# Magellan Longevity Open Evidence Dataset

**Version 1.0.0 · released 2026-08-11 · Creative Commons Attribution 4.0 International (CC BY 4.0)**

130 longevity and healthspan products graded on the strength of human evidence, each linked to the primary literature behind the grade, plus 2692 study-level records covering 2646 distinct PubMed identifiers (2346 with DOIs).

Landing page: <https://magellanlongevity.com/data/>

## Files

| file | rows | fields |
|---|---:|---:|
| `products.csv` | 130 | 32 |
| `products.json` | 130 | 32 |
| `products.jsonl` | 130 | 32 |
| `studies.csv` | 2692 | 32 |
| `studies.json` | 2692 | 32 |
| `studies.jsonl` | 2692 | 32 |
| `data-dictionary.json` / `.md` | — | field definitions and provenance |
| `datapackage.json` | — | Frictionless tabular-data-package descriptor |
| `dataset-metadata.json` | — | schema.org Dataset descriptor |
| `CITATION.cff` / `citation.bib` / `citation.ris` | — | citation metadata |
| `checksums.txt` | — | SHA-256 for every file above |

## What makes it unusual

- The **rationale and the caveat behind every grade** ship with the grade. 130 of 130 rows carry a published caveat.
- **Null and negative results are kept.** 144 study rows report no difference; 1 product is graded "Tested — did not work".
- **Doses are only present where a human trial range exists** (104 of 130 rows). Nothing is inferred.
- Grades are produced by executing the same code that renders the public site, so the export cannot drift from what readers see.

## Quick start

```bash
sha256sum -c checksums.txt
```

```python
import pandas as pd
p = pd.read_csv("products.csv")
s = pd.read_csv("studies.csv")
p.groupby("evidence_grade").size()
s[s.is_primary_record_for_pmid].shape   # (2646, 32)
```

## Limitations

This is an evidence **map**, not a systematic review: no protocol, no exhaustive search, no risk-of-bias tool, no meta-analysis.
`reported_direction` is a published rule cascade over the source's own words, not an appraisal — 1365 rows are `unclear` because no rule fired, which is not evidence of absence. Prices are a 2026-08-11 snapshot. Full limitations: <https://magellanlongevity.com/data/>

## Citation

```
Radu, G. (2026). Magellan Longevity Open Evidence Dataset (Version 1.0.0) [Data set]. Magellan Longevity. https://magellanlongevity.com/data/
```

```bibtex
@dataset{magellan_longevity_evidence_2026,
  author       = {Radu, Gabriel},
  title        = {Magellan Longevity Open Evidence Dataset},
  year         = {2026},
  version      = {1.0.0},
  publisher    = {Magellan Longevity},
  url          = {https://magellanlongevity.com/data/},
  license      = {CC-BY-4.0},
  note         = {130 graded longevity products and 2692 linked study records}
}
```

## Licence

Creative Commons Attribution 4.0 International (CC BY 4.0) — <https://creativecommons.org/licenses/by/4.0/>. Attribution required:

> Magellan Longevity Open Evidence Dataset v1.0.0 (Gabriel Radu, DO, 2026), https://magellanlongevity.com/data/ — CC BY 4.0

Article titles, journal names, DOIs, PMIDs and short quoted findings are facts about or brief quotations from third-party publications, included for identification and attribution. No full texts or complete abstracts are redistributed.

**Not medical advice.** Evidence grades describe the strength of published human research. They are not clinical recommendations and nothing here is intended to diagnose, treat, cure or prevent any disease.