Writing the Acknowledgements Section of a Medical Manuscript: What Journals Now Require in 2026

For most of academic publishing history, acknowledgements were simple. You thanked the statistician who ran the model, the research coordinator who managed the database, and the funding body whose grant made the study possible. Two or three sentences, done. Journals neither policed the section carefully nor required much from it beyond basic completeness.

That simplicity is gone. The acknowledgements section of a medical manuscript now carries at least four distinct obligations that did not exist a decade ago: a disclosure of any AI tool used during manuscript preparation, an acknowledgement of professional medical writing support that meets standards set by AMWA, EMWA, and ISMPP, a patient and public involvement statement that a growing number of journals require regardless of whether any involvement occurred, and funder-specific language that satisfies the terms of grant agreements and open-access mandates. Each of these has its own vocabulary, its own placement logic, and its own failure mode that triggers an editorial query or delays acceptance.

None of this is particularly hard if you plan for it. The problem is that most authors treat acknowledgements as an afterthought, composing them in the hour before submission when the team is exhausted and ready to move on. That is when details slip, disclosure language gets vague, and funders end up named wrong. This guide is meant to be read before the paper is finished, not after.

What Acknowledgements Are For: The ICMJE Framework

The ICMJE recommendations define acknowledgements as the place for recognizing contributors who do not meet the four criteria for authorship. The four criteria are: substantial contribution to conception or design, or acquisition, analysis, or interpretation of data; drafting or critically revising the manuscript for important intellectual content; final approval of the version to be published; and agreement to be accountable for all aspects of the work. Any person who meets all four is an author. Any person who meets some but not all belongs in acknowledgements.

In practice, this covers a wider range of contributions than authors usually consider at the time of writing. A laboratory technician who ran the assays but had no role in designing the study or interpreting the results. A clinical research coordinator who recruited and consented participants. A biostatistician who provided code review but whose intellectual contribution was limited to checking the analysis the authors had specified. A colleague who read and commented on a late draft but did not revise it substantially. All of these people contributed to the work. None of them qualifies as an author unless their contribution was broader than it sounds here. All of them should be named in acknowledgements.

The ICMJE also specifies that acknowledged individuals should consent to being named, because acknowledgements carry an implied endorsement. In practice, most authors send a courtesy email or message. The main risk of skipping this step is acknowledging someone who disagrees with the paper's conclusions and who might later need to be removed, which is an awkward editorial correction to request.

Professional Medical Writers: What Must Be Disclosed and How

Professional medical writers are writers, editors, and publication specialists who assist sponsors, academic groups, and individual researchers in preparing manuscripts, abstracts, and presentations. They are common in industry-sponsored clinical trials, increasingly present in academic medicine, and often hired by authors whose first language is not English. Their use is completely legitimate. The obligation is to acknowledge them clearly.

The joint position statement from AMWA, EMWA, and ISMPP specifies that any professional medical writing assistance should be acknowledged in the acknowledgements section with the name of the writer, their professional affiliation or employer, and the specific nature of their support. The statement should also confirm who funded the writing support. A standard phrasing recommended by these organizations runs roughly as follows: "Medical writing support was provided by [Name], [Company/Affiliation], and funded by [Sponsor Name]."

This seems straightforward but generates some common errors. Authors sometimes omit the funding source for writing support, either because they find it awkward to name a pharmaceutical company in that context, or because they genuinely forgot who paid for the service. Both versions are problems. Failing to disclose the funding source for writing assistance is the specific error that tends to attract scrutiny during post-publication integrity reviews, because it obscures the relationship between sponsor and author. The disclosure is not about stigmatizing the use of writing support. It is about making the sponsor's role visible.

A professional medical writer who contributed substantially to the intellectual content of the paper, not merely to its clarity and structure, may qualify as an author rather than an acknowledged contributor. The same four ICMJE criteria apply. In practice, this is uncommon for professional medical writers, whose role is usually to take content the research team has generated and shape it into a publishable manuscript. But it is a distinction worth thinking through if you are using someone who has been involved from the study design phase.

