Berberine, cinnamon, banaba and olive leaf: the blood sugar supplement ledger
In a cafeteria you compare suppliers on a spreadsheet: price per case, delivery days, how often they short you. I wanted the same kind of sheet for the blood sugar ingredients that keep showing up in supplements, including the memory supplements that now lean on them. So I made one.
Every row comes from a published human trial. Where the evidence is only in animals, I say so.
The ledger
| Ingredient | Dose studied | Who was studied | What changed | Main caution |
|---|---|---|---|---|
| Berberine | 500 mg, 3 times a day | 36 adults, new type 2 diabetes, 3 months | A1c fell from 9.5% to 7.5%, similar to metformin | Stomach upset in about 1 in 3; interacts with many drugs |
| Ceylon cinnamon | 120 mg to 6 g a day | 543 people with type 2 diabetes, 10 trials | Fasting glucose down about 25 mg/dL; A1c not clearly changed | Adds to diabetes medication |
| Banaba (corosolic acid) | 10 mg, single dose | 31 adults | Lower blood sugar 60 to 120 minutes after a glucose drink | Can cause lows with medication |
| Olive leaf | about 50 mg oleuropein a day | 46 overweight men, 12 weeks | Insulin sensitivity up about 15% | May lower blood pressure a little |
| Kudzu flower | 200 to 300 mg a day | 81 adults with obesity, 12 weeks | Less belly fat; blood sugar in lab studies only | Acts a little like estrogen |
Berberine
The heavyweight of the group. In the best-known trial, adults with newly diagnosed type 2 diabetes took berberine or metformin, and their A1c dropped by about the same amount over three months (Yin, 2008). That is a real effect, not a rounding error. About a third of people had stomach trouble, mostly temporary.
The catch: berberine slows down liver enzymes that clear many prescription drugs. It is also unsafe in pregnancy. If you take anything regularly, this is the ingredient to ask your pharmacist about.
Ceylon cinnamon
Cinnamon has been tested many times, mostly in small trials. Pooled together, ten trials with 543 people showed fasting blood sugar going down by roughly 25 mg/dL, but no significant change in A1c, the longer-term number (Allen, 2013). Doses were all over the place, from a pinch to six grams a day.
Ceylon is the right kind for a supplement. The cheap cassia cinnamon in most grocery stores has much more coumarin, which can strain the liver in large amounts.
Banaba leaf
The extract is sold as corosolic acid. In a small Japanese study, 31 people took either 10 mg of it or a placebo just before a sugary test drink, and blood sugar was lower over the next two hours with the extract (Fukushima, 2006). One dose, one afternoon. Promising, not proven for daily use.
Olive leaf
A New Zealand trial gave 46 overweight middle-aged men olive leaf extract or placebo for 12 weeks each, swapping halfway, and found insulin sensitivity improved about 15% with the extract (de Bock, 2013). The extract maker supplied the capsules, which is worth noting, but the design was solid.
Kudzu flower
The only human trial I found was about belly fat: 81 Japanese adults with obesity lost a little visceral fat over 12 weeks. Its blood sugar effects so far are from lab work. Kudzu isoflavones act a little like estrogen, which matters for anyone with a hormone-sensitive condition.
What happens when they are combined
This is where the ledger turns from a list into a warning. Several memory supplements on the market now combine four or five of these ingredients in one tablet. In a healthy person, that might mean a gentle push on blood sugar. In someone already taking metformin or insulin, the pushes add up, and the risk is a low blood sugar episode. The research I cover on the first page of this ledger found that lows themselves are linked to higher dementia risk, so this is not a small detail.
If you are weighing a combination product, I found it helpful to see doses and studies for each Phytomem One ingredient laid out together, since that product is a good example of the trend. And the specific issue of stacking blood sugar ingredients with metformin is covered there with each interaction listed.
The NCCIH page on diabetes and dietary supplements makes the official point plainly: supplements should not replace diabetes treatment, and you should tell your health care providers about anything you take.
How I read a supplement trial
After a year of reading these studies, I have a short checklist I run through before I believe a headline. You can use it on any supplement, not just these five.
How many people? Thirty people is a start. Three hundred is a study. Three thousand is evidence. Most supplement trials are in the first group.
For how long? A single afternoon, like the banaba study, tells you about one meal. Twelve weeks tells you more. Blood sugar changes that matter for the brain play out over years.
Compared with what? A placebo is the minimum. A real drug, like the berberine-versus-metformin trial, is better.
Who were the people? Results in people with diabetes do not automatically apply to people with prediabetes, and results in overweight middle-aged men do not automatically apply to a 70-year-old woman.
Who paid? Companies fund a lot of supplement research. That does not make it wrong, but it is worth noticing, and a good paper says so in the fine print.
What exactly was measured? "Improved insulin sensitivity" and "prevented dementia" are very different claims. Almost all of the research in this ledger measures the first kind of thing.
Run the five rows of my table through that checklist and you will see why I rank berberine first and kudzu last.
The combination I would be most careful with
Berberine plus banaba plus cinnamon, on top of a prescription diabetes drug. Each one alone pushes blood sugar down modestly. Together, in a tablet whose label does not list the amounts, there is no way to estimate the total push in advance. That is the situation where I would insist on home glucose checks for the first couple of weeks, and where I would ask the doctor whether the medication dose needs watching.
My own ledger entry
I do not take any of these. My numbers came down with food and walking, and I did not want to muddy the picture while I was tracking it. If my A1c starts climbing again, berberine is the one I would ask my doctor about first, because it has the most evidence. But I would ask first, and I would bring this table.