A triple-blind randomised placebo-controlled trial at a single Iranian centre gave 50 adults with type 2 diabetes and confirmed diabetic peripheral neuropathy either 200 micrograms a day of organic selenium, as selenium-enriched yeast, or placebo for 8 weeks. Outcomes were fasting glucose, HbA1c, insulin resistance measured by the homeostatic model assessment, quality of life and sleep quality, analysed with adjustment for baseline values, age, sex, body mass index and neuropathy score.
HbA1c fell by an adjusted 1.20 percentage points more than placebo (95% CI -1.79 to -0.61), fasting glucose by 33.95 mg/dL (-64.2 to -5.5) and the insulin resistance index by 1.57 (-3.06 to -0.08). Quality of life and sleep did not differ between groups except on one sleep subscale. Adverse events were minimal in both arms.
A 1.2 point HbA1c fall is what you expect from starting a GLP-1 receptor agonist, not from a trace element over eight weeks. HbA1c reflects roughly the preceding two to three months of glycaemia, so an eight-week study cannot fully register the change it is reporting. The authors say as much themselves, flagging the modest sample, short follow-up and single-centre design.
Do not start prescribing selenium on this. Larger trials of antioxidant supplements in diabetes have repeatedly failed to reproduce promising small-trial results, and selenium has a narrow therapeutic window: higher intakes have been linked to increased type 2 diabetes risk in other cohorts. If a patient is already taking a supplement, ask the dose, because 200 micrograms a day sits well up the range once dietary selenium is added on top.
- Do not recommend selenium for glycaemic control on this evidence.
- Ask specifically about supplements at every review; patients often do not count them as medicines.
- The tolerable upper intake for selenium in adults is 400 micrograms a day from all sources: check the total, not just the tablet.
- For painful diabetic neuropathy the evidence-based steps remain glycaemic optimisation and a trial of duloxetine, pregabalin, gabapentin or amitriptyline.
- Recheck HbA1c no sooner than three months after any change; anything earlier is a partial reading.
The statistics, in plain English
The confidence interval for the fasting glucose effect runs from -64.2 to -5.5 mg/dL: the lower end is a dramatic effect, the upper end is barely anything, and a spread that wide is what 50 patients buys you. Small trials do not merely give imprecise answers, they systematically overstate effects, because a modest true effect only crosses the significance threshold in a small sample when chance happens to exaggerate it. The insulin resistance interval, -3.06 to -0.08, almost touches zero, which makes that finding marginal rather than solid. And HbA1c integrates about 8 to 12 weeks of glucose, so an 8-week endpoint measures a change that is still in progress.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for diabetes & endocrinology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free