239 Health claim verifications avg. score 4.4/10 69 rated true or mostly true 169 rated false or misleading
“An artificial intelligence model can detect early-stage breast cancer with approximately 94% accuracy, surpassing the average performance of radiologists.”
The claim conflates AUC/AUROC scores (~0.93) with "accuracy," which are different metrics. The best available meta-analytic evidence reports pooled AI sensitivity of 0.85 and AUC of 0.89 — not 94%. Critically, 2025 RSNA studies show AI misses approximately 14% of cancers, with false negatives concentrated in smaller, early-stage tumors in dense breasts — the very cases the claim highlights. While AI can match or modestly exceed average radiologists in some contexts, the specific "~94% accuracy for early-stage detection" framing significantly overstates the evidence.
“Higher cholesterol levels in the body lead to higher testosterone production.”
While cholesterol is a necessary biochemical precursor for testosterone synthesis inside cells, the claim that "higher cholesterol levels in the body" lead to higher testosterone production is not supported by human evidence. Multiple population-level studies (including NHANES data) find no association—or even an inverse relationship—between circulating cholesterol and testosterone levels. The rate-limiting step is intracellular cholesterol transport into mitochondria, not the amount of cholesterol in the bloodstream. Research also shows that low testosterone can itself raise circulating cholesterol, reversing the claimed causal direction.
“Heating olive oil produces carcinogenic compounds that pose a cancer risk to humans.”
Heating olive oil — especially past its smoke point or during prolonged, repeated frying — can generate compounds like aldehydes and PAHs that are classified as potentially carcinogenic. However, the claim is misleading because no epidemiological evidence establishes that heating olive oil poses a demonstrated cancer risk to humans under normal cooking conditions. Olive oil actually produces fewer toxic byproducts than many other cooking oils, and overall olive oil consumption is associated with reduced cancer risk in meta-analyses.
“Severe COVID-19 infection can increase the risk of faster lung cancer development.”
Multiple peer-reviewed studies and academic medical centers report an association between severe COVID-19 (hospitalization-level) and increased subsequent lung cancer incidence, supported by plausible inflammatory mechanisms. However, the claim omits critical context: the epidemiological evidence is correlational, not causal; the strongest mechanistic research involves metastatic cancer cell reactivation rather than new lung tumor formation; no major health authority has confirmed SARS-CoV-2 as a lung carcinogen; and vaccination appears to mitigate the risk. The association is real, but the causal framing is premature.
“The Apple Watch can predict heart failure with high accuracy using an AI model that analyzes peak oxygen uptake (pVO2) data.”
The claim overstates what current evidence supports. While the TRUE-HF AI model uses Apple Watch data to estimate daily fitness surrogates correlated with pVO2, the Apple Watch does not directly measure peak oxygen uptake — it estimates submaximal VO2max with known error and bias. Published findings show promising risk associations (e.g., threefold higher event risk per 10% fitness drop), but no validated "high accuracy" prediction metrics (AUC, sensitivity, specificity) for heart failure have been reported for this specific pVO2-based approach. The research is promising but preliminary.
“A newly developed drug has demonstrated the ability to reverse cognitive decline associated with Alzheimer's disease in animal models.”
The claim is accurate on its own terms. Multiple independent research groups have reported newly developed compounds — including GL-II-73, P7C3-A20, NU-9, and FLAV-27 — that reversed cognitive deficits in rodent models of Alzheimer's disease. However, the claim omits critical context: animal models are widely recognized as poor proxies for human Alzheimer's, no such reversal has been demonstrated in humans, and the history of translating preclinical AD successes to clinical benefit is marked by repeated failure.
“High sugar intake is associated with a 30% increased risk of developing depression.”
The claim overstates the evidence. A ~30–31% increased risk has been found specifically for sugar-sweetened beverage consumption, but the most comprehensive meta-analyses of overall sugar intake report a smaller association of roughly 21%. One prospective-cohort meta-analysis of total sugar found no statistically significant link at all. Presenting "30%" as the general figure for "high sugar intake" conflates a subgroup-specific finding with the broader scientific picture, and all results reflect associations, not proven causation.
“Individuals with Type 1 diabetes have nearly three times the risk of developing dementia compared to individuals without Type 1 diabetes.”
Type 1 diabetes is associated with elevated dementia risk, but "nearly three times" overstates the typical finding. The most comprehensive quantitative synthesis — a 2025 systematic review and meta-analysis — reports a pooled hazard ratio of approximately 1.50 (a 50% increase), while a large nationwide cohort study found roughly double the risk. The ~2.8× figure comes from one specific recent study and media reports echoing it, not from the broader evidence base. The claim cherry-picks the highest estimate rather than reflecting the range of peer-reviewed findings.
“Diets high in fast-acting carbohydrates are associated with an increased risk of developing dementia.”
The claim is well-supported by multiple independent, peer-reviewed human studies — including a large UK Biobank prospective cohort — showing that diets high in fast-acting (high glycemic index/load) carbohydrates are associated with increased dementia risk. The association is further backed by plausible biological mechanisms including insulin resistance and neuroinflammation. However, the evidence is observational, effect sizes are modest, genetic factors like APOE4 status modify the risk, and the claim omits that low-GI carbohydrates may be protective.
“Exercise Pegasus, a pandemic simulation, either caused or predicted the United Kingdom meningitis B outbreak.”
