The headlines on GLP-1, brain, dementia and mood swing between two poles. One week these medicines are a future Alzheimer treatment. The next week a major trial says they are not. Both stories are real. They simply answer different questions. This piece sorts the proven from the promising, because the difference matters most to women, who carry the heavier share of both Alzheimer disease and depression.
The 30-second summary
- A large analysis of more than 2 million people on GLP-1 medicines linked them to lower risk of dementia, addiction, suicidal ideation and self-harm. This is observational, not proof of cause.
- The rigorous EVOKE and EVOKE+ Phase 3 trials found oral semaglutide did not slow early Alzheimer disease. Protecting a healthy brain is not the same as treating an established disease.
- Almost two-thirds of people living with Alzheimer disease are women, and women carry roughly double the lifetime risk of depression, so this evidence lands hardest on us.
What the observational data actually shows
In January 2025, a team at Washington University in St. Louis led by Dr. Ziyad Al-Aly published one of the broadest looks yet at these drugs in the journal Nature Medicine. They tracked 175 health outcomes across more than 2 million people with diabetes and compared those who took GLP-1 medicines, such as semaglutide (Ozempic, Wegovy) and tirzepatide (Mounjaro, Zepbound), with those who took older diabetes drugs.
The brain findings were striking. GLP-1 use was linked to lower risk of neurocognitive disorders, including Alzheimer disease and dementia. It was also linked to lower risk of addiction to alcohol, cannabis, stimulants and opioids, and to lower risk of seizures, suicidal ideation, self-harm, bulimia and psychotic disorders. The study also flagged real risks, including pancreatitis and kidney problems that can build quietly, which is why monitoring matters.
Here is the honest caveat. This was an observational study. It can show a pattern, a strong and consistent one, but it cannot prove the medicine caused the lower risk. People who stay on a treatment, who attend appointments, who have access to care, differ in many ways from those who do not. The signal is encouraging. It is not a verdict.
The reality check: EVOKE and EVOKE+
If observational data hints, a randomised trial decides. In early 2026 The Lancet published EVOKE and EVOKE+, two large Phase 3 trials run across 566 sites in 40 countries. They tested oral semaglutide up to 14 mg daily in people aged 55 to 85 with amyloid-confirmed early Alzheimer disease, meaning mild cognitive impairment or mild dementia.
The result was sobering. On the main cognitive measure, the CDR-SB score, the change over two years was essentially the same on semaglutide as on placebo. The differences were not statistically significant. Oral semaglutide did not slow the clinical progression of early Alzheimer disease.
This is the heart of the GLP-1 brain, dementia and mood story, and it is easy to misread. The trial does not cancel the observational signal. It refines it. Lowering the long-run risk of developing dementia in a broadly healthy brain is a different task from reversing or slowing a disease that has already taken hold. A seatbelt lowers your risk of injury. It does not heal a broken bone. The same medicine can plausibly do the first and still fail at the second.
Mood, food noise and the cycle
Beyond dementia sits the question women ask most often: how will this affect my mood and my relationship with food? The biology gives reasons for cautious optimism. GLP-1 receptors sit in brain regions tied to reward, impulse control and inflammation. Researchers think this is part of why the medicines quiet what many women call "food noise," the constant background chatter about the next snack. When the noise drops, the mental load of eating drops with it, and that alone can lift mood.
But the picture is not uniformly bright, and you deserve the full one. Some women report flatter mood, less joy, or low energy as appetite fades. Rapid loss of interest in food can shade into eating too little, which starves both body and brain. Mood also moves with the menstrual cycle and with perimenopause, and a dip in the late-luteal week can be mistaken for a medication effect. The drug, the cycle and life stress braid together, and untangling them takes tracking, not guesswork. If low mood is persistent, worsening, or comes with thoughts of self-harm, that is a same-day call to your clinician, not a number to log.
Why women should read this first
The reason to put women at the centre is arithmetic. Almost two-thirds of Americans living with Alzheimer disease are women, and women are nearly twice as likely as men to develop it. Women also carry roughly double the lifetime risk of depression. So when a class of medicines touches dementia risk and mood at the same time, women are the population with the most at stake and, too often, the least studied. Most of this brain evidence was not designed around female hormones, perimenopause, or the cycle. Read every promising headline through that lens.
The practical takeaway is steady, not sensational. Do not start or stay on a GLP-1 medicine for brain protection alone. The trial evidence does not support that. Do pay attention to mood, energy, sleep and appetite as you go, because those are the signals that change first and matter most day to day. And bring patterns, not single bad days, to the people who prescribe for you.
What Steady does with this
- Steady tracks mood-adjacent signals over time, including energy, sleep and appetite changes, so a real downward trend stands out from one rough week tied to your cycle.
- Steady logs 14 GLP-1 symptoms and overlays your cycle phase, which helps separate a medication effect from a predictable late-luteal dip rather than blaming the drug for both.
- Steady turns a month of mood, symptom and dose notes into one clear page you can hand to your prescriber, so a conversation about brain and mood starts from your data, not your memory.
Read next: GLP-1 and mood: the real link, GLP-1 and food noise: why cravings go quiet, GLP-1 and inflammation
Sources
- Xie Y, Choi T, Al-Aly Z. Mapping the effectiveness and risks of GLP-1 receptor agonists. Nature Medicine, January 2025. Nature Medicine article
- WashU Medicine. Study identifies benefits, risks linked to popular weight-loss drugs. The Source, January 2025. WashU Medicine summary
- EVOKE and EVOKE+ investigators. Efficacy and safety of oral semaglutide 14 mg in early-stage symptomatic Alzheimer's disease: two phase 3 trials. The Lancet, 2026. The Lancet article
- Alzheimer's Association. Women and Alzheimer's. Women and Alzheimer's
- Frontiers in Global Women's Health. Women and the risk of Alzheimer's disease, 2023. Review article
Medical disclaimer: Articles in the Steady research hub are educational, not medical advice. They cannot account for your personal history, medications, or symptoms. Always talk with the clinician who prescribes your GLP-1 medication before making any change, and seek urgent help for persistent low mood or thoughts of self-harm. See our full medical disclaimer.