One Health research starts from a simple observation: the health of people, animals, and the environments they share is one connected system, not three separate ones. A pathogen that circulates in livestock, a resistance gene that moves through a farm's water supply, or a food chain that carries contamination from abattoir to table does not respect the boundaries between veterinary medicine, human medicine, and environmental science. Journals increasingly want studies that work across those boundaries. This guide explains what One Health research is, which studies fit, how to present an interdisciplinary manuscript so reviewers from different fields can all see its value, and the mistakes that most often hold such papers back.
One Health is an approach to designing and carrying out research, surveillance, and health programmes in which specialists from several disciplines work together on problems that sit at the interface of human, animal, and environmental health. It is not simply a veterinary topic with a new label, and it is not a synonym for zoonotic disease research, although zoonoses are a central part of it. The defining feature is integration: the question itself cannot be answered properly from inside a single discipline.
In practice, a One Health study usually has three characteristics. First, it involves at least two of the three domains — people, animals (domestic or wild), and their shared environment. Second, the link between those domains is part of the study design, not an afterthought in the discussion: samples, data, or interventions are deliberately collected across the interface. Third, the findings are meant to inform action in more than one sector — for example, a control measure that protects both a herd and the farming families alongside it.
Several pressures have moved One Health from a conference theme to a publishing priority. Zoonotic diseases — infections that pass between animals and people — continue to emerge and re-emerge, and the earliest signals are usually detected in animal populations or in the people who work closely with animals. Detecting and characterising those signals early is animal-health work with direct human-health consequences.
Antimicrobial resistance is the second major driver. Resistance that develops under antibiotic use in livestock production does not stay on the farm: resistant bacteria and resistance genes move through food, water, soil, and direct contact. Research that tracks resistance across animals, food products, farm environments, and human patients is exactly the kind of evidence that policy in both agriculture and medicine now depends on.
Food safety is the third driver. Contamination can enter at production, during transport and processing, or at retail, so a study that samples only one point in the chain sees only part of the risk. Finally, climate and livestock are increasingly studied together: shifting rainfall and temperature patterns change grazing systems, water availability, vector distributions, and disease patterns, affecting animal productivity and the livelihoods — and nutrition — of the households that depend on those animals. Work that connects these threads speaks to funders, editors, and readers in several fields at once.
If you are unsure whether your work qualifies, the most common study types in One Health publishing are a useful guide:
Disease surveillance across species. Prevalence and monitoring studies that sample animals alongside their human contacts, or that pair animal surveillance data with human case data from the same area and period.
Antimicrobial resistance studies. Work comparing resistance patterns in isolates from livestock, meat or milk, farm environments, and people — or evaluating interventions, such as changed farm antibiotic practices, and their downstream effects.
Zoonosis investigations. Outbreak investigations, risk-factor studies, and knowledge-and-practice surveys among farmers, abattoir workers, veterinarians, and other exposed groups.
Food-chain safety research. Studies that follow a hazard — bacterial contamination, residues, or spoilage organisms — through production, processing, and sale, and identify where control is most effective.
Wildlife and environmental interface studies. Research on disease exchange where livestock, wildlife, and people meet: shared grazing and water points, live-animal markets, peri-urban farming, and similar settings.
Studies that are purely clinical, purely production-focused, or purely ecological can still be excellent papers — but they are single-discipline papers, and they are usually better framed honestly as such than dressed in One Health language the design does not support.
The most common weakness editors see in One Health submissions is not the science — it is the framing. A manuscript written entirely for veterinary readers, with one paragraph at the end claiming human-health relevance, will disappoint a public-health reviewer; the mirror image fails the other way. Framing begins with the title and abstract: name both sides of the interface (for example, the animal population studied and the human exposure or outcome it connects to), so a reader from either field immediately understands why the paper matters to them.
In the introduction, build the case in both directions. Explain the animal-health problem, then the human or environmental connection, citing literature from both sides. State the gap as an interface gap: what is unknown precisely because previous work looked at only one side. Your objectives should then answer that gap explicitly.