AI Use: What Goes in Acknowledgements and What Goes in Methods

The placement of AI disclosures between acknowledgements and methods is one of the more genuinely confusing requirements in the current landscape, and different publishers have settled on different answers. Understanding the logic behind the distinction helps authors navigate journals that have not spelled it out explicitly.

The working principle, which both ICMJE and the major publishers endorse, is this: AI use that affected the research itself goes in methods. AI use that affected only the writing of the paper goes in acknowledgements or a dedicated AI disclosure statement. If you used a large language model to help classify qualitative interview responses, code adverse event data, or extract information from a dataset, that belongs in the methods section as part of the description of how you collected or processed your data. It affected what you found and how. Reviewers need to see it to evaluate your methods. If you used the same kind of model to improve the readability of your introduction or to rephrase the discussion section, that belongs in the acknowledgements or the dedicated AI disclosure statement. It affected how you wrote, not what you found.

Translation assistance occupies an interesting middle position. If you used an AI tool to translate your manuscript from the language in which you drafted it into English, most journals treat this as a writing-level disclosure that belongs in acknowledgements. If you used AI translation to translate instruments, patient-facing materials, or original source documents as part of the data collection or analysis process, that may belong in methods as well.

Some journals now include a dedicated field in their online submission system specifically for AI disclosure, separate from the acknowledgements text. Whatever you enter there should be consistent with what appears in the manuscript. An AI disclosure in the submission system that does not match the acknowledgements in the paper creates a discrepancy that reviewers and editors sometimes flag.

Funding Statements: What Grantors Require in 2026

Most major research funders now specify the language they require in the acknowledgements of papers arising from their funding. Deviating from that language, even innocently, can create compliance problems for your grant and, in some cases, trigger a correction request from the journal after publication when the funder notices.

NIH requires that acknowledgement language include the specific grant number, the name of the granting institute or center, and language confirming that the content is solely the responsibility of the authors and does not necessarily represent the official views of NIH. The exact formulation is available in the NIH grants policy statement and varies slightly between programs. For papers subject to the 2024 NIH Public Access Policy, the acknowledgements should also confirm the NIH manuscript submission identifier (NIHMSID) once it has been obtained, though some journals manage this on the authors' behalf.

Wellcome Trust, the UK Research and Innovation bodies, and most European Research Council grants have similarly specific acknowledgement templates. Authors working under Wellcome grants should check whether their grant was funded before or after January 2025, when Wellcome's hybrid APC policy changed, because the required acknowledgement language differs between the two cohorts. ERC grants require a specific disclaimer about the European Union's responsibility (or absence of it) for the research outcomes.

Industry-funded studies have their own conventions. Pharmaceutical and device company sponsors typically specify acknowledgement language in the publication agreement, and following their preferred language is both contractually appropriate and a clear signal to readers about the nature of the funding relationship. Some journals additionally require that industry-funded studies include the study registration number in the acknowledgements or the abstract, even if it also appears in the methods.

Patient and Public Involvement Statements

The BMJ began requiring a patient and public involvement statement in 2015. BMJ Open followed. By 2026, this requirement has spread to a significant number of journals in the UK, Canada, and Northern Europe that publish clinical research, particularly those aligned with National Institute for Health and Care Research funding in the UK, where patient involvement in research design is increasingly expected rather than optional.

The statement asks authors to describe whether patients or members of the public were involved in the design, conduct, or reporting of the research, and if so, how. This does not mean whether patients were enrolled as study participants. It means whether patients as advisors, co-investigators, or design contributors had any role in shaping the study. If the answer is no, the statement says so directly, and that is entirely acceptable at most journals. The requirement is for the statement to exist, not for the involvement to have occurred.

