📚 McLean Hospital, a Harvard Medical School affiliate in Massachusetts, has announced plans for a new Center for Metabolic Psychiatry. A $20 million grant from the Baszucki Group supports the center and related education and training initiatives exploring how metabolic treatments might contribute to psychiatric care.
An investment that size caught my attention this week.
It also brought me back to a conversation during a break at a recent PMHNP review event. A student asked about GLP-1 medications and the role of the psychiatric nurse practitioner.
Where do we fit? Should we be prescribing them? Monitoring them?
My answer started with the basics: know whether your patient is taking one, know why, and understand how it might affect the rest of their treatment.
But I have kept thinking about the question.
We have explored GLP-1s and mood and what PMHNPs should do with the nutrition evidence in earlier issues of Eva’s Tea. Those conversations bring us to a more specific question:
Could treating a metabolic problem improve the psychiatric illness itself and could that eventually help us choose treatment?
🖥️ The CGM is already in the room
Picture a patient sitting across from us with a continuous glucose monitor on the back of their upper arm. We ask about mood, sleep, medication adherence, and side effects.
Do we ask about the monitor?
What prompted its use? Who is following the results? Has anything changed in the patient’s eating, energy, or medication routine?
We teach the biopsychosocial model. This is an opportunity to put it into practice.
And metabolic psychiatry asks us to consider how much those answers might eventually influence the treatment plan.
Could metabolism help us choose treatment?
One small study offers an early reason to investigate.
In TRIO-BD, researchers randomized 45 people with treatment-resistant bipolar depression and insulin resistance to metformin or placebo. Insulin resistance reversed in 10 metformin participants and one placebo participant.
☕️ Those whose insulin resistance reversed experienced greater improvements in depression and functioning.
There is an important limitation: that psychiatric comparison grouped patients by whether insulin resistance reversed, rather than simply by their randomized treatment assignment. It cannot establish that metformin broadly treats bipolar depression.
Still, the question is compelling. Could insulin resistance help identify a group of patients who might benefit psychiatrically from a targeted metabolic intervention?
If larger studies establish that relationship, a future treatment discussion could include which metabolic findings are relevant, whether addressing them offers a reasonable chance of improving psychiatric symptoms, and how we will measure the response.
We do not yet have a validated metabolic profile that reliably tells us which psychiatric treatment to select.
Could nutrition become a more specific prescription?
We have already discussed why “eat better” is an incomplete clinical recommendation. The next question is whether a particular nutritional intervention could become the defined treatment for a particular psychiatric presentation.
Ketogenic therapy is being studied in that role.
In a 2026 trial of 88 adults with treatment-resistant depression, the ketogenic group had an additional 2.18-point improvement on the PHQ-9 at six weeks compared with a control diet. Both groups received dietary support and both improved. At 12 weeks, the difference between groups was no longer statistically significant. That is a modest short-term signal, with durability still uncertain.
Research in schizophrenia-spectrum and bipolar I disorders is also preliminary. A 2026 trial randomized 58 participants for one month and found metabolic benefits. A subgroup completing four months of ketogenic treatment improved in psychiatric symptoms and cognition, but that longer phase lacked a concurrent control group.
What interests me is what a usable treatment pathway might eventually look like: a defined indication, a clinician and dietitian with relevant training, medication review, agreed-upon outcomes, and a plan for deciding whether to continue.
Before asking someone to take on a demanding intervention, I want to be able to explain what we hope will improve, when we will reassess, and what we will do if it is not helping.
That is where future research could make this much more useful in an ordinary practice.
Where GLP-1s fit into this conversation
GLP-1s bring us back to the student’s question and illustrate why we need to distinguish metabolic benefits from psychiatric benefits.
In HISTORI, 154 antipsychotic-treated adults with schizophrenia, prediabetes, and overweight or obesity received semaglutide or placebo. Over 30 weeks, semaglutide produced approximately 9.2 kg greater weight reduction and improved glycemic control without worsening psychiatric symptoms. That supports treating metabolic illness in this population; it does not establish semaglutide as a treatment for psychosis.
☕️ For the psychiatric prescriber, there are also immediate medication questions. Lurasidone requires at least 350 calories with each dose. If appetite suppression makes that meal difficult, taking every tablet can still mean receiving less medication exposure than intended.
