Coffee and Tea Linked to Lower Dementia Risk Over Time: A long-running observational study reports that people who regularly drank moderate amounts of caffeinated coffee or tea had a lower risk of developing dementia. The work relies on self-reported beverage habits tracked over years, so it can’t prove cause and effect, but the pattern is consistent with earlier hints that caffeine and other compounds in coffee and tea may influence brain health. Researchers emphasize that “more” isn’t necessarily better: the signal appears tied to everyday, moderate intake rather than extreme consumption. The findings add another thread to the larger dementia-prevention tapestry—alongside sleep, physical activity, blood-pressure control, and social engagement—while also underscoring a key limitation of lifestyle research: associations can be robust even when mechanisms and causality remain unresolved. (Science Daily)

Mutation Explains Rare Adenovirus Vaccine Clotting Syndrome: Scientists report they have pinpointed the molecular trigger behind vaccine-induced immune thrombocytopenia and thrombosis (VITT), the rare clotting-and-low-platelet syndrome linked to adenovirus-based COVID-19 vaccines. In a New England Journal of Medicine study published yesterday, teams in Australia, Canada, and Europe used mass spectrometry and molecular analyses to trace how a mutated autoantibody gene in certain B cells can misread a normal adenovirus protein as platelet factor 4 (PF4), a human clotting-related protein. That mistaken identity prompts pathogenic antibodies that drive dangerous thrombosis—often in the brain or abdomen—alongside platelet depletion and bleeding risk. VITT was estimated to occur in roughly 1 in 200,000 recipients of AstraZeneca or J&J vaccines. Researchers say prior adenovirus infection may prime susceptible B cells. The discovery suggests developers could redesign or remove the implicated adenovirus protein region to prevent VITT while preserving vaccine efficacy. (Ars Technica)

Moderna mRNA Flu Vaccine Review Blocked Amid FDA Turmoil: Reports from Stat News and The Wall Street Journal say Vinay Prasad, the Trump administration’s top vaccine regulator at the FDA, personally refused to review Moderna’s mRNA flu vaccine, overriding career scientists who were prepared to proceed. Stat reports that a senior career official, David Kaslow, wrote a memo objecting and outlining why review should continue. The Journal adds that FDA scientists met with Prasad in early January and warned that blocking the review was the wrong approach. Prasad’s objection reportedly centered not on safety or efficacy issues with Moderna’s vaccine, but on the trial’s comparator choice—despite FDA having previously deemed the plan acceptable with modifications. The episode is fueling industry fears about unpredictability, litigation risk, and internal dysfunction, alongside reports of staff complaints about Prasad’s conduct. (Ars Technica)

Flu Shots Tied to 18% Lower Odds of Heart Attack, Meta-analysis Finds: A large meta-analysis summarized by CIDRAP links influenza vaccination with significantly lower odds of myocardial infarction. Drawing on data from more than 23 million people, the analysis found an overall association suggesting vaccinated individuals experienced fewer heart attacks than those unvaccinated. The authors call for future studies to clarify mechanisms—whether protection comes primarily from preventing influenza-triggered inflammation and clotting risk, from broader immune effects, or from differences in who chooses vaccination. Even with open questions, the findings strengthen the public-health case for flu vaccination as cardiovascular protection, not just respiratory illness prevention—especially for people with elevated baseline risk. The CIDRAP brief frames the result as another reason clinicians may want to discuss flu shots as part of comprehensive heart-health planning during respiratory virus season.(CIDRAP)

Midseason Flu Vaccine Effectiveness in Canada: Low-to-Mid Range Protection: An interim Canadian analysis (test-negative design) estimates this season’s flu vaccine effectiveness at 40% against medically attended influenza A(H3N2), 37% against the newly predominant H3N2 subclade K, and 31% against H1N1. CIDRAP notes that specimens were collected from October 26, 2025 to January 10, 2026 across multiple provinces and that the report was published in Eurosurveillance. The story details why H3N2 performance is closely watched: viral evolution and antigenic mismatch concerns can erode protection, yet the interim numbers still indicate meaningful benefit—especially in some age groups (higher in ages 18–29 than 30–49 in this dataset). The piece also highlights that most vaccines used were inactivated and egg-based, adding context for ongoing strain-selection and vaccine-update debates. (CIDRAP)

Morning Immunotherapy May Improve Outcomes in Lung Cancer Trial: Science News highlights a randomized trial suggesting that timing immunotherapy earlier in the day can improve outcomes for advanced lung cancer—without changing drugs, doses, or adding new agents. The article frames the result as a long-discussed concept (circadian biology affecting treatment response) finally tested with more rigorous clinical design. It reports that morning administration better “revved up” immune responses against tumors and extended survival, supporting the idea that the immune system’s daily rhythms can alter therapeutic effectiveness. The story notes that chronotherapy has circulated for decades but has lacked large, definitive trials in many settings; this study strengthens the case for incorporating time-of-day into oncology protocols where feasible. If replicated, a scheduling change could become a low-cost lever to improve outcomes—though implementation would require clinic workflow adjustments. (Science News)

Using AI Chatbots for Health Advice? Yale Clinician Shares Practical Guardrails: Yale News interviews clinician-educator Shaili Gupta about the benefits—and risks—of relying on chatbots such as ChatGPT for medical guidance. The piece explains why patients turn to these tools (simplifying medical language, summarizing vast information, and providing a “personal” conversational feel) and why that same experience can fuel overtrust. Gupta emphasizes using chatbots for efficiency and education rather than diagnosis, warning that confident-sounding outputs can mislead, especially around urgent symptoms and mental health. The article offers concrete tips: don’t replace your doctor, avoid treating chatbot output as definitive, ask for clarifications and summaries, and use tools for support tasks like reminders or question prep for appointments. The central message is pragmatic: chatbots can boost health literacy, but users need skepticism, context, and clinician partnership to stay safe.(YaleNews)