A simple no-involvement statement would read: "Patients or members of the public were not involved in the design, conduct, or reporting of this research." If there was involvement, the statement should describe what form it took: whether patients reviewed the study protocol, co-designed the outcome measures, contributed to interpretation of results, or helped develop the plain language summary. The GRIPP2 short-form reporting checklist offers a structured template for more extensive involvement that some journals prefer.

Many authors working in basic science or with purely retrospective administrative data skip this statement because they assume it does not apply to their work. Check your target journal before deciding. If the journal asks for it and you have not included it, you will receive an editorial query. The statement takes one sentence to write if there was no involvement, and it is better to include it proactively than to respond to a request during revisions.

Statistical Support and Other Technical Acknowledgements

There is genuine ambiguity about when a biostatistician is an author and when they are acknowledged. The ICMJE criteria apply the same way here as anywhere else: if the statistician contributed substantially to the study design or interpretation, revised the manuscript, approved the final version, and accepts accountability, they qualify as an author. If they ran specified analyses, checked code, or reviewed a statistical methods section without significant intellectual contribution to the direction of the work, acknowledgement is appropriate.

In practice, many research teams have productive statistical collaborations that fall somewhere in the middle, and the decision is sometimes made by convention rather than by careful application of ICMJE criteria. The most common error is listing a statistician as an author on a paper where they have not reviewed or approved the final version and would not agree to be accountable for the clinical claims it makes. When in doubt, acknowledge rather than author, and ask the statistician which they prefer.

Similar logic applies to bioinformaticians, data engineers, laboratory managers, and clinical research associates. If a person ran analyses, built pipelines, or managed the data under direction from the investigative team without independent intellectual contribution to the research questions or their interpretation, their contribution goes in acknowledgements. They should be named individually with a brief description of their role rather than listed collectively (for example, "the data management team at [Institution]"), which most journals now prefer.

Journal-Specific Variations That Catch Authors Off Guard

The ICMJE framework establishes the floor. Individual journals build upward from there, and the variation is genuinely wide. Authors who submit to multiple journals across a career can fall into the trap of assuming the acknowledgements format from one submission works for the next.

JAMA and its network journals ask authors to specify in the acknowledgements section whether the paper benefited from external statistical review, and if so, who conducted it. This is not standard across all major journals. The JAMA instructions also request that acknowledgements distinguish between colleagues who were involved in the research itself versus those who only reviewed the manuscript, which is a more granular requirement than most journals impose.

The Lancet journals have moved toward requiring a separate data sharing statement that may appear adjacent to or within the acknowledgements section. This is distinct from the data availability statement that many journals now require in a separate field, though the information overlaps. Some Lancet submissions require both, and the acknowledgements-adjacent version is specifically about what data are available to other researchers, while the data availability statement confirms that data are deposited in a named repository.

Nature journals, including Nature Medicine and the Nature Portfolio specialty journals, request a contributions section that functions similarly to CRediT author contribution statements but also covers acknowledged contributors in some submission workflows. If you are submitting to a Nature journal, read whether they ask for contributions to be listed for all named contributors or only for authors, because the requirements vary by journal and have been updated several times since 2024.

BMJ and its family of journals, as noted above, require the patient and public involvement statement as a separate element adjacent to acknowledgements, and specifically ask for a word count so they can confirm it is present. Several BMJ specialty journals have also adopted a section for equity, diversity, and inclusion considerations in their recent author instructions, noting whether the study team reflects the population under study and whether recruitment was designed to reach underserved groups. This is not yet a universal requirement but appears in enough journal instructions now that authors in clinical medicine should scan for it.

How Submission Systems Have Changed the Picture

Most major clinical journals now handle at least some acknowledgements-adjacent disclosures through structured fields in the submission system rather than through free-text manuscript sections. These fields include funding declaration, competing interests, author contributions, data availability, and AI use disclosure. The existence of these fields does not eliminate the need for corresponding text in the manuscript. It adds a layer where inconsistencies between system fields and manuscript text are caught by editorial staff and returned for correction.