☕️ “Are you taking it?” needs a follow-up: “What food are you taking the Lurasidone with?”
Whether we prescribe metabolic medications depends on our preparation, competence, applicable scope, and practice setting. We can still recognize when another clinician’s treatment changes something relevant to ours.
A center like McLean can bring disciplines together. A private PMHNP may be coordinating with a PCP across town and a dietitian in another health system.
🔄 An integrated practice can make coordination much easier. In solo practice, we need to build those relationships deliberately.
During a behavioral medication follow-up, we can connect metabolic concerns to the psychiatric picture. Are changes in weight, appetite, or glucose occurring alongside changes in mood, energy, or functioning? Could the medication regimen be contributing? What should the clinician treating the metabolic condition know?
Coordination can begin with something practical:
“Our shared patient continues to experience depressive symptoms and has had recent changes in glucose and weight. I’m reviewing whether the psychiatric regimen may be contributing and would appreciate your assessment. If you’ve already ordered metabolic labs, I’d like to use those results for the patient’s antipsychotic monitoring whenever possible, so we avoid duplicate testing and an extra blood draw.”
If metabolic treatment begins, our contribution includes documenting the psychiatric baseline and following it over time. Are symptoms improving? Is the patient participating more in daily life? Has anything changed beyond the laboratory results? Metabolic and psychiatric improvements are separate outcomes worth tracking. When both improve, that is useful clinical information, even if we cannot determine exactly what caused the change.
What evidence would help me use this in practice?
Before recommending a metabolic intervention specifically to improve psychiatric symptoms, I would want research that answers three practical questions:
Who is most likely to benefit? Which patients, diagnoses, or metabolic findings would guide that recommendation?
How much does it help? Does the treatment produce meaningful improvements in psychiatric symptoms, daily functioning, or quality of life?
Does the improvement last? What do longer studies tell us about sustained benefit, risks, and the support patients need to continue treatment?
Those answers would help me explain why I am recommending a particular intervention, what the patient might reasonably expect, and how we would evaluate whether it is helping.
Those answers would help us decide when a metabolic intervention belongs in a psychiatric treatment plan and how to judge whether it is helping.
For now, we can bring metabolic health into the assessment, consider how our prescribing fits, and follow psychiatric outcomes when another clinician initiates treatment. The possibility ahead is that metabolic findings may help us choose additional treatments for specific patients.
That is what interests me about McLean’s investment.
And it is what I would add to my answer to that student: start by knowing what your patient is taking and why. Then ask how that information belongs in your treatment plan.
☕️ The CGM is already in the room. We can ask about it.
Selected references
Abram, S. V., Kyner, J. M., Vu, A., et al. (2026). Metabolic improvements with a ketogenic diet correlate with symptom improvement in psychosis: A randomized controlled trial. Schizophrenia Bulletin. https://doi.org/[add DOI]
Calkin, C. V., Chengappa, K. N. R., Cairns, K., et al. (2022). Treating insulin resistance with metformin as a strategy to improve clinical outcomes in treatment-resistant bipolar depression (The TRIO-BD study): A randomized, quadruple-masked, placebo-controlled clinical trial. The Journal of Clinical Psychiatry. https://doi.org/[add DOI]
Ganeshalingam, A. A., Uhrenholt, N., Arnfred, S., et al. (2025). Semaglutide treatment of antipsychotic-treated patients with schizophrenia, prediabetes, and obesity: The HISTORI randomized clinical trial. JAMA Psychiatry. https://doi.org/[add DOI]
Gao, M., Kirk, M., Knight, H., et al. (2026). A ketogenic diet for treatment-resistant depression: A randomized clinical trial. JAMA Psychiatry. https://doi.org/[add DOI]
Mass General Brigham. (2026). McLean Hospital to establish new clinical and research center with $20 million investment. https://[add URL]
Sethi, S., Wakeham, D., Ketter, T., et al. (2024). Ketogenic diet intervention on metabolic and psychiatric health in bipolar and schizophrenia: A pilot trial. Psychiatry Research, 335, 115866. https://doi.org/10.1016/j.psychres.2024.115866
U.S. Food and Drug Administration. (2025). Latuda (lurasidone hydrochloride) [Prescribing information].
