Plant compounds

Berberine At 15 mg Against A 600 mg Study Dose

Berberine has one of the larger human trial records of any compound on this panel, in both directions: benefits reported in diabetes and cholesterol trials, and a documented interaction with a transplant drug. This article puts the 15 mg on the label beside the amounts those trials actually gave.

The row as printed

The panel line reads Berberine HCL, 15 mg. It is one of the simplest rows on the label to read, because it is a plain weight of a single compound salt with no percentage to apply and no ratio to interpret. The hydrochloride is a common form of berberine, and the 15 mg is the weight of that salt. There is no arithmetic to do before the number can be compared with a trial.

That simplicity is the reason the row is worth an article. When a row is easy to read, the comparison with the research is easy to make, and the result is clear. Berberine is also a compound with a large human literature, so the comparison is not empty in the way it is for a rarely studied plant.

The diabetes trials, pooled

Guo 2021 in Oxidative Medicine and Cellular Longevity is a systematic review and meta-analysis of randomised trials of berberine in people with type 2 diabetes. It searched eight databases, four of them Chinese, and assessed 46 trials, in which berberine was given alone or with standard diabetic therapy against a control.

The pooled results were favourable on every headline measure. HbA1c fell by 0.73, fasting plasma glucose by 0.86 and two-hour postprandial glucose by 1.26, with lower fasting insulin, a lower insulin resistance index and a lower BMI, by 1.07. Triglycerides, total cholesterol and LDL fell and HDL rose. The authors conclude there is strong evidence supporting berberine's clinical efficacy and safety in type 2 diabetes.

Three qualifiers belong beside that conclusion. The trials included people who were also on standard diabetic treatment, so berberine was often an addition to a drug, not a replacement for one. The abstract does not state the doses the trials gave, and a reader should not assume a figure it does not print. And every participant had type 2 diabetes, which is the population the effect was measured in.

The cholesterol trials, pooled

Blais 2023 in Drugs looked only at placebo-controlled trials in adults, and only at lipids. It found eighteen studies with 1,788 participants, treatment lasting from four to twenty-four weeks, and fifteen of the eighteen conducted in mainland China and Hong Kong.

Berberine reduced LDL cholesterol by 0.46 mmol/L, total cholesterol by 0.48 mmol/L, triglycerides by 0.34 mmol/L and apolipoprotein B by 0.25 g/L. HDL rose by 0.06 mmol/L overall, but the effect was different by sex: 0.11 in women and −0.07 in men, a difference the authors flag as a sex-specific finding worth further study. No serious adverse events were reported, and the authors summarise the effect as small reductions in LDL, triglycerides and apolipoprotein B.

The paper also reports the cost side. Gastrointestinal adverse events were reported in twelve studies and tended to be more frequent with berberine than placebo, at 2 to 23% against 2 to 15%. That is the ordinary price of the ingredient in the doses the trials used, and it is why the next sections care about how much was given.

Who those trials were done in

The two pooled reviews share a feature that matters for a capsule sold to healthy adults. They are trials in patients: people with type 2 diabetes in the first, adults with dyslipidaemia in the second. A reduction in HbA1c in someone whose HbA1c is raised does not describe what the same compound does in someone whose HbA1c is normal, because there is little to reduce.

The second feature is geography and duration. Most of the lipid trials were run in one region and lasted from a month to six months. Neither is a fault, and both are reasons the results should be read as findings about those populations rather than as a general property of the compound.

A third feature is that the reviews report averages. An average across 46 trials describes the group, and the spread inside it is not something a reader of an abstract can see. Some participants moved a great deal on berberine and some did not move at all, which is true of most treatments and is the reason a pooled figure is a statement about a population rather than a prediction for a person. A buyer reading a single mean difference should hold it with that in mind.

The cyclosporin study

The interaction literature is where a single small study says the most, and it should be read for its dose. Xin 2006 in Methods and Findings in Experimental and Clinical Pharmacology examined the effect of berberine on the pharmacokinetics of cyclosporin A, an immunosuppressant given to transplant recipients, in healthy male volunteers.

Six volunteers took 0.3 g of berberine twice daily for ten days, and the drug was measured before and at the end. With cyclosporin at 6 mg/kg, berberine caused no significant change. In a second group given cyclosporin at 3 mg/kg followed by a single 0.3 g dose of berberine, the area under the blood concentration curve rose on average by 19.2% and the twelve-hour level from 104 to 123 µg/l. The authors suggest the increase may be partly due to inhibition of the enzyme CYP3A4 in the liver and gut wall, and possibly to slower emptying of the stomach and small intestine, and they say the speculation needs further investigation.

Two lessons come from the dose. First, the study gives the only berberine dose this article can state from a source it has read: 0.3 g twice a day, or 600 mg a day, which is forty times 15 mg. Second, the effect was not uniform: it was absent after ten days of 600 mg a day with the higher cyclosporin dose and about a fifth after a single 0.3 g with the lower one, which is the reason a prescriber wants to hear about a supplement rather than assume it makes no difference.

The label asks anyone taking a medicine to consult a physician before use, and that sentence has a specific meaning here. The suggested use and the caution are printed in full on the panel page.

What the tolerability data show

The tolerability picture is the same shape as the benefit picture: it comes from trials at trial doses. The gastrointestinal events in the lipid review, 2 to 23% against 2 to 15% for placebo, are the figure a buyer can weigh, and the absence of serious events across sixteen reporting studies is the reassuring half of it.

What neither review can say is what a fortieth of the studied dose does to the gut, because no trial reported here gave it. It would be reasonable to expect fewer of the effects that scale with dose and reasonable to expect fewer of the benefits, and this desk offers both as expectations, not findings. Effects rarely fall in a straight line, and in either direction an assumption is not a result.

The arithmetic of a fortieth

The only stated berberine dose in the sources read for this article is the 0.3 g twice daily of the cyclosporin study. Dividing gives 600 mg a day against a printed 15 mg, a ratio of forty to one. It is a comparison of amounts, made once, with the source named.

It is worth saying what the ratio is not. It is not a claim that trials of diabetes and cholesterol used 600 mg, because the abstracts of those reviews do not state their doses, and this desk will not fill in a number from memory. It is not a claim that the capsule does nothing. It is a statement that the one dose this article can vouch for is forty times the dose on the label, and that a result at one amount is not automatically a result at another.

What each source gave

The table prints only what each source states. Where an abstract is silent on the dose, the cell says so.

What this desk will not say

• That 15 mg of berberine lowers blood sugar or cholesterol. Those trials were in patients and none of the doses printed in the sources read here is close to 15 mg.
• That 15 mg interacts with cyclosporin or any other drug. The only interaction study read here found no effect after ten days of 600 mg a day with one cyclosporin dose and a rise of about a fifth after a single 0.3 g with another.
• That 15 mg is free of interactions. No study read here tested the amount, and the label's advice to consult a physician about medicines stands.
• That berberine is unsafe. Sixteen studies reporting adverse events found none serious.
• That the row is decorative. It is printed as an amount of a compound with a real record, and that record is exactly why the gap is worth stating.

What can be said is in the table: one printed amount, three sources, and one stated dose forty times larger. The ingredients page reads every row the same way, and the eleven-amounts article shows where the berberine row sits among the rest. The neighbouring small rows are read the same way in the resveratrol article and the mangosteen and prickly pear article.