Editorial Article
On 20 January 2026 Brussels Refused a Piglet Feed Additive Because Two Good Trials Is Not Three
On 20 January 2026 the European Commission adopted Implementing Regulation (EU) 2026/119. It runs to a few pages and does one thing: it denies authorisation of a preparation of carvacrol as a feed additive in the zootechnical category for weaned piglets. It is, on its face, a routine refusal about one applicant and one substance. Read alongside two earlier decisions, it is the third act of a twenty-three-year regulatory story, and it tells you how the fastest-growing category in animal nutrition is actually governed.
Carvacrol is a monoterpenoid phenol and the principal active in oregano essential oil. It is the archetypal botanical alternative to zinc oxide in weaned piglets — the product a European nutritionist reaches for when the medicine is no longer available. And the Commission refused it.
The reasoning is worth quoting precisely, because the numbers are the story. The European Food Safety Authority, in its opinion of 17 March 2020 and again in its follow-up opinion of 31 January 2024, could not conclude on the efficacy of the preparation of carvacrol, as only two long-term efficacy studies showing significant effects could be considered by the Authority, when Commission Regulation (EC) No 429/2008 and the Authority’s guidance required three long-term efficacy studies to show significant effects.
Two good trials. Not three. And two further studies that might have supplied the third were set aside, in the Commission’s words, due to high incidence of diarrhoea and high mortality in piglets. The applicant proposed removing some data from the analysis of one of those discarded studies; the Commission held that not duly justified.
Read that once more. The two trials that failed were excluded because the animals got sick. That is the field condition the product exists to address. A trial conducted in a commercial pig unit where disease and mortality happen is not a badly run trial; it is a realistic one. The evidentiary standard, applied literally, prefers a clean trial in a controlled unit over a messy one in a real one, and then refuses the product for not having three clean ones.
The full regulatory record, and the commercial consequences that follow from it, is set out in our report on the Global Feed Additives Market. This article follows the regulatory thread on its own.
Act one: the Union defines the problem out of the category
Regulation (EC) No 1831/2003 establishes the authorisation procedure for feed additives in the European Union. It divides them into technological, sensory, nutritional and zootechnical additives, plus coccidiostats and histomonostats. Only substances that have cleared the procedure may be placed on the market or used, and authorisations are granted for specific species and specific conditions of use.
The Commission’s own guidance on the legislation contains the sentence that reorganised the industry: antibiotics, other than coccidiostats or histomonostats, are not feed additives under European legislation. The product category that had built the modern feed additive industry was removed from the definition of the category. This is a stronger move than a ban. A banned substance can be reconsidered. A substance outside the category cannot be authorised back into it.
Recital 26 of the same regulation shows the legislator knew what it was doing. It required a date after which antibiotics still authorised as growth promoters would be forbidden, while allowing sufficient time for the development of alternative products to replace those antibiotics, and provided that no further antibiotics would be authorised as feed additives.
The Commission announced the completion of that process in July 2003, prohibiting the four remaining growth promoters — monensin sodium, salinomycin sodium, avilamycin and flavophospholipol — and the ban took effect from 1 January 2006. The substitution clock started then.
Act two: zinc, and a number that explains a whole category
Zinc oxide was used as a veterinary medicine to control post-weaning diarrhoea in piglets, typically at 100 mg per kg body weight per day for 14 consecutive days. The European Medicines Agency’s referral documentation equates that to 2,500 ppm zinc in feed. Separately, the nutritional maximum for zinc in complete feed for piglets was 150 mg/kg.
The ratio is 16.7 times. That is the size of the gap between what was legal as a nutrient and what was legal as a medicine — and it is the size of the hole that closing the medicinal route would leave.
The Committee for Medicinal Products for Veterinary Use concluded on 16 March 2017 that the overall benefit-risk balance for orally administered zinc oxide in pigs was negative, because the benefits in preventing diarrhoea did not outweigh the environmental risks from zinc accumulation, and because a risk of co-selection for antimicrobial resistance could not be excluded. The Commission adopted Implementing Decision C(2017) 4529 on 26 June 2017. Article 1 required Member States to withdraw the marketing authorisations, and permitted deferral where immediate withdrawal would cause adverse impact given the lack of available alternatives — with the deferral period not exceeding five years from the date of adoption. The medicinal route to 2,500 ppm closed by 26 June 2022. The nutritional ceiling of 150 mg/kg stayed.
Recital 4 of that decision is the most commercially consequential sentence in the entire record. It records the need to allow time to develop alternatives that are available and affordable and change pig farming practices without increasing the use of antimicrobials.
The regulator named the substitution market it was creating, and named the risk it was trying to avoid: that removing zinc would push producers back onto antibiotics. It allowed five years for alternatives to arrive. What arrived was a category worth, on our model, USD 6,900.0 million in 2025 growing to USD 13,400.0 million by 2033 at 8.65 per cent — the fastest-growing part of the feed additives market.
Act three: the alternative is measured against a bar it cannot clear
Then comes January 2026. The botanical alternative is refused, not because it is unsafe and not because it does not work in the field, but because the evidence standard requires three long-term efficacy trials with significant effects and the applicant could produce two that survived scrutiny.
The three acts form a closed loop. In 2003 the Union legislated a substitution market into existence and allowed time for alternative products. In 2017 it withdrew the medicine and again allowed five years for alternatives. In 2026 it refused the leading botanical alternative because the evidence bar could not be met in commercial conditions. The category created by regulation is now constrained by regulation, and the constraint is evidentiary rather than commercial.
One point of scope matters and should not be over-read. The refusal in 2026/119 denies the zootechnical claim for weaned piglets. It is not a prohibition of carvacrol, and it does not by itself remove the substance from every other use that other legislation governs. What it removes is the claim — and the claim is the thing that justified the price.
