Lower brain choline may mark anxiety, not a supplement fix

A large brain-imaging meta-analysis has given the choline-anxiety conversation a more precise frame, and a less exciting one than the headlines suggest. In a 2026 Molecular Psychiatry meta-analysis, Richard J. Maddock, Jason Smucny and colleagues pooled 25 proton magnetic resonance spectroscopy datasets and found that people with anxiety disorders tended to show lower total choline in cortical regions, especially areas involved in emotional control. The effect was not trivial. Among the higher-quality studies, the gap reached a Hedges’ g of -0.64, and the average reduction in cortical total choline was about eight percent.

It sounds like the start of a supplement story. It is not there yet. Choline is an essential nutrient, it contributes to cell-membrane phospholipids and acetylcholine biology, and it already has a health halo in the supplement market. But the new paper is best read as a biomarker paper first. The sharper question is not whether anxious adults should rush to buy phosphatidylcholine capsules. It is whether lower brain choline looks like a repeatable chemical signature of anxiety, and whether that signature points to a mechanism worth testing in trials.

Even the authors have drawn that line carefully. In the UC Davis summary carried by ScienceDaily, Smucny called it “the first meta-analysis to show a chemical pattern in the brain in anxiety disorders.” Maddock was even plainer about the next step: “We don’t know yet if increasing choline in the diet will help reduce anxiety. More research will be needed.” That is the right level of restraint. Imaging can tell readers something important about where to look. It cannot, on its own, tell them how to treat themselves.

What the brain-scan paper actually found

Magnetic resonance spectroscopy does not diagnose a vitamin deficiency in the way a blood test might. It estimates concentrations of certain brain chemicals in living tissue. In this case, the paper looked at total choline, often shortened to tCho, a signal tied more to membrane turnover and phospholipid metabolism than to “how much choline you ate last week.” The distinction matters because a low tCho readout can reflect altered brain chemistry without proving a simple dietary shortfall.

Pooled across 25 published datasets, the paper carries more weight than a one-off scan study. The authors combined 370 patients with anxiety disorders and 342 controls, reducing some of the noise that has made psychiatric imaging hard to interpret. The signal also appeared most clearly in prefrontal and cortical regions, which fits the brain systems usually implicated in threat processing, regulation and decision-making. For a field that often produces mixed small studies, that degree of convergence is useful. It suggests anxiety may carry a metabolic footprint, not just a symptom label.

The study is transdiagnostic, too, rather than narrowly attached to one brand of anxiety. That broad lens makes the finding more interesting as a research clue, but it raises a complication for readers looking for a consumer takeaway. A biomarker that travels across several anxiety disorders may be telling researchers something about chronic stress, altered neural energetics or membrane signaling. It is not automatically telling the public that one nutrient deficiency sits underneath every case.

Why the supplement leap is too fast

Outside the scanner, there is some prior literature linking choline status to mental health. The Hordaland Health Study, associated with Ingvar Bjelland and published in The American Journal of Clinical Nutrition, examined choline in relation to anxiety and depression and helped establish that the idea did not come out of nowhere. A newer adult dietary choline and betaine paper reached the same topic from the intake side rather than the brain-imaging side. Taken together, those studies make it reasonable to ask whether choline biology and anxiety are connected.

Reasonable is not the same thing as settled. Nutrition psychiatry is full of associations that look compelling in observational work and then weaken once researchers try to change outcomes in real people. Diet quality, sleep, alcohol use, medication exposure, socioeconomic status and chronic illness all travel with mental health. So does reverse causation. Anxious people may eat differently, sleep less, move less or use more substances, and those patterns can change nutrient intake or metabolism without nutrients being the root cause of the disorder.

Read this way, the paper is a better map of the question, not an answer to it. Brain choline could be a causal factor. It could be a downstream effect of long-running anxiety. It could also be a marker of some broader process, such as inflammation, stress physiology or altered membrane turnover, that happens to overlap with choline pathways. The new meta-analysis narrows the field. It does not choose among those explanations.

What readers should and should not take from it

Most readers will hear “lower choline in the brain” and translate it into “I am low in choline.” That step is too neat. The paper did not test whether the participants met dietary choline requirements, whether blood choline tracked with brain tCho in a predictable way, or whether supplementation normalized the imaging signal. It also did not show that changing the signal would change panic, worry or avoidance symptoms.

For clinicians and researchers, though, the finding is not small. Psychiatric biomarkers are rare, especially ones that hold up across pooled datasets. If anxiety disorders do show a reproducible choline-related signature in the cortex, that gives the field something concrete to interrogate. One could imagine future work separating generalized anxiety from panic disorder, stratifying by medication use, or testing whether the people with the lowest tCho respond differently to therapy, diet or sleep interventions. Those are better questions than the retail version of the story.

Soon enough, the market version will arrive. Choline already appears in prenatal formulas, nootropic stacks and phospholipid products, so an anxiety headline is likely to be recycled into marketing copy. Readers should be skeptical. No randomized trial in the fact bundle shows that choline supplements reduce anxiety symptoms, and the lead authors explicitly warned against making that leap. A useful rule in supplement reporting is simple: if the most careful people attached to the paper are not recommending the intervention yet, the copywriter should not be either.

What a real next study would need to show

The convincing next study would not be another speculative headline. It is an interventional study that starts with well-characterized patients, measures baseline diet and possibly blood biomarkers, tracks brain tCho, and then tests whether changing choline intake actually changes symptoms. Ideally, that trial would also separate people who are plausibly low in choline from people who are not. If everyone is lumped together, a real effect could vanish into the average.

The strongest reading, for now, is narrow but important. The new meta-analysis suggests that anxiety disorders may be associated with a measurable reduction in cortical choline-containing compounds. That makes choline biology more relevant to anxiety research than it was before this paper. It does not make choline supplementation an evidence-based anxiety treatment. For readers deciding what to do on Monday morning, that distinction is the whole story.

References

1. Maddock RJ, Smucny J, et al. Transdiagnostic reduction in cortical choline-containing compounds in anxiety disorders: a 1H-magnetic resonance spectroscopy meta-analysis. Molecular Psychiatry. 2026. https://www.nature.com/articles/s41380-025-03206-7
2. Bjelland I, et al. Choline in anxiety and depression: the Hordaland Health Study. American Journal of Clinical Nutrition. 2023. https://ajcn.nutrition.org/article/S0002-9165(23)23277-3/fulltext