This claim is false. Exercise Pegasus simulated a fictional novel enterovirus (a virus), while the UK meningitis B outbreak is caused by Neisseria meningitidis serogroup B (a bacterium) — two biologically unrelated pathogens. The MenB strain had been circulating in the UK for roughly five years before the exercise even took place. Full Fact and UK government officials have explicitly dismissed the alleged connection as a conspiracy theory with "simply no evidence." The only source supporting the claim is a low-authority conspiracy blog.
“Consumption of tomatoes causes inflammation in the human body.”
The claim that tomatoes cause inflammation is not supported by the scientific evidence. Multiple peer-reviewed studies and systematic reviews show that tomatoes and their key compound lycopene are either neutral or actively anti-inflammatory, reducing biomarkers like CRP and IL-6 in controlled human trials. The only supporting arguments rely on unproven hypotheses about solanine, a study protocol with no published results, and anecdotal reports from specific patient subgroups — none of which establish general causation.
“Regular use of dry Finnish sauna improves cardiovascular health markers, including blood pressure and arterial flexibility.”
Multiple peer-reviewed systematic reviews and prospective studies consistently associate regular Finnish sauna use with lower blood pressure and reduced arterial stiffness, supported by plausible biological mechanisms. However, the claim overstates certainty: much of the evidence is observational, at least one randomized controlled trial in coronary artery disease patients found no improvement, and acute post-session effects may not translate to lasting benefits for all populations. The association is well-established, but calling it a proven general improvement goes slightly beyond what the current evidence firmly supports.
“The United States Food and Drug Administration has approved leucovorin as a broad treatment for autism.”
This claim is false. The FDA approved leucovorin in March 2026 only for cerebral folate deficiency (CFD), an ultra-rare genetic condition affecting roughly 1 in a million people — not for autism. Leucovorin remains investigational for autism, the American Academy of Pediatrics does not recommend its routine use for autistic children, and a key study supporting leucovorin's autism benefits was retracted in January 2026 due to data irregularities. No FDA-approved broad treatment for autism spectrum disorder exists.
“Ultra-processed foods account for the majority of calories consumed by American adults as of March 2026.”
The claim is well-supported. A 2025 CDC report found American adults consumed 53% of their calories from ultra-processed foods during 2021–2023, and peer-reviewed research consistently places the figure above 50%. However, the most recent primary data doesn't extend to March 2026 specifically — it's an extrapolation from a 2021–2023 survey window. No evidence suggests the trend has reversed below the majority threshold, but the "as of March 2026" framing implies more current measurement than exists.
“AI chatbots, such as ChatGPT, provide medical advice that is consistently reliable and safe for users.”
The claim that AI chatbots like ChatGPT provide "consistently reliable and safe" medical advice is not supported by the evidence. Multiple high-quality studies from 2024–2026 show ChatGPT gave incorrect advice in over 51% of medical emergencies, exhibited hallucination rates of 50–82%, and correctly identified conditions in fewer than 34.5% of real-world cases. ECRI designated AI chatbot misuse as the top health technology hazard for 2026. While chatbots show promise in narrow, controlled tasks, their performance is neither consistent nor safe for general medical advice.
“Long-term use of wireless earbuds may negatively affect brain function due to electromagnetic field exposure.”
No peer-reviewed study has demonstrated that wireless earbuds impair brain function. Bluetooth earbuds emit roughly 100–1,000 times less RF radiation than cell phones held to the head. The WHO, CDC, and Bluetooth-specific research consistently find no adverse neurological effects at these power levels. The claim's key supporting evidence comes from cell phone studies on children—a fundamentally different exposure scenario. While long-term earbud-specific research is limited, presenting speculative extrapolation as plausible risk is not supported by current science.
“Eating chocolate every day reduces the risk of heart disease.”
The claim overstates the evidence. While observational studies link moderate chocolate consumption to lower cardiovascular risk, the strongest randomized trial (COSMOS) found no significant reduction in total cardiovascular events. Benefits appear limited to modest amounts of high-flavanol dark chocolate — not "chocolate every day" broadly. The claim conflates correlation with causation, ignores dose-dependent risks (a J-shaped curve where excess intake may be harmful), and equates cocoa flavanols with everyday commercial chocolate.
“Vaccines contain ingredients that are harmful to human health.”
This claim is misleading. While it's true that rare allergic reactions to vaccine excipients (like gelatin or PEG) occur in roughly 1 per million doses, the unqualified statement implies vaccines are broadly dangerous. The overwhelming scientific consensus — including WHO, the CDC, the AAP, and a landmark study of 1.2 million children — confirms that vaccine ingredients like aluminum adjuvants and thimerosal are safe at the doses used, with no causal link to autism, neurological disorders, or systemic harm.
“Consuming pineapple core around the time of embryo transfer increases the success rate of IVF implantation.”
There is no scientific evidence that eating pineapple core around embryo transfer increases IVF implantation success. Multiple fertility clinics and medical sources confirm no published human studies support this claim. The idea stems from bromelain's general anti-inflammatory properties, but theoretical plausibility is not proof of clinical benefit. The claim presents an unproven folk remedy as an established fact.
“It takes exactly 21 days to form a new habit.”
The claim that it takes "exactly 21 days" to form a habit is a widely debunked myth. The figure originated from a plastic surgeon's anecdotal observations in 1960, not from any scientific study. Multiple peer-reviewed studies and a 2024 meta-analysis show habit formation typically takes 59–66 days, with individual variation ranging from 4 to 335 days. Caltech researchers explicitly stated the 21-day estimate "was not based on any science." There is no fixed universal timeline for habit formation.