Why The US And EU Disagree By A Factor Of Eight On Vitamin B6

What this label prints

The Gelagen Supplement Facts panel lists Vitamin B6 as pyridoxine HCl, 2 mg per two-gummy serving, 118 per cent of a day’s value. The percentage is calculated against the adult Recommended Dietary Allowance of 1.3 mg, and on that measure 2 mg is an unremarkable, moderately generous multivitamin dose.

The RDA is not the number this page is about. A Tolerable Upper Intake Level — the ceiling above which a nutrient starts causing measurable harm rather than doing good — exists for vitamin B6, and there are currently two different, both-current, both-official versions of it in circulation, set by two different regulatory bodies reviewing substantially the same underlying question. They do not agree. They do not agree by a wide margin. This page explains why.

One hundred divided by twelve is a little over eight. That is the factor in this page’s title, and it is arithmetic, not exaggeration.

How the US arrived at 100 mg, in 1998

The Food and Nutrition Board set the US upper limit as part of its 1998 Dietary Reference Intake report covering the B vitamins. The Board’s method for any upper limit is the same in outline: pick the health outcome most likely to occur first as intake rises — the “critical endpoint” — find the highest intake at which that outcome does not appear in the available data, and divide by an uncertainty factor to build in a safety margin.

For vitamin B6, the critical endpoint the Board chose was sensory neuropathy: nerve damage that shows up first as numbness, tingling or pain, typically in the hands and feet, and is the health effect most consistently linked to very high vitamin B6 intakes across the case reports available. Working from the human clinical data available at the time, the Board identified a No Observed Adverse Effect Level of 200 mg a day — the highest studied intake at which sensory neuropathy was not observed — and applied an uncertainty factor of 2 to it, arriving at 100 mg a day. One study from the 1980s reporting neuropathy at a lower dose was available but was judged too methodologically unreliable to serve as the basis for the limit, and was set aside rather than used.

That is the entire derivation: one human NOAEL, one modest twofold safety margin, and 100 mg a day as the result. It has stood essentially unchanged since 1998.

A third number the US and EU figures both sit above

The US and EFSA figures are not the only two ceilings ever published for vitamin B6, and the third one complicates the “which country is right” framing usefully. The United Kingdom’s own Committee on Toxicity reviewed vitamin B6 safety in 1995 and restated its conclusion in a November 1997 statement, a year before the US figure was finalised: after considering more than one hundred scientific papers, the Committee set a guidance level of just 10 mg a day from supplements, starting from a Lowest Observed Adverse Effect Level of 50 mg a day and dividing by a safety factor of 5.

Two details about that 1997 UK review make it more than a historical footnote. First, the Committee explicitly drew on animal toxicity data — a 1978 study in Beagle dogs — to support its human guidance level, the same category of evidence, in the same species, that EFSA would lean on a quarter-century later. Second, the Committee gave real weight to the 1987 Dalton and Dalton case-series data that the US Food and Nutrition Board considered the following year and set aside as too methodologically unreliable to anchor a limit. Two expert panels looked at overlapping evidence within about twelve months of each other and reached conclusions roughly ten times apart, which is itself a preview of exactly the kind of divergence this page is about.

The UK figure is lower than either the US or the EU number, and both of the figures this page is built around — 100 mg and 12 mg — sit above it. That ordering is worth holding onto going into the next section: it shows the low-dose, animal-data, wide-safety-margin approach is not something EFSA invented in 2023. It is closer to a return to a position UK regulators had already reached in 1997, using data the US had in hand at the same time and chose to treat differently.

How the EU arrived at 12 mg, in 2023

EFSA’s Panel on Nutrition, Novel Foods and Food Allergens revisited vitamin B6 in full and published its own scientific opinion on the upper limit in 2023, a quarter of a century after the US figure was set. The Panel was not working from the same starting point. A body of case reports and pharmacovigilance data had accumulated in the intervening decades describing sensory neuropathy at intakes considerably lower than the 200 mg human NOAEL the 1998 US review relied on — the Panel’s own review of nutrivigilance data recorded cases at daily intakes as low as 74 to 75 mg and at 100 mg, doses that sit at or below the entire US safety margin, not above it.

