What to take for blood sugar

Graded by the strength of the evidence — strongest first. Nothing here is sold to you.

Berberinegrade Blowers fasting glucose and HbA1c; GI side effects common
Myo-Inositolgrade Bimproves insulin sensitivity and ovulatory function in PCOS in multiple RCTs
Omega-3grade BMeta-analysis in type 2 diabetes: results for omega-3 on glucose control were mixed and, overall, no consistent improvement in glycemic control was demonstrated. The authors note the EPA/DHA ratio and early intervention may modify effects — a subgroup signal, not a general glucose benefit.
Magnesiumgrade Cmodest improvement in insulin sensitivity / fasting glucose, mainly in deficiency or diabetes
Ashwagandhagrade CSystematic review and meta-analysis of Withania somnifera (ashwagandha) in diabetes: pooled data suggest a possible glucose-lowering effect, but the authors conclude the anti-diabetic profile is not yet established and call for larger, longer randomized trials measuring blood glucose, HbA1c and insulin. Signal, not confirmation.
Creatinegrade CSystematic review and meta-analysis (9 studies): creatine supplementation had no significant overall effect on fasting blood glucose or insulin resistance. Some benefit appeared in the small number of studies in people with diabetes, so any effect is likely context-specific rather than general.
Ginkgograde CMeta-analysis: Ginkgo biloba extract significantly reduced fasting blood glucose and improved insulin resistance (HOMA-IR) versus control, plus reduced BMI and waist circumference, but did NOT significantly change HbA1c or fasting insulin levels — a partial biomarker signal, not comprehensive glycemic control.
Melatoningrade CMeta-analysis of 12 RCTs: melatonin supplementation significantly reduced fasting glucose and improved insulin sensitivity (QUICKI index), but had no significant effect on insulin levels, HOMA-IR, or HbA1c. A partial biomarker signal on fasting glucose specifically, not comprehensive glycemic control.
Vitamin Dstudied, no effectLarge randomized trial (D2d ancillary analysis) in older US adults: vitamin D supplementation did NOT significantly reduce incidence of type 2 diabetes in this trial alone. When pooled with prior trials in meta-analysis, a modest reduction was observed — so the evidence is suggestive at the pooled level but the single largest trial itself was a null result.
Glucosaminesafety-flaggedSystematic review (11 studies, 6 RCTs): several studies found DECREASED insulin sensitivity or increased fasting glucose in people taking glucosamine, particularly those with baseline impaired glucose tolerance or insulin resistance. This is a caution signal, not a benefit — people with pre-diabetes or diabetes should be aware glucosamine may worsen glucose metabolism in some studies.
NMNstudied, no effectReview of 8 human RCTs (342 mainly non-diabetic middle-age/older adults, 14 days to 12 weeks, 250-2000 mg/day): NMN supplementation showed no significant benefit on fasting glucose, fasting insulin, HbA1c, insulin resistance (HOMA-IR), or lipid profile. Short-term NMN supplementation did not meaningfully improve glucose control in this population.