The Sweetener That Waited Sixteen Years

Aspartame took sixteen years to move from a laboratory accident to the approval that finally stuck in the United States, and the delay was not caused by the chemistry. A chemist at G. D. Searle discovered the compound in 1965, the US Food and Drug Administration (FDA) approved it for dry goods in 1974, stayed that approval in 1975, and finally cleared it for dry products in 1981, over the objections of the independent inquiry set up to review it. The chemistry stayed the same. What changed, repeatedly, was the argument about how much evidence was enough, and who was entitled to decide 2.

That argument has never quite stopped. In July 2023 two bodies of the World Health Organization looked at the same sweetener and reached verdicts that sounded contradictory and were not. This piece traces that regulatory history from the documents, and it is about regulation, not a verdict on safety. It says what each body concluded and on what kind of evidence, and leaves the rest to the reader.

1965: an accident in an ulcer-drug programme

James Schlatter was a chemist at Searle, a pharmaceutical company in Illinois. In 1965 he was making fragments of gastrin, a peptide hormone that stimulates acid secretion in the stomach, as part of a search for a drug against ulcers. One of the intermediates was a small compound of two amino acids. As the story is usually told, and Schlatter himself told it, he licked a finger to pick up a sheet of paper and noticed an intensely sweet taste. He traced it to the intermediate on his bench.

Tasting laboratory chemicals is not something any safety officer would recommend, and the episode survives mostly because it happened to turn out well. What matters for this account is what the accident led to: a company with an ulcer-drug programme found that it held a candidate sweetener, and had to decide whether to pursue it. Searle did, and the sweetener became the company's most valuable product.

Two amino acids and a methyl group

Chemically, aspartame is L-aspartyl-L-phenylalanine methyl ester. Two amino acids, aspartic acid and phenylalanine, are joined by a peptide bond, and a methyl group caps the phenylalanine end. It is therefore a dipeptide, the smallest possible peptide, with a small modification. It is roughly two hundred times sweeter than table sugar by weight, which is why very little is needed.

A 1969 paper by Searle chemists in the Journal of the American Chemical Society, 'Structure-taste relationships of some dipeptides', set out the discovery in scientific terms: certain dipeptide esters taste sweet, and the sweetness depends on the amino acids and their spatial arrangement 1. In the body, aspartame is digested like other small peptides, to aspartic acid, phenylalanine and a small amount of methanol. The phenylalanine is the reason products containing it carry a warning for people with phenylketonuria, a rare inherited condition in which phenylalanine cannot be metabolised.

1974 approval, 1975 stay

Searle filed a food additive petition and submitted a body of safety studies to the FDA, and in 1974 the agency's Commissioner approved aspartame for use in dry foods. The decision did not hold. Objections were filed by John Olney, a neuroscientist who had studied the effects of glutamate on the developing brain, and by James Turner, a public-interest lawyer. They questioned the safety data and asked for a public hearing. In December 1975 the FDA stayed the approval, which prevented Searle from marketing the product until the objections were resolved.

Under the food additive provisions of the US law, the sponsor of an additive carries the burden of showing that it is safe, and the agency must evaluate the studies the sponsor provides. A stay is not a ruling that the additive is unsafe. It is a ruling that the question cannot be considered closed. In this case it also opened a second line of inquiry: not into aspartame itself, but into the studies Searle had relied on.

The audit and what it did to laboratory rules

In 1977 an FDA task force examined a set of Searle's studies, including some for aspartame. The reviewers found serious problems of record-keeping and quality control: poorly documented procedures, inconsistencies between raw data and reports, and lapses in how animals and tissues were handled. Subsequent review by a panel of independent pathologists concluded, according to later accounts, that the inconsistencies would not have changed the studies' conclusions, but the picture the audit gave of the laboratory's practices was damaging.