Faced with human data it judged too inconsistent in dosing and duration to anchor a precise new NOAEL, the Panel leaned instead on a controlled animal study: a subchronic toxicity study in Beagle dogs, which identified a Lowest Observed Adverse Effect Level of 50 mg per kilogram of body weight per day. Converting that to a human-equivalent dose and applying an uncertainty factor of 300 — a far larger margin than the US review’s factor of 2, reflecting both the cross-species extrapolation and the LOAEL-rather-than-NOAEL starting point — against a default 70 kg adult body weight produced a figure of 11.7 mg a day. A second line of reasoning drawn from the human data separately pointed to about 12.5 mg a day. EFSA took the midpoint of the two and rounded down, publishing a final upper limit of 12 mg a day for all adults, including during pregnancy and lactation.

Why three careful reviews landed so far apart

None of the three numbers is an error, and none of the three panels was careless. The spread is the visible result of three separate choices, each defensible on its own terms, stacking in the same direction across the UK, US and EU reviews:

• Different starting data. The 1998 US review had a human NOAEL of 200 mg and set aside the one lower-dose report it had access to as too unreliable to use. The 2023 EU review had a quarter-century more case and pharmacovigilance data, including reports of neuropathy at 74–75 mg, and judged that newer evidence sufficient to weigh seriously rather than exclude.
• A NOAEL versus a LOAEL. The US figure starts from a dose at which no effect was seen. The EU figure’s animal-derived anchor starts from a dose at which an effect was seen, which by design needs a larger downward adjustment to reach a level believed to be safe.
• A twofold margin versus a 300-fold margin. Even from similar starting points, an uncertainty factor of 2 and one of 300 produce wildly different endpoints. The size of that factor is itself a judgment call about how much the available data can be trusted to generalise to everyone, and the two bodies made different calls.

The honest summary is not that one agency is right and the other is wrong. It is that the US figure has not been revisited with the newer case-report evidence EFSA reviewed, and EFSA’s figure leans on an animal study and a large safety margin rather than a directly observed human ceiling. Both are legitimate ways to set a limit under genuine uncertainty, and they were never going to land on the same number once the uncertainty factors alone differ by a hundred and fifty times.

It is also worth being precise about what the gap is not evidence of. An eightfold difference between two ceilings does not mean the US figure has been shown to be unsafe, and it does not mean every product formulated around the higher US number is therefore risky. It means two different, defensible methodologies, applied to overlapping but not identical evidence, produced two different answers to the same question — and a reader comparing a label’s dose to “the” upper limit for vitamin B6 needs to know which one is being used, because the honest answer to “what is the limit” is currently “it depends which regulator you are asking.”

Where 2 mg sits against all three ceilings

This is the question a label actually needs answering, and it has one answer regardless of which upper limit is used. Gelagen’s 2 mg B6 row sits at 2 per cent of the US 100 mg ceiling, at about 17 per cent of the far stricter EFSA 12 mg ceiling, and at 20 per cent of the UK Committee on Toxicity’s 1997 guidance level, the strictest of the three. Under every regulator’s math on record, a single day’s serving is a small fraction of even the most conservative number any of them has ever published.

That comparison is arithmetic against a printed serving, not a statement that stacking several B6-containing products is automatically safe. Vitamin B6 shows up in multivitamins, in B-complex supplements and in some energy and pre-workout formulas, and whichever ceiling a reader considers the relevant one, adding several products’ B6 rows together, rather than reading one label in isolation, is the calculation that actually matters once more than one bottle is in the cabinet.

The short version: a single printed ceiling is not always a settled fact. Vitamin B6’s upper limit has two current, official, defensible answers eight times apart, and the reason is traceable — different data, a NOAEL against a LOAEL, and two very different uncertainty factors — rather than mysterious.

The other twelve rows on this panel are checked against their own reference intakes the same way. All thirteen amounts are set against their reference intakes here, and the ingredients page reproduces the panel in label order.

Sources