Relying on AI to write essays may weaken brain engagement over time; new USPSTF recommendation for domestic violence screening; the deadliest types of heart disease – Morning Medical Update

Relying on ChatGPT for essays may weaken brain engagement over time

A new study using electroencephalography (EEG) brain scans found that people who relied on large language models (LLMs) like ChatGPT to help write essays showed lower cognitive activity and weaker brain connectivity than those who used search engines or no tools at all. Over four sessions, LLM users consistently underperformed across neural, linguistic and behavioral metrics. When switched back to writing without AI, they showed signs of cognitive under-engagement. The findings suggest long-term use of AI writing assistants may lead to “cognitive debt,” raising questions about their role in education. Find the full study here.

USPSTF: Screen for intimate partner violence, but not enough evidence yet on elder abuse

The U.S. Preventive Services Task Force (USPSTF) now recommends that clinicians screen women of reproductive age — including those who are pregnant or postpartum — for intimate partner violence (IPV) and refer those who screen positive to multicomponent support interventions. The Task Force found moderate net benefit in these efforts. However, it concluded there is still insufficient evidence to recommend for or against screening for caregiver abuse of older or vulnerable adults.

Heart disease deaths are down, but new risks are on the rise

Heart attacks are no longer the top killer in U.S. heart disease deaths, according to a new study in the Journal of the American Heart Association. Researchers found heart disease mortality has dropped 66% over the past 50 years, largely due to advances in emergency care and prevention. But deaths from heart failure, arrhythmias and hypertensive heart disease are rising. The shift suggests Americans are living longer after initial cardiac events but increasingly dying from chronic heart conditions tied to aging, obesity and diabetes.