0663 gives |z| ≈ 4.22 and two-sided p ≈ 0.000024, rather than 0.007. → Conclusion holds
  - higher cognitive reserve was associated with better cognitive function (p = .005), independent of AD-related biomarkers: "The direct association between cognitive reserve and cognitive function was maintained (β = 0.27, p = 0.039; 95% CI: 0.14, 0.41)." (Results, Sensitivity analyses, SCD/MCI analysis) vs "Higher cognitive reserve was significantly associated with better cognitive function (β = 0.27, p = 0.005; 95% CI: 0.14, 0.41)." (Results, Structural equation modeling, full sample). The two analyses print identical coefficients and confidence intervals despite different p-values, and a normal-Wald reconstruction from either rounded interval gives SE ≈ 0.0689 and p ≈ 0.000089, although interval repetition alone does not establish a copying error. → Conclusion holds
  - Minor reporting notes (1):
    - Descriptive: "The association between higher CSF Aβ42 (inverted) and lower cognitive function was mediated by hippocampal neurodegeneration (indirect effect: β = −0.06, p = 0.039; 95% CI: −0.10, −0.02)." (Results, Structural equation modeling). Assuming a symmetric normal-Wald interval for the reported indirect coefficient, SE ≈ (−0.02 − (−0.10))/3.92 = 0.0204 gives |z| ≈ 2.94 and two-sided p ≈ 0.0033, rather than 0.039. → Conclusion holds
- **Reporting and transparency:** Noted. Path-specific APOE between-group tests are not presented alongside the moderation claim. The supplied text also lacks explicit ethics, consent, funding and conflict-of-interest declarations and identifies cohort registration rather than preregistration of this SEM analysis.
  - Conflict of interest: not stated · Ethics approval: not stated · Preregistration: not stated · Data availability: not stated · Code availability: declared
- **Reference integrity:** Clean
  - 65 references in manuscript
  - 63 of 63 checkable references found in an index
  - 2 are books, websites or datasets — counted, but not index-checkable
  - Citation details diverge for 1 reference:
    - #37 Papp KV Rentz DM Orlovsky I Sperling RA Mormino EC Optimizing the preclinical Alzheimer's cognitive composite with semantic processing: The PACC5 Alzheimers Dement (N Y) 2017 3 4 668 677 10.1016/j.trci.2017.10.004 29264389 PMC5726754 Papp KV, Rentz DM, Orlovsky I, Sperling RA, Mormino EC. Optimizing the preclinical Alzheimer’s cognitive composite with semantic processing: The PACC5. Alzheimers Dement (N Y). 2017;3(4):668–77. 29264389 10.1016/j.trci.2017.10.004 PMC5726754 (DOI resolves to: “Optimizing the preclinical Alzheimer's cognitive composite with semantic processing: The PACC5” — Papp, 2017 (Alzheimer's &amp; Dementia: Translational Research &amp; Clinical Interventions). Differs from the DOI’s record on: container, authors) https://doi.org/10.1016/j.trci.2017.10.004
  - References not verified automatically are resolved manually within 24 hours.

## Post-publication record

- **Status:** Clean
- **Summary:** No retraction or expression of concern found in Crossref (includes Retraction Watch).
- **Last checked:** 2026-10-17