In the discussion, resist the temptation to write two separate halves. Compare your findings with evidence from both disciplines, explain what each sector should do differently because of your results, and be candid about which side of the interface your data describes more weakly. Reviewers from the "other" discipline are quick to spot claims that outrun the data on their side — and equally quick to value authors who acknowledge the limits honestly. A short, concrete "implications across sectors" passage — for veterinary services, for public-health practice, and for producers or policymakers — is one of the most effective devices in this kind of paper.
Single-discipline framing. The study samples one population, in one sector, yet the title and conclusions claim One Health significance. Either strengthen the design or narrow the claim.
Missing or mismatched ethics approvals. Interface studies often need more than one approval: animal ethics for the veterinary component, human research ethics for interviews, questionnaires, or human samples, and sometimes institutional permission for farm or market access. Missing approvals are a frequent cause of delay and desk rejection.
Unclear data sharing and authorship. Multi-team studies generate shared datasets. State who collected what, how data and samples are stored and shared, and how authorship reflects each group's contribution.
Disconnected methods. Animal, laboratory, and questionnaire components described as if they were separate projects, with no explanation of how sampling locations, timing, or identifiers link them. Reviewers should be able to see how the pieces join up.
Overclaiming impact. A prevalence study in one district is a valuable baseline; it is not, by itself, a national control strategy. Proportionate claims are a mark of credibility in interdisciplinary review.
Most One Health manuscripts work well in a familiar structure, adapted at a few points. The title should name the hazard or question and both sides of the interface. The abstract should state where and when the study ran, which populations or materials were sampled on each side of the interface, the headline findings for each, and the cross-sector conclusion. The introduction moves from the disease or problem, to the interface where it operates, to the specific gap your study fills. The methods section benefits from subsections per component — study area and design, animal sampling, human participants and instruments, laboratory procedures, and analysis — plus a short paragraph explaining how the components were linked. Results are clearest when presented in the same order, ending with the findings that connect the components, such as matching isolates or shared risk factors. The discussion interprets across sectors, and the paper closes with limitations and practical recommendations. If you are preparing a veterinary or animal sciences manuscript generally, our animal sciences manuscript guidelines cover the formatting and preparation standards that apply across the journal.
The Global Journal of Veterinary and Animal Sciences (GJVAS), published by Global Academic Publishers, includes public health and One Health within its stated scope, alongside veterinary medicine, animal health, production, welfare, reproduction, and infectious diseases. Studies of the kinds described above — zoonoses, antimicrobial resistance, food safety, and surveillance at the animal–human interface — therefore sit naturally within the journal's remit, provided the veterinary or animal sciences component is substantial and the manuscript is prepared to the journal's standards. GJVAS is an open access journal publishing under a CC BY 4.0 licence, with double-blind peer review, so articles are free for anyone to read on publication — well suited to interdisciplinary work that must reach practitioners and policymakers as well as academics. Before submitting, read the step-by-step companion to this article, our guide to publishing veterinary and animal sciences research with GJVAS, and check your manuscript against the journal's scope and author instructions.
A strong One Health study is designed from the interface outward, not assembled afterwards from separate pieces. The starting point is a problem that genuinely spans sectors. A useful test is to ask whether answering the question properly requires evidence from more than one side. If a veterinary survey alone would settle it, the study may be veterinary research with a human-health context, which is a perfectly good paper but a different claim. If the answer depends on comparing, linking, or combining evidence across sectors, the One Health framing is justified and should shape every later decision.
Begin by defining the interface in concrete terms. Name the animal population, the human group, and the environment or pathway that connects them. "Livestock, farmers, and shared water" is a designable interface; "animals and people in general" is not. Specify the place, the time period, and the mechanism you are investigating: direct contact, a food product, an insect vector, contaminated water, or an occupational exposure. This precision determines sampling, ethics, and analysis, and it is what reviewers look for first when they judge whether the One Health label is earned.
Next, write objectives that cross the interface. Rather than one veterinary objective with a human-health hope attached, set objectives that each name their sector and the link between them. For example, an objective might be to describe a hazard in animals, to describe exposure or infection in a defined human group, and to examine whether the animal and human findings are connected in time, place, or by laboratory characterisation. When objectives are written this way, the methods, results, and discussion almost organise themselves, because each objective has a matching method and a matching result.
