How Missing Medication Data Contributes to Bias in Alzheimer’s Disease Machine Learning Models
Bibliographic record
Abstract
Alzheimer's disease (AD) is the most common cause of dementia, yet many cases go undiagnosed due to limited access to expensive brain scans and lab tests. This study investigated whether medication data could help identify AD. Using data from 1,785 participants in the US-representative National Health and Nutrition Examination Survey 2013– 2014, we identified 105 individuals (5.9%) with memory test scores suggesting possible AD. We evaluated seven machine learning models using medication features. Models that incorporated contextual prescription information, including the reasons for medication use and conditions being treated, achieved the best performance (area under the receiver operating characteristic curve [AUC] 0.61–0.63). In contrast, models using only basic drug names or provider information performed poorly (AUC 0.46–0.51). This performance difference was statistically significant (t = 14.98, p < 0.0001). Our findings suggest that medication data, when analyzed with attention to clinical context, could serve as a low-cost tool for identifying individuals at risk of AD. This approach may help address diagnostic disparities in settings with limited access to advanced testing.
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 | 0.120 | 0.242 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.003 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.002 | 0.003 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.003 | 0.004 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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".