The other regulator chose the opposite lever and arrived at the same place
The United States did not ban anything. Its instrument was Guidance for Industry #213 and the Veterinary Feed Directive final rule, and the FDA’s own implementation record is unusually precise.
Of 292 new animal drug applications initially affected, 84 were completely withdrawn. Of the remaining 208, 93 oral products intended for use in water were converted from over-the-counter to prescription status and 115 products intended for use in feed were converted from over-the-counter to veterinary feed directive status. And production, including growth promotion, indications were withdrawn from all 31 applications that included such indications.
In January 2017 the agency reported that every affected application had either aligned with the guidance or had its approval voluntarily withdrawn. The molecules stayed. The indication that made them worth buying for growth went.
| European Union | United States | |
|---|---|---|
| Instrument | Regulation (EC) No 1831/2003 | Guidance for Industry #213 and the VFD final rule |
| Effective | 1 January 2006 | 1 January 2017 |
| What was removed | The substance — antibiotics defined out of the feed additive category | The claim — production and growth-promotion indications |
| Recorded reasoning | Recital 26 allowed time for alternative products to replace those antibiotics | Indications withdrawn from all 31 affected applications |
| Application count | Four remaining growth promoters prohibited in 2003 | 84 withdrawn; 115 to VFD; 93 to prescription |
| Commercial result | Gut-health substitution market created | Molecules retained, indication deleted |
What the two regimes have in common
Set side by side, the two largest markets look like opposites and behave identically. The European Union removed the molecule from the category. The United States removed the wording from the label. Neither removed the product from the farm.
What both removed was the claim, and the claim is the asset. That is the mechanism behind the central finding of our sizing work: enzymes and gut-health additives hold 26.0 per cent of the 2025 market and deliver 50.0 per cent of the growth, because demand for a claimed product is created by a prohibition and priced against a substitute that is no longer legal to buy.
It also explains an oddity that a purely commercial reading would miss. Methionine has nine suppliers worldwide and three of them hold almost 80 per cent of capacity; gut health has dozens of participants and no dominant supplier. Yet the fragmented category grows at 8.65 per cent against 3.43 per cent for amino acids, and on participants’ own reporting earns the better gross margin. The reason is that amino acids are priced against the soybean meal they replace, while a gut-health product is priced against nothing — the substitute is illegal. That pricing mechanism is examined in the companion article on how this market actually sets its prices.
Why the January 2026 refusal is a template, not an incident
The instinct is to read 2026/119 as one applicant’s bad luck. It is better read as the standard now applied to every zootechnical claim in the pipeline.
Three things follow. First, the fastest-growing category in this market now carries the highest regulatory risk, and the risk is evidentiary rather than commercial. Second, innovation in claim-based additives migrates to jurisdictions without a pre-market efficacy assessment in the European sense — which is why the gut-health category grows fastest outside the Union and carries a European hole in the middle. Third, the category’s pricing power is contingent rather than structural. If buyers shift from accepting supplier evidence to demanding independent verification, the mechanism stops working, not because demand falls but because the product becomes comparable.
What to watch is specific and cheap to monitor: the next three refusals, or the first successful re-authorisation of a botanical that previously failed. Either would reprice the category.
A note on what this article does not do
This article does not argue that gut-health products do not work. Many of them do, and the Commission’s refusal was of a claim on evidence, not a finding that the substance is ineffective in practice. Nor does it treat the European and American regimes as exhaustive: feed additive perimeters are set nationally, and we have not verified registration standards in other markets to the standard applied here, so we have not published a comparison we could not defend.
The regulatory thread runs through the whole category. It is set out in full, alongside the value chain, competitive landscape, investment activity and market risk — and with the three deliberate perimeter exclusions disclosed — in the report on the Global Feed Additives Market, which sizes the category at USD 43,500.0 million in 2025 rising to USD 62,900.0 million in 2033 at 4.72 per cent on compound feed tonnage growing only 1.20 per cent.
Frequently Asked Questions
What did the EU decide about carvacrol in January 2026, and why?
Commission Implementing Regulation (EU) 2026/119 of 20 January 2026 denies authorisation of a preparation of carvacrol as a feed additive in the zootechnical category for weaned piglets. EFSA's opinion of 17 March 2020 and its follow-up opinion of 31 January 2024 could not conclude on efficacy because only two long-term efficacy studies showing significant effects could be considered, when Regulation (EC) No 429/2008 and the Authority's guidance require three. Two further studies were set aside due to high diarrhoea incidence and high mortality in the piglets.
Does the refusal mean carvacrol can no longer be sold?
It denies the zootechnical claim for weaned piglets. It is not a prohibition of the substance, and it does not by itself remove carvacrol from every other use governed by other legislation. What it removes is the performance claim on the label, and the claim is what justified the product's price rather than its sale.
How much zinc was legal as a medicine compared with as a nutrient?
2,500 mg/kg as a veterinary medicine for post-weaning diarrhoea, which the European Medicines Agency's referral documentation equates to 100 mg per kg body weight per day for 14 days, against a nutritional maximum of 150 mg/kg in complete feed for piglets. The ratio is 16.7 times, and it is the size of the gap the 2022 withdrawal of the medicinal route left to be filled.
How do the EU and US approaches differ?
Different instruments. Regulation (EC) No 1831/2003 removed antibiotics from the feed additive category altogether from 1 January 2006. FDA Guidance for Industry #213 kept the molecules but withdrew production and growth-promotion indications from all 31 applications that carried them, moved 115 feed products to veterinary feed directive status and 93 water products to prescription, and saw 84 applications withdrawn outright. Both made the claim the regulated object, which is where the value sits.
