Can Two Papers Use the Same Participants Without Being Duplicate Publications?

If Two Papers Analyze the Same People, Are They Really Two Different Studies?

A research team recruits 1,000 participants and collects enough information to answer several questions. One paper examines the primary outcome. Another analyzes a different endpoint. A third investigates a mechanism that was not central to the first article.

All three papers contain the same participants. Does that make the later publications duplicates? No. Participant overlap alone does not determine duplicate publication. The important questions concern what each paper analyzes, what it contributes, and whether readers can see that the sample is shared.

The Same Participants Can Legitimately Appear in Multiple Papers

Participant Overlap and Publication Overlap Are Not the Same Thing

ICMJE explicitly allows multiple papers from the same participant sample

ICMJE's recommendations directly address this situation. For large trials, it notes that researchers may plan from the beginning to produce numerous separate publications addressing separate research questions while using the same original participant sample.

The recommendation emphasizes transparency rather than requiring every question involving those participants to appear in one enormous paper.

More broadly, ICMJE states that manuscripts based on the same dataset may warrant separate publication when they differ meaningfully in analytical methods or conclusions and add substantially to one another.

Duplicate publication is about substantial publication overlap, not participant identity

ICMJE defines duplicate publication as publication of a paper that overlaps substantially with one already published without clear, visible reference to the previous publication.

Notice what that definition does not say: it does not define duplicate publication as any article involving participants who have appeared in another paper.

The two can coexist, but one does not automatically establish the other.

One cohort may support many legitimate research questions

Longitudinal cohorts provide an obvious example. Researchers may follow thousands of participants for decades while collecting information on health, education, employment, family life, behavior, social conditions, and numerous outcomes.

One paper might examine educational attainment. Another could investigate cardiovascular outcomes. A third could examine mental health trajectories. Requiring every analysis involving the cohort to appear in the same article would make little scientific sense.

The participant overlap is extensive, perhaps complete, but the research questions can be genuinely distinct.

Trials can also generate planned secondary publications

A clinical trial may have one primary endpoint and several prespecified secondary outcomes or mechanistic questions. Separate publications can sometimes provide enough space and methodological detail to examine those questions properly.

ICMJE states that secondary analyses of clinical trial data should cite the primary publication, clearly identify themselves as containing secondary analyses or results, and use the same identifying trial registration number and persistent dataset identifier where applicable.

Those requirements help readers recognize that the articles derive from the same underlying trial.

The second paper should add substantially to the first

Shared participants are acceptable, but trivial differences between papers are not rescued by giving each manuscript a different title.

ICMJE states that manuscripts based on the same dataset should add substantially to one another to warrant consideration as separate papers. This is the more useful test.

A paper examining treatment effectiveness and another examining a theoretically important mediation mechanism may add substantially to one another. A second paper repeating the primary outcome with a slightly different set of covariates may have a much weaker claim to independent publication.

A different outcome can justify another paper, but not automatically

Suppose the same participants completed ten outcome measures. It would be possible to assign each outcome to a different manuscript, but that does not mean ten papers are scientifically justified.

Several closely related outcomes may belong to one coherent research question. Conversely, one outcome may represent a sufficiently distinct theoretical problem to warrant separate analysis.

The same reasoning applies when deciding how many papers one dataset can support. The substantive contribution matters more than the number of variables available.

The shared sample should be discoverable from the papers themselves

Readers should not need to notice that the recruitment dates, sample size, and baseline characteristics look suspiciously familiar before realizing that two articles analyze the same people.

ICMJE recommends appropriate citation of previous publications from the same dataset. For trial-based secondary analyses, it further recommends retaining the same trial registration number and dataset identifier.

Depending on the study and journal, the methods section may also explicitly state that participants were drawn from a previously described cohort, trial, survey, or project.

Why sample transparency matters for evidence synthesis

Participant overlap becomes particularly consequential in systematic reviews and meta-analyses. Suppose two articles each report an association using the same 800 participants, but neither makes the relationship obvious. A reviewer may incorrectly treat the studies as independent evidence from 1,600 people.

ICMJE warns that duplicate publication of original research can lead to inadvertent double-counting of data or inappropriate weighting of the results of one study.

Even when the papers are legitimate separate analyses, clearly identifying their shared sample helps evidence reviewers avoid treating correlated evidence as independent.

Partial participant overlap also deserves transparency

The samples do not have to be identical for overlap to matter. One paper may analyze the entire cohort, another a subset with complete biomarker data, and another participants who completed a follow-up assessment.

Those can be legitimate publications. Yet describing their relationship helps readers understand why sample sizes differ and prevents them from assuming that each article represents an entirely new recruitment effort.

Follow-up papers can use the same cohort at different time points

Longitudinal research naturally produces publications involving the same people at different stages. A baseline paper, two-year follow-up, and ten-year follow-up are not automatically duplicates merely because the cohort persists.

The later paper should make the longitudinal relationship clear and distinguish newly collected observations or analyses from findings already reported.

New research questions can emerge after the original study

Not every legitimate secondary analysis needs to have been the primary purpose of the original data collection. Existing data can support valuable new questions, provided the analysis is methodologically appropriate and consistent with ethical, consent, governance, and data-use requirements.

For shared clinical-trial datasets, ICMJE requires secondary users to explain completely how their analyses differ from previous analyses and to reference the data source using its persistent identifier.

That principle captures the broader expectation well: tell readers where the data came from and what is genuinely new about what you did with them.

Same participants plus same findings is a different situation

If the second paper uses the same participants, same exposure, same outcome, substantially the same analysis, and substantially the same conclusion as the first paper, the fact that some wording or formatting has changed offers little protection against a duplicate-publication concern.

The issue then shifts toward substantial republication of previously published work.

Same participants plus excessive fragmentation can become salami slicing

Even when each paper reports a technically different result, researchers should consider whether the distinctions justify separate publications.

Publishing one closely related outcome per paper from the same participants can become salami slicing if the fragments add too little independently and would be more informative when reported together.

Participant reuse is therefore not the problem. The scientific rationale and transparency of the resulting publications are.

Related papers should be disclosed during submission

ICMJE's submission recommendations call for authors to make a full statement to the editor about submissions and previous reports that might be regarded as redundant publication of the same or very similar work. Related work should be specifically referenced, and copies should be provided where appropriate so the editor can evaluate the relationship.

Disclosing companion papers does not make a legitimate analysis look suspicious. It gives the editor the information needed to see why the papers are different.

Three Papers, the Same 2,000 Participants, Different Contributions

Common Misunderstandings About Reusing Participants Across Papers

Reuse Participants Transparently, Not Invisibly

When planning another paper from the same participants, begin with the contribution rather than the sample. Ask what new question the paper answers and why readers need it separately from the publications that already exist.

A useful writing test is whether the methods section can state the relationship plainly: "Participants were drawn from the cohort previously described in..." If that sentence feels dangerous because it makes the papers look too similar, the solution is not to remove the sentence. It is to examine whether the new paper is sufficiently distinct.

Check Before Publishing Another Paper Using the Same Participants

Frequently Asked Questions About Using the Same Participants in Multiple Papers

The Same People Can Generate Different Evidence Without Becoming Different People

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.