A paper published last June in Science that highlighted the dangers of microplastics on fish health has been retracted.

An investigation by Sweden’s Central Ethical Review Board found the study authors “guilty of scientific dishonesty” and hinted that some of the research described in the paper may not have been conducted, at least not to the extent described.

The paper, “Environmentally relevant concentrations of microplastic particles influence larval fish ecology,” was published June 3, 2016 and authored by Peter Eklöv and Oona Lönnstedt, both of Uppsala University.

In the original summary of the paper, the duo stated, “…exposure to environmentally relevant concentrations of microplastic polystyrene particles (90 micrometers) inhibits hatching, decreases growth rates, and alters feeding preferences and innate behaviors of European perch (Perca fluviatilis) larvae. Furthermore, individuals exposed to microplastics do not respond to olfactory threat cues, which greatly increases predator-induced mortality rates.”

The researchers, and many media reports covering the results of the paper, said that the larvae in the study actually preferred eating plastic over natural prey, comparing the act to teenagers eating junk food.

The abundance of media attention quickly prompted a group of researchers to speak out about “critical” flaws within the study, including missing data, inconsistencies in sample sizes, and “large disparities” in the way the experiment was reported compared with reports of eye witnesses.

An official complaint letter, signed by seven researchers on June 20, 2016, was sent to Uppsala University urging officials to conduct an investigation.

The complaint letter noted the group had asked Eklöv and Lönnstedt to provide the necessary data to reproduce their results, but they did not oblige. The Science paper stated that original data could be found in the supplementary material to the article, but it was not actually made available.

The group also sent a list of 20 questions regarding the methodologies used and results of the study.

In August 2016, just a couple months after initial publication, Uppsala University conducted a preliminary investigation but stated evidence was lacking to launch a full misconduct investigation.

However, the expert group from the ethical review board has now disagreed with the University’s conclusion. In a recently-released report, the board found Eklöv and Lönnstedt guilty of “scientific dishonesty.”

“The Expert Group considers the deficiencies in terms of original data to be of such severity in a research ethics perspective, that Peter Eklöv and Oona Lönnstedt, by their inability to present original data for the research, have been guilty of scientific dishonesty,” reads the report.

The expert group also stated it was “remarkable” that both the paper was accepted by Science, and that Uppsala University found no support for the presence of dishonesty in the research.

“The Expert Group notes initially that several questions were repeatedly put to Peter Eklöv and Oona Lönnstedt in order to resolve the question marks in this case. The answers received have been in all essentials deficient, at times contradictory and have not infrequently given rise to further questions,” wrote the panel.

Eklöv and Lönnstedt blamed the theft of a computer for the reason why they could not produce all original data from the study, but said data was saved on the university’s databank. A “glitch” in the database caused much of the information needed to defend their case to be lost.

The report highlights suspicions originally raised by the complaint letter authors of whether Lönnstedt was actually present at the research station during experiments to the extent that was reported. The June complaint letter referenced eyewitness accounts contradicting her reported tasks, and Lönnstedt was unable to provide supporting evidence documenting her tasks to the expert group.

Other questions about the work included the size of beakers used, how often the water was changed and if the microplastic particles used were pre-treated. The study found significant frequency of mortality in eggs exposed to the microparticles, but because the particles were not pre-treated, the mortality rates reported in the study could have been caused by the concentration of detergent in the particles.

In addition to concern of original data and methodologies used, questions on whether the researchers actually gained animal ethics committee approval were raised.

The panel made it clear that although Lönnstedt was responsible for the larger part of the alleged research, Eklöv is not free of responsibility with his role as senior researcher.

Uppsala University released a statement in response to the expert group’s decision, stating:

“The conclusions reached by the two reports differ, but they also proceed from different regulatory frameworks and cover different quantities of material, and the University will have to take this into account in its decision. However, the reports agree that the data was inadequately backed up.”

According to Nature, Eklöv and Lönnstedt still defend their work, and say they did indeed have an ethical permit to cover their experiments, and the stolen laptop and mishap with original data not being properly backed up “was an unfortunate mistake and not dishonesty.”

Microplastics have since been banned from use in cosmetics. Other researchers contend this retraction should not threaten to reverse the ban, as there is still a strong body of previous evidence showing harmful effects of microplastics on marine life.