Evidence report library
Browse health and science claims by evidence strength.
Search deHype reports by topic, grade, study type and clinical relevance. Each card gives the practical verdict first, then links to the full source-chain review.
Explore by topic
Top public report topics, ranked by current archive coverage.
Cancer
A guide to interpreting news and research about cancer, with evidence checks on headlines,…
7 reports
Dementia
Evidence checks about dementia headlines, Alzheimer’s claims, supplements, biomarkers, and…
7 reports
Diet and nutrition
Diet and nutrition headlines often exaggerate the strength of evidence, particularly when …
2 reports
Heart health
Evidence checks about cardiovascular headlines, cholesterol, blood pressure, stroke, stati…
1 report
Longevity
Explore evidence and critical reviews relating to ageing, biological age markers, and anti…
1 report
Mental health
Evidence checks about depression, anxiety, stress, psychedelics, apps, lifestyle intervent…
1 report
Latest evidence reports
Concise archive cards help readers decide whether to open the full source-chain report.
Evidence check: Copper drug clears toxic Alzheimer’s proteins and restores memory
Findings are promising but limited to laboratory mouse models; human benefit, safety, and effectiveness in Alzheimer’s are untested.
Animal/lab only Limited human evidence Hypothesis-generating only
Evidence check: Millions take calcium and vitamin D for stronger bones. A major review finds little benefit
Meta-analysis of nearly 154,000 adults found little to no clinical benefit from calcium and vitamin D supplements for most older adults, supporting cautious re-evaluation of routine use.
Human evidence Randomised human evidence Potentially actionable with caution
Evidence check: Silica Nanoparticles Induce Ferroptosis, Reprogram Immunity in Prostate Cancer Models
The evidence is from mouse prostate cancer models only; while the approach is intriguing, there is no basis yet for clinical claims or implications for patients.
Animal/lab only No human evidence Hypothesis-generating only
Evidence check: Dementia: Joint pain supplement may speed progression
The evidence for glucosamine accelerating dementia is preliminary, indirect, and partly observational; findings require replication and do not justify clinical changes at this stage.
Early-stage only Observational human evidence Not clinically actionable
Evidence check: Science & Medicine: A pill that slows aging? San Antonio researchers are putting it to the test
The article is based on promising animal data and a description of an ongoing human trial, but does not present published human outcomes or direct clinical evidence for slowed ageing in people.
Preclinical/early-stage Limited human evidence Not clinically actionable
Evidence check: New psychedelic-like drugs could treat depression without making you trip
This is an appropriately enthusiastic report of early-stage drug discovery. It references robust lab work and novelty in chemical scaffolds, but there is no clinical, human, or meaningful disease-modifying evidence at this time. Human relevance remains hypothetical.
Early-stage only
Evidence check: A hidden map in your nose could explain how smell works
This article reports an important technical advance in understanding smell mapping in mice. There is no evidence that the organizational map or its manipulation directly restores smell or applies in humans at this time. The clinical impact is speculative.
Early-stage only No human evidence None
Evidence check: Can fasting fight gum disease? Scientists find surprising link
This article reports encourageing pilot trial findings linking fasting-style diets to reduced gum inflammation markers, but evidence is limited to a small sample and surrogate endpoints, not direct clinical benefit.
Early-stage only Limited human evidence Not clinically actionable
Evidence check: Poor sleep linked to rising cancer risk in under-50s
The study suggests an association between poor sleep and early-onset cancers, but evidence is observational and confounding factors prevent causal inference.
Observational, not causal Observational human evidence Not clinically actionable
Evidence check: Your nose could detect Alzheimer’s years before symptoms begin
The research offers scientific insight into early immune changes and olfactory loss in Alzheimer’s, but the article’s diagnostic and therapeutic framing exceeds the current clinical evidence.
Over-framed Limited human evidence Clinically actionable
Evidence check: A personalized vaccine for melanoma cut the risk of cancer returning after five years
While initial trial results for the melanoma vaccine are promising, claims outpace the available peer-reviewed evidence and longer-term follow-up.
Early-stage only Limited human evidence Not clinically actionable
Evidence check: Blood and spinal fluid proteins reveal distinct fingerprints of four brain diseases
This is a robust molecular profiling study in humans, but proposed diagnostic and therapeutic applications are not yet validated, making clinical conclusions premature.
Early-stage only Limited human evidence Hypothesis-generating only