A common source of delay is submitting a funding declaration in the system that lists different grants than the acknowledgements section in the manuscript, usually because a late-stage grant was added to one place but not the other. The same problem occurs with author affiliations entered in the system and those appearing on the title page. These are not intellectual problems, but they cost revision time that most authors would prefer to spend elsewhere.

Some submission systems now specifically prompt for AI use declarations as a mandatory step before final submission, and the absence of an entry in that field blocks progression to the next step. If you encounter this and you genuinely did not use any AI tool in writing or analysis, you should enter a statement confirming that fact, rather than leaving the field blank and hoping the system accepts it. Most systems expect a response, even if the response is negative.

The Most Common Errors That Trigger Editorial Queries

Running through the editorial queries that most frequently cite the acknowledgements section reveals a predictable short list. Understanding these helps authors proof their own manuscripts more effectively before submission.

The first is missing funding language. Authors name a grant but omit the grant number, or name the granting body without the institute, or use informal language rather than the funder-required template. Any of these will trigger a request for correction. The fix is to look up the required acknowledgement language directly from the funder, not to reconstruct it from memory.

The second is undisclosed writing support. A manuscript submitted with clean, polished English by a non-native-speaking team without any acknowledgement of editing or translation assistance will, in some cases, prompt an editor to ask. The question is not accusatory. The editor is trying to ensure the paper meets the journal's transparency requirements. The answer is to disclose the support upfront rather than wait to be asked.

The third is vague AI disclosure. An acknowledgements entry that says "AI tools were used in preparation of this manuscript" without naming the tool, the version, and the nature of the use will be returned for revision at most major journals. The standard has moved toward specific disclosure, and journal editors have both the instructions and the inclination to ask for clarification.

The fourth, increasingly, is the absent patient involvement statement at journals that require one. Authors working outside the UK sometimes miss this requirement because it is more deeply embedded in UK and some European journal culture. A brief scan of the target journal's author instructions for any mention of PPI, patient involvement, or lay advisory panels takes two minutes and prevents an avoidable back-and-forth.

The fifth is a mismatch between acknowledged contributors and CRediT roles listed for authors. If the CRediT taxonomy section credits "Formal analysis" to an author but the acknowledgements name a biostatistician who ran the main analyses, editors notice the apparent contradiction. Either the biostatistician's contribution rises to the level of authorship under ICMJE criteria, or the CRediT role needs reassignment. Sorting this out before submission is straightforward. Sorting it out in response to a reviewer comment during peer review is much more awkward.

A Practical Approach to Writing the Acknowledgements

The most efficient way to write a clean acknowledgements section is to divide it into logical blocks, each covering one category of contributor or obligation, and to draft each block from a written record rather than from memory. Start with a list maintained during the project of everyone who contributed and in what capacity. Separate authors from contributors at the beginning of the manuscript drafting process, not at the end, so the decision does not have to be made under time pressure.

The blocks to include, in the order that most journals expect them: contributions from named non-author contributors; professional medical writing or editing support (if any); AI use disclosures for manuscript preparation (if any); the patient and public involvement statement; funding acknowledgements with complete grant numbers and required funder language; and a brief data availability or sharing note if the journal does not provide a separate field for this.

Before submission, the corresponding author should circulate the acknowledgements draft to all named contributors and to the team's grants administrator to verify grant numbers. This takes one email and prevents the most common categories of post-submission corrections. It also gives named individuals the chance to flag any inaccuracies in how their contribution was described, which is both courteous and professionally sound.

The acknowledgements section is not where you settle disputes about authorship, explain why someone who expected to be an author was not included, or add qualifying language about the limitations of the contributions you are crediting. It is a transparent record of who helped and how, written in plain declarative prose. Keep it factual, keep it specific, and write it before the submission deadline rather than during it.

Further Reading