A case report proposes that muscle lost while dieting can push blood sugar back up
In short
This paper is a case-driven hypothesis rather than a trial. A 53-year-old man with type 2 diabetes lost weight from 84kg to 70kg on semaglutide, developed muscle weakness and reduced mobility, and ended with refractory hyperglycaemia — fasting glucose 300 mg/dL and HbA1c 9% — that persisted after the drug was stopped. Because skeletal muscle is the body's primary glucose sink, the authors propose that rapid muscle loss narrows that exit and creates a self-reinforcing loop, and they call for prospective study of muscle-preserving interventions such as resistance training. The percentage figures in the mechanism are magnitudes the authors posit, not measurements taken from the patient.
Start with what this paper is not. It is not a trial and not a cohort study. It takes one patient's course as a starting point and proposes a mechanism, and the authors put the word hypothesis in their own title. Nothing here is proven; a question worth testing has been put on the table.
What actually happened in the case?
A 53-year-old man, diagnosed with type 2 diabetes in 2014. Glycaemic control kept deteriorating through metformin, glipizide, sitagliptin and empagliflozin; he moved to dulaglutide for a year and a half, then to semaglutide. Over the following year his weight fell from 84kg to 70kg — accompanied by muscle weakness, reduced mobility, and refractory hyperglycaemia at fasting glucose 300 mg/dL and HbA1c 9%. That hyperglycaemia persisted after the drug was discontinued.
What loop are the authors proposing?
The hinge is that skeletal muscle is the tissue that absorbs most of the glucose in the body. The hypothesis is that rapid muscle loss narrows that exit, the raised glucose then makes conditions worse for muscle, and a muscle-glucose feedback loop forms. Layer amplified hepatic glucose production on top and the result is hyperglycaemia that standard drugs struggle to bring down.
The paper carries figures — muscle glucose uptake down 35–45%, anabolic signalling suppressed 25–35%, mitochondrial mass down 20–25%. These are magnitudes the authors posit for the proposed mechanism, not values measured in the patient. Anyone quoting them has to carry that condition along.
What supports it beyond the single case?
The authors assemble corroborating data. A 24-month study of 432 patients showed declines in appendicular skeletal muscle index and grip strength; a longitudinal analysis of 141 men showed rising markers of neuromuscular junction degradation; a secondary analysis of 51 MASLD cases showed 9.3% psoas volume loss. Case reports of fatigue in a 74-year-old and rhabdomyolysis in a 47-year-old are also cited. All of this aligns with the hypothesis; none of it tests it.
One case, one hypothesis, no established causation. Do not reduce or stop a prescribed medication on the strength of an article — uncontrolled blood glucose is dangerous in itself, and any adjustment belongs with the treating clinician. GLP-1 use and training performance in adults is covered in GLP-1 drugs and training performance, and exercise prescription for muscle loss in how well exercise works for sarcopenia.
If the hypothesis holds, the response is already known
The intervention the authors want tested prospectively is resistance training. What makes this easy is that the response is the same whether the hypothesis survives or not — how to protect muscle during weight loss is settled, and it comes down to keeping the load on and the protein in. If the hypothesis is wrong, muscle was preserved; if it is right, a metabolic problem was headed off with it.
Catching muscle loss in numbers instead of symptoms
In this case the muscle loss was found as a symptom — it surfaced only once strength had dropped and walking had become hard. A far earlier signal sits in a training log. During a cut, absolute Big 3 loads falling three to four weeks running is the first alarm that muscle is going. A bodyweight-adjusted strength score works against you here: because it corrects for body weight, it can rise even as strength slips slightly, simply because the scale went down. Watching only the score during a cut means the picture looks fine while muscle leaves. That is why the absolute loads go in the log too.
Frequently asked questions
Does semaglutide cause sarcopenia?
This paper cannot answer that. It is a case-driven hypothesis built from a single patient, and the authors present a proposal for prospective study rather than established causation — the title itself says case-driven hypothesis.
What happened in the case?
A 53-year-old man with type 2 diabetes lost weight from 84kg to 70kg on semaglutide, developed muscle weakness and reduced mobility, and ended with fasting glucose of 300 mg/dL and HbA1c of 9% that persisted after the medication was discontinued.
Can losing muscle raise blood sugar?
Theoretically yes, because skeletal muscle is the body's largest glucose sink. The hypothesis here is that muscle loss narrows that exit and the resulting hyperglycaemia in turn worsens conditions for muscle, forming a feedback loop. It remains a proposed mechanism, not established causation.
Are the percentages in the paper measured values?
No. Figures like a 35–45% reduction in glucose uptake or 25–35% suppression of anabolic signalling are magnitudes the authors posit for the proposed mechanism, not measurements taken from the patient. Quote them with that caveat attached.
What should someone on a GLP-1 drug do?
Medication decisions belong with the treating clinician, but the training side is clear: keep resistance training in, keep protein intake adequate, and log absolute Big 3 loads so the point at which strength starts slipping during weight loss does not go unnoticed.
Source: PubMed