GLP-1 Drugs and the Brain: Dementia and Alcohol-Use Research

Less loss of brain volume sounds like good news. But what does it tell you about memory, daily life or whether a drug worked? In the ELAD trial of liraglutide, an encouraging MRI observation appeared alongside a null primary result. Understanding how those findings fit together helps you judge what the study actually established.

GLP-1 stands for glucagon-like peptide 1, a hormone involved in regulating blood glucose. An agonist activates a receptor; semaglutide, for example, activates the same receptor as the body's GLP-1, as its prescribing information explains. The ELAD investigators pursued brain research partly because of earlier findings in animal models. Whether those findings would translate into useful changes in people was a question for the trial.

Keep the drug, formulation and condition attached to each result. Liraglutide and semaglutide are different medicines. An injection and a tablet have different delivery requirements. Those differences can generate hypotheses, but they cannot explain a trial's outcome without evidence testing the explanation.

How to read several outcomes from one trial

The primary endpoint is the main measure chosen to judge whether a study meets its objective. Secondary endpoints address additional effects; exploratory analyses can suggest questions for future research. Testing many outcomes creates more chances for an apparently positive finding even when there is no real effect, unless the analysis accounts for those multiple comparisons. The FDA's guidance on multiple endpoints explains that problem.

That hierarchy lets you keep an interesting observation in view while asking how much confidence it deserves.

ELAD: a null primary result with secondary signals

The ELAD trial of liraglutide randomized 204 people with mild to moderate Alzheimer's disease syndrome to daily injections or placebo for 52 weeks. Its primary endpoint was a PET measure of cerebral glucose metabolism. That outcome showed no significant difference between groups.

A secondary cognitive composite favored liraglutide, while daily-function and clinical-rating outcomes did not significantly differ. Exploratory MRI analyses suggested less loss of temporal-lobe and total gray-matter volume. These were hypothesis-generating findings, with unadjusted multiple comparisons and other limitations.

The positive MRI observation concerns volume. It should not be described as a positive PET finding in glucose metabolism. Nor can the trial establish prevention of dementia in healthy people, a population it did not study.

The semaglutide trials tested another question

The evoke and evoke+ phase III trials studied oral semaglutide in early symptomatic Alzheimer's disease. Their reported primary clinical-rating results at 104 weeks did not show a significant treatment advantage over placebo.

That is a clinically important result. Comparing it with ELAD still requires care: the trials differed in medicine, population and design. Saying that the oral route alone explains the difference would turn a possibility into a conclusion that these comparisons cannot establish.

The prescribing information for oral semaglutide describes absorption predominantly through the stomach and an elimination half-life of about a week. A tablet should not be characterized as a brief exposure that simply disappears after a few hours. Its pharmacology also does not prove a brain-treatment benefit.

Alcohol-use research contains an early positive signal

A 2025 randomized trial by Hendershot and colleagues studied 48 adults with alcohol use disorder over nine weeks. Semaglutide reduced laboratory alcohol self-administration and favored some craving and drinking measures. Other drinking outcomes did not significantly differ.

The trial was small and short, with participants who were not seeking treatment. It offers a reason for further study, rather than a complete current assessment of GLP-1 treatment for alcohol use disorder. It also cannot establish effects on gambling, scrolling, ADHD or every behavior involving reward.

The habit-change article distinguishes everyday routines from substance-use problems that require more specific care.

Keep prescribing decisions with the right indication

If you already take one of these medicines, discuss questions about cognition, appetite, side effects and other medication with the prescriber. A new research finding is a reason for a conversation, not a reason to change a dose or formulation independently.

Oral semaglutide's label also describes delayed gastric emptying and potential effects on the absorption of other oral medicines. The relevance depends on the actual drugs involved; it should not become a universal explanation for changes in attention medication.

Return to the question behind that encouraging scan: what changed, how was it measured, and did the study establish a benefit that matters in daily life? Keep an exploratory volume finding, a cognitive test and everyday function separate as you read. For broader brain-health decisions, the dementia-risk guide helps place this developing drug research alongside other evidence.