PLOS One retracts Parkinson’s model paper after image and data integrity concerns
The PLOS One Editors have retracted a preclinical study on Parkinson’s disease after post-publication scrutiny identified problems in the paper’s tables, image material, and individual-level data. The article had examined the anti-neuroinflammatory and neuroprotective potential of Cissus tuberosa ethanol extract in a Parkinson’s disease model, but the record now carries a formal retraction after the journal assessed the supporting evidence and found it unreliable.
Concerns first centered on numerical patterns in Tables 6 and 7, where the reported standard errors of the mean appeared unusually repetitive. In animal and laboratory studies, standard errors can repeat by chance, yet repeated values across multiple outputs can also raise concern when the pattern looks more regular than expected from biological variation alone. In a preclinical neurodegeneration study, where small shifts in inflammatory markers or behavioral outcomes can change the interpretation of a treatment effect, numerical irregularities are not a minor editorial nuisance. They can compromise confidence in the entire experimental chain.
Reused panels and the problem of experimental provenance
Image review added a second layer of concern. Several panels in unpublished material submitted with the original manuscript appeared similar to panels used in other figures and in another publication, despite being presented as different experimental conditions. The similarities involved the C Heart panel in the unpublished material and Fig 2 CG:Heart in a later paper, the C Liver panel and Fig 2 TR Liver in that same later paper, and the CG panel in the unpublished material compared with a CG panel in a separate paper that was later retracted.
Those overlaps matter because image panels in biomedical research are not decorative. They function as evidence of experimental state, tissue response, or treatment effect. When a heart or liver panel is reused, the reader loses any assurance that the visual field represents the stated condition. In a Parkinson’s model paper, where the stated biological claim depends on modulation of neuroinflammatory markers, image provenance is part of the evidentiary structure. If the visual record cannot be trusted, the biological interpretation built on it collapses with it.
The corresponding author said errors occurred during figure preparation in the original version of the manuscript and were corrected in later revisions. The same author also stated that material from the later paper had been used without the data owner’s authorization. Those explanations address process, not reliability. They do not restore confidence in the figures as originally submitted, and they do not resolve the broader questions raised by the repeated panel similarities.
Individual-level data raised deeper reliability questions
The journal then reviewed the individual-level data underlying the graphs and tables. That assessment raised additional concerns about the integrity and reliability of the dataset itself. This step is critical. Figure-level problems can sometimes be traced to editing mistakes, file mix-ups, or poor archiving. Concerns in the raw data are more serious because they suggest the issues are not confined to presentation. They go to the foundation of the study’s results.
For a preclinical paper making claims about anti-neuroinflammatory and neuroprotective effects, the underlying data are the entire basis for any mechanistic claim. If the dataset is unstable, then the downstream conclusions about marker modulation, tissue protection, or disease attenuation are not dependable. The journal’s editorial review concluded that these concerns were sufficient to retract the article.
Why this retraction matters in Parkinson’s research
Parkinson’s disease research relies heavily on animal models that attempt to reproduce oxidative stress, inflammatory signaling, dopaminergic injury, and motor dysfunction. Natural products and plant-derived extracts are often explored in that setting because of their apparent anti-inflammatory or antioxidant properties. Those studies can be useful when they are carefully controlled, transparently reported, and reproducible. They can also mislead quickly when figures are reused, datasets are unstable, or published comparisons borrow material without permission.
The biology itself makes rigor essential. Neuroinflammation is not a vague background phenomenon in Parkinson’s disease; it is one of the mechanisms investigators often target to explain neuronal injury and possible neuroprotection. If an extract appears to shift inflammatory markers, the result must be supported by clean experimental provenance, consistent tables, and verifiable raw data. Otherwise, the field risks building mechanistic narratives on artifacts rather than on biology.
Author responses and editorial action
The corresponding author provided a response to the concerns. Other listed authors either did not respond directly or could not be reached. ZC and MA responded without clarifying agreement or disagreement with the retraction. MS, ASehar, RS, HT, SAB, and ASethi did not respond directly or could not be reached.
The editorial decision was straightforward once the data concerns were assessed in full. The editors retracted the paper in light of the problems observed with the individual-level data underlying the study. The retraction notice does not preserve the article as a reliable source of evidence for Cissus tuberosa in Parkinson’s disease models.
What readers should take from the record
This case is not about a disputed interpretation at the margins. It is about whether the experimental record can support any interpretation at all. Once image similarities appear across apparently different conditions and the raw data themselves raise integrity concerns, confidence in the manuscript falls apart.
Researchers working in preclinical neurology should treat this retraction as a reminder to safeguard figure provenance, archive source files, and maintain traceable raw datasets from the outset. Reviewers and editors, meanwhile, are left with a familiar but consequential lesson: in a model of neurodegeneration, the quality of the evidence must be as carefully defended as the hypothesis it is meant to test. Here, that evidentiary standard was not met.