HMDR Cannot Be Used as a Sole Diagnostic Tool
The hippocampal-to-medial temporal lobe ratio (HMDR) and related volumetric measures should not be used as the sole diagnostic tool for Alzheimer's disease or cognitive impairment assessment—they must be supplemented with clinical evaluation and other diagnostic modalities. The evidence consistently demonstrates inadequate sensitivity despite reasonable specificity, making standalone use clinically inappropriate.
Key Evidence Against Sole Reliance
Poor Sensitivity Despite Good Specificity
The most compelling evidence comes from a 2025 study showing that automated hippocampal volumetric measures (including ratios) demonstrate highly variable sensitivity (0.31-0.77) despite better specificity (0.68-0.91) for detecting objective memory impairment 1. This means these measures will miss a substantial proportion of patients with actual disease—an unacceptable failure rate for a sole diagnostic tool.
A 2022 clinical validation study reinforces this limitation, explicitly concluding that "solely using hippocampal volumetric data to predict diagnostic group membership or memory test failure was not supported" and emphasizing the need for supplementation by other tools 2. While extreme hippocampal volume loss was rare in cognitively intact individuals (good specificity), the poor sensitivity necessitates additional diagnostic information.
Modest Correlation with Clinical Outcomes
Even when volumetric measures correlate with memory testing, the relationships are modest at best. The 2025 NeuroQuant study found correlations with the California Verbal Learning Test ranging only from r = 0.20-0.44 1—explaining less than 20% of variance in most cases. This weak predictive relationship undermines confidence in using these measures independently.
Variability in Cut-off Values and Measurement Approaches
A critical 2025 analysis revealed considerable overlap between adjacent MTA scores in publicly available reference images, with inconsistencies in how different systems define atrophy thresholds 3. This variability directly addresses your concern about differing cut-off values across studies and creates confusion that reduces inter-rater reliability.
The 2024 ILV/Hip ratio study, while showing the ratio outperformed individual volumes, still only demonstrated "good performance without statistically significant differences" compared to visual ratings 4—suggesting equivalence rather than superiority, and certainly not standalone adequacy.
When Volumetric Measures Add Value
As Complementary Biomarkers
Volumetric hippocampal measurements perform best as one component of a multi-modal diagnostic approach:
- They effectively distinguish between diagnostic groups at the population level 2, 5
- The 2013 DESCRIPA study showed hippocampal measurements achieved AUC of 0.71 for predicting AD conversion in MCI patients 5
- A 2026 automated volumetry study demonstrated 82% sensitivity and 100% specificity when used alongside clinical assessment 6
Age-Dependent Interpretation Required
The 2021 validation study established that age-dependent thresholds are essential: MTA threshold of 1 for patients <75 years and 1.5 for those ≥75 years 7. Using fixed cut-offs across age groups will generate false positives in elderly patients and false negatives in younger ones.
Critical Pitfalls to Avoid
Ventricular Bias
The 2025 congruency study revealed that many visual rating systems are more consistent with temporal horn area than hippocampal area 3. This means ventricular enlargement from normal pressure hydrocephalus or other causes can falsely elevate atrophy scores, leading to misdiagnosis. Always differentiate hippocampal atrophy from ventricular dilatation.
Heterogeneous Populations
Most validation studies involved research cohorts with known or probable AD. The 2021 study specifically noted that validity must be established in heterogeneous memory clinic populations including subjective cognitive impairment, vascular dementia, and mixed pathologies 7—exactly the real-world scenario where diagnostic uncertainty exists.
Limited Predictive Value in Early Disease
The 2025 preliminary study in cognitively normal individuals found that while hippocampal indices showed trends toward correlation with amyloid PET positivity, between-group differences did not reach statistical significance 8. This suggests limited utility in the earliest disease stages where intervention might be most valuable.
Recommended Diagnostic Algorithm
Use volumetric measures as supporting evidence within this framework:
- Clinical assessment remains primary: Detailed cognitive history, informant report, and neuropsychological testing establish the diagnostic foundation
- Apply volumetric analysis as confirmatory: When clinical suspicion exists, hippocampal volumes/ratios can support or refute the diagnosis
- Interpret with age-appropriate norms: Use established age-dependent thresholds rather than fixed cut-offs
- Correlate with functional measures: Compare volumetric findings with objective memory test performance
- Consider alternative pathology: When volumetric findings conflict with clinical presentation, investigate vascular disease, normal pressure hydrocephalus, or other etiologies
- Serial measurements for progression: Volumetric changes over time provide more diagnostic value than single time-point measurements
The evidence unequivocally demonstrates that while HMDR and related volumetric measures offer valuable objective data, their limitations in sensitivity, variability in methodology, and modest clinical correlations preclude their use as standalone diagnostic tools 2, 1.