What Penn Medicine's 'food noise' case study actually showed
Casey Halpern's Nature Medicine case study recorded nucleus-accumbens activity in a single patient on tirzepatide. Activity was suppressed early, then returned at 5 months alongside renewed cravings.
Why we wrote this. When a credentialed researcher cautions against 'miracle drug' framing on the record, that's load-bearing context for any reader following the obesity coverage right now.
On 17 November 2025 Penn Medicine published a brief covering a Nature Medicine case study by Casey H. Halpern, MD, Kelly Allison, PhD, and Wonkyung Choi, which used implanted intracranial electrodes to record nucleus-accumbens activity in a single patient on tirzepatide. The case is unusual because direct human nucleus-accumbens recordings during pharmacotherapy are rare; most of what we know about how GLP-1 and GIP agonists affect appetite circuits comes from rodent work or imaging.
What the recordings showed
Early in tirzepatide treatment, nucleus-accumbens activity associated with food preoccupation was suppressed. The patient reported resolution of what the field has started calling "food noise," the intrusive thought pattern around food that GLP-1 class drugs are widely credited with quieting. At five months, the NAc activity returned, and the patient's food preoccupation returned alongside it.
What the team said
Halpern, in the Penn Medicine brief, framed the finding plainly: "This study offers major insights into how these drugs may work inside the brain and will guide us as we explore new indications." His caution to the wider conversation was sharper still:
Until we better understand their action on the brain, it's far too soon to call GLP-1 and GIP inhibitors miracle drugs for more conditions beyond type 2 diabetes and obesity.
Allison added context on what the recordings did and did not show: the drugs are "amazing medications at doing what they were developed for," but the case study suggests the food-craving suppression mechanism may not be durable over time.
Why this matters
Two reasons. First, the case study is a counterweight to the broader "miracle drug" framing that has accompanied the GLP-1 obesity literature, and one delivered by clinicians whose research interests overlap with the same therapeutic space (Halpern's lab works on deep-brain stimulation for obesity, a competing intervention). The caution is on-record from someone with skin in the broader game.
Second, the durability question. The headline efficacy of tirzepatide in trials like SURMOUNT-1 is set by 72-week endpoints. SURMOUNT-4, the discontinuation trial, showed substantial weight regain when participants switched to placebo at week 36. The Halpern recordings hint at a mechanism for the regain: the food-craving circuit re-engages even while the patient remains on therapy. If reproduced in larger studies, that finding shapes how clinicians and patients think about long-term use, particularly when combined with the recent TRIUMPH-1 readout for retatrutide at 28.3% mean weight loss at 80 weeks.
What this is not
A reason to stop treatment. A reproduced finding. A conclusion about most patients (n=1 is the literal sample size). Or evidence that tirzepatide's weight effect is short-lived; the case study tracks craving suppression specifically, not body-weight outcomes. The pharmacology is doing real work on weight even when the food-noise effect fades.
What it is: a high-quality methodological piece, named experts on record with a cautious framing, and the kind of mechanistic depth that hardens the long-horizon question about chronic GLP-1 class use.
Frequently asked
What is 'food noise'?
An informal term, increasingly used in the GLP-1 literature, for intrusive thought patterns around food: planning meals, thinking about snacks, mental preoccupation with eating. Patients on tirzepatide and semaglutide commonly report it quieting. The Halpern case study connected the subjective experience to nucleus-accumbens activity in one patient.
Does this mean tirzepatide stops working at 5 months?
No. The case study tracks one specific brain-circuit signal (food-noise suppression) in one patient, not body-weight or HbA1c outcomes. The pharmacological effect on weight in larger trials continues well past 5 months. The finding is that one specific mechanism may attenuate, not that the drug stops working.
Is this a Penn Medicine endorsement of stopping treatment?
No. Halpern's caution is against overreach in extending GLP-1 class indications beyond type-2 diabetes and obesity without more mechanistic detail. He is not advising patients to stop their prescribed medicine. Any decision about continuing or discontinuing belongs with a clinician who knows the individual case.