Then map the data you need before you collect anything. For each objective, list the population, the sample or respondent, the measurement, and the identifier that will allow linkage. If animal samples and human questionnaires are collected in the same villages but cannot be linked at household, farm, or site level, the study will end up as two parallel reports rather than one integrated analysis. Decide the linkage unit early — farm, household, market, water point, or district — and use it consistently in field forms, laboratory labels, and the dataset. This single decision prevents many of the "disconnected methods" problems editors see.
Finally, plan the analysis as an integrated analysis. Decide in advance which comparisons cross sectors: for instance, comparing the same organism or resistance pattern in animal and human isolates, testing whether a farm practice is associated with both animal and human outcomes, or triangulating laboratory results with interview findings about practice and exposure. Pre-specifying these cross-sector analyses in the protocol also strengthens ethics applications and makes the paper easier to review, because reviewers can see that the interface analysis was planned, not improvised after the data were collected.
Different One Health questions call for different designs, and the choice should follow the question rather than habit. A cross-sectional survey is often the right starting point when little is known about a problem in a place: it can estimate how common a condition, organism, or practice is in animals and in people at the same time, and identify factors associated with it. Its limit is that it captures one point in time, so it cannot by itself prove that an exposure caused an outcome. That limit should be stated plainly, not hidden behind interface language.
When the question is about sequence — whether an exposure precedes an outcome, or how a problem changes over seasons — a cohort or longitudinal design is stronger. Following the same farms, herds, or households over time allows researchers to observe new infections, changes in resistance, or the effect of an intervention as it unfolds. These designs cost more time and follow-up effort, and loss of participants or herds must be planned for, but they answer "what happens next" questions that a single survey cannot.
Case-control designs are useful when an outcome is uncommon or when an outbreak needs rapid investigation: compare animals, farms, or people with the outcome against comparable ones without it, and look back at exposures that differ. Surveillance designs, by contrast, are built for ongoing detection rather than a single answer — repeated sampling at abattoirs, markets, clinics, or veterinary practices, ideally with human and animal data reported side by side so that signals in one sector can be checked against the other. Intervention studies test a change, such as improved hygiene, altered antibiotic use, vaccination, or safer handling practices, and measure its effect across sectors; here the design must define the intervention precisely, decide who receives it and who does not, and measure outcomes on both sides of the interface, or the claim of a cross-sector benefit will be unsupported.
Qualitative and mixed-methods work also has a clear place. Interviews and focus groups with farmers, workers, veterinarians, and clinicians can explain why a practice persists, why a control measure is or is not adopted, and where official guidance conflicts with daily reality. Combined with laboratory or survey data, such findings often explain patterns that numbers alone leave puzzling — for example, why a recommended practice is rare despite being well known. Whatever the design, the manuscript should name it, justify it for the question, and describe sampling, inclusion criteria, and sample size reasoning in terms a reviewer from each discipline can follow. A design that is obvious to a veterinarian but opaque to a public-health reviewer, or vice versa, will slow review and weaken confidence in the paper.
One Health research is collaborative by definition, and the collaboration needs structure. The most effective teams agree early on who leads each component, who holds the combined dataset, and how decisions are made when disciplines disagree about methods or interpretation. A veterinarian, a human-health researcher, a laboratory scientist, and — where relevant — an environmental scientist or social scientist each bring standards, vocabularies, and expectations that differ. Naming those differences at the start is more productive than discovering them at the writing stage.
Agree authorship and data ownership before fieldwork begins, not after results arrive. Record who designed each component, who collected and analysed which data, and how shared data and samples may be used in future papers. Decide which publication comes first, who is corresponding author, and how each partner's institution is credited. Journals increasingly ask about contributions, and a written agreement protects both junior researchers and partner institutions. It also prevents the common situation in which one team's data are used to support claims that team has not reviewed.
Communication routines matter as much as agreements. Regular joint meetings during fieldwork — even brief ones — allow problems to be caught while they can still be fixed: a sampling site that cannot be matched across components, a questionnaire item that field staff interpret differently, or a laboratory delay that affects the human sampling schedule. A shared protocol, shared field forms, and a single data dictionary, in which every variable is defined once and used identically by all teams, do more for integration than any amount of post-hoc harmonisation. When partners are in different institutions or countries, agree data-sharing and sample-transfer arrangements, including permissions and material transfer terms, before samples move.