The FDA also asked the US Attorney's office in Chicago to consider whether the company had concealed or misrepresented data. No grand jury was convened, the statute of limitations on the alleged offences ran out, and the case was dropped in December 1977 for lack of evidence. The Searle investigation is commonly cited, alongside other cases of unreliable toxicology laboratories, among the events that led the FDA to adopt formal Good Laboratory Practice regulations in the late 1970s. A rule that governs how every safety study is documented is, in part, a legacy of this episode.

A Public Board of Inquiry says no; a Commissioner says yes

The objections had to be answered, and the route chosen was a Public Board of Inquiry, a panel of three independent scientists convened in the late 1970s to hear the evidence. The board heard Olney's claims that aspartame could damage the brain, including that of the foetus, and rejected them. It did, however, say that a question about a possible link with brain tumours needed further study, and on that basis it recommended against approval at that point.

The decision then went to the FDA Commissioner, Arthur Hull Hayes. He sought advice from a panel of FDA scientists and from agency lawyers, considered further data, including a Japanese study of tumours in rats, and in July 1981 approved aspartame for use in dry products. The approval is described in later accounts as one of the most contested decisions in the agency's history. A Commissioner overruling the board that his own agency had convened is unusual, and it supplied critics with an enduring grievance 2.

1987: Congress asks the GAO to check

Doubts did not subside, and in 1987 the General Accounting Office, now the Government Accountability Office, produced a report at the request of Congress, titled 'Food Additive Approval Process Followed for Aspartame'. The published summary describes its scope: the agency examined the scientific review conducted at the FDA's food safety centre of the issues raised about the aspartame studies, and how the agency was monitoring continuing safety concerns 2.

Secondary accounts summarise its finding as being that the approval protocol had been followed, together with a survey of scientists that showed a wide spread of confidence. The full report runs to more than a hundred pages, and the details should be read from the report itself rather than from summaries. The title is itself informative. A review that asks whether a process was followed answers a procedural question. It does not answer whether the process was good, or the evidence sufficient.

2013 and 2023: Europe re-evaluates, and two WHO bodies disagree in public

The European Food Safety Authority carried out a full re-evaluation of aspartame, which it published in 2013. The opinion reviewed the animal, human and mechanistic literature available at that point, concluded that aspartame and its breakdown products were safe for the general population at current exposure levels, and left the existing acceptable daily intake unchanged. It also noted that people with phenylketonuria must control their intake of phenylalanine from all sources 3.

In 2023 two assessments appeared within days of one another. The International Agency for Research on Cancer, meeting in June 2023, classified aspartame as Group 2B, possibly carcinogenic to humans. The basis was limited evidence for a type of liver cancer, hepatocellular carcinoma, in humans, together with limited evidence in experimental animals and limited mechanistic evidence. The Joint FAO/WHO Expert Committee on Food Additives, meeting between late June and early July, reaffirmed the acceptable daily intake it had set earlier, finding no sufficient reason to change it and describing the evidence of an association with cancer in humans as not convincing 4.

What a dipeptide sweetener teaches about how regulators decide

The aspartame record offers a few lessons that apply well beyond sweeteners. The first is that approval is a decision, not a measurement. The same body of studies produced a yes in 1974, a stay in 1975, a no from an advisory board in 1980, and a yes from the Commissioner in 1981, and each step can be defended in its own terms. Data quality can matter as much as data quantity, and the 1977 audit changed how studies are documented for every later sponsor.

The second lesson is about evidence tiers. The literature on aspartame includes animal toxicology, human observational studies, and mechanistic work, and the agencies weight them differently. A 2007 review of the toxicological and epidemiological literature reached a reassuring conclusion; as with any review of a commercially important substance, readers should check who funded and who wrote it before leaning on it 5. The 2023 verdicts show that even careful bodies, using stated methods, can emphasise different parts of the same record.

The third lesson is that a small molecule on a label can carry a surprisingly long institutional memory. A food additive that began as a sheet of paper and a licked finger has now been the subject of reviews in the United States, in Europe and at the World Health Organization, and each has left a document. Reading those documents in order is a way of seeing how regulatory certainty is built and revised.