Community and stakeholder involvement should also be planned, not assumed. Farmers, market workers, abattoir staff, and local health workers are participants and partners, not only sources of samples. Explaining the study in plain language, agreeing how results will be fed back, and involving local veterinary and health services in interpretation makes recruitment more honest, findings more credible, and recommendations more likely to be usable. A short statement in the manuscript about how communities and local services were involved, where that applies, helps reviewers judge the study's conduct as well as its results.
Consider an illustrative study — not a report of real results — designed to examine antimicrobial resistance where dairy production, farm families, and milk handling meet. The interface is concrete: cattle, the people who milk and handle them, the milk itself, and the farm environment, including water and waste. The research question is whether resistance patterns in bacteria from these sources are connected, and which farm practices are associated with them. Framed this way, the study is One Health from the first sentence, because no single sector's data could answer it alone.
The design might combine a cross-sectional survey of farms with laboratory characterisation. On each participating farm, the team collects samples from animals, from milk, and from the environment, alongside a structured interview with the farm household about antibiotic use, milking hygiene, waste disposal, and illness in people and animals. Human samples are included only where ethically approved and genuinely relevant — for example, from consenting household members with occupational exposure — and human and animal components are linked at farm level through a shared farm identifier, so that patterns can be compared within the same setting rather than only in aggregate.
In the laboratory, isolates from each source are identified and tested by the same methods, so that comparisons across sources are valid. The analysis then has three layers: a description of resistance in each source, a comparison across sources within farms, and an examination of which practices are associated with resistance. The discussion can then speak to each sector separately and together: what veterinary services should advise on antibiotic use, what milk handlers and households should change in practice, and what further surveillance would be needed before any wider policy claim could be made. Note what this example does not claim: it does not prove transmission from animals to people, and it does not support a national strategy on its own. Its value is a clearly bounded, linked picture of one system — exactly the kind of evidence that later, larger studies and policy can build on.
A second illustrative example shows how a zoonosis investigation differs. Suppose a veterinary service notices cases of an animal disease known to affect people in a farming district, and a joint team sets out to understand exposure and risk. The animal component establishes how the disease is distributed among herds and which herds are affected. The human component surveys the people most exposed — farmers, herders, and abattoir or market workers — about contact patterns, protective practices, and symptoms, with appropriate consent and confidentiality. An environmental component, where relevant, records shared grazing, water points, or market conditions that could connect herds and households.
The strength of this design is in the linkage. If animal cases, human exposures, and shared sites are recorded with common place and time identifiers, the team can examine whether human risk is higher where animal disease is present, which practices distinguish higher-risk from lower-risk households, and where an intervention — such as vaccination, safer birthing or slaughter practices, or protective equipment — would have the greatest effect. Interviews add the human explanation: whether protective measures are available, affordable, and actually used, and what barriers prevent their use.
When writing such a paper, the authors should be careful with causal language. An association between animal disease in a herd and risk in its household is informative, but it does not by itself prove the direction of spread, and a single-district study cannot be generalised to a whole country without qualification. The paper's contribution is strongest when it states the finding, its limits, and the next study that would be needed — for example, longitudinal follow-up or laboratory confirmation linking isolates — rather than overstating what one investigation has established.
Sampling is where One Health designs most often succeed or fail. Each component needs a defined sampling frame: which farms, markets, clinics, or herds, selected how, and why that selection represents the population the paper claims to describe. Convenience sampling — taking whatever animals or respondents are easiest to reach — is common in field research, but it must be described honestly and its limits acknowledged. If animal and human samples come from different places or periods, the paper cannot compare them as if they were matched, and reviewers will notice.
Laboratory comparability is equally important. Using different identification or susceptibility-testing methods for animal and human isolates undermines cross-sector comparison, even if each method is standard in its own field. Where methods must differ, explain why and how results were made comparable, or limit the comparison accordingly. Record sample handling, transport, and storage conditions, because these affect recovery and can create apparent differences between sources that are really artefacts of handling. A methods section that a laboratory reviewer from either discipline can reproduce is a strong signal of quality.
Data management deserves the same care as fieldwork. A single data dictionary, consistent identifiers, and a documented process for cleaning and merging datasets prevent errors that are difficult to detect later. Decide how missing data, unlinked records, and discordant results will be handled before analysis, and report how many records were excluded at each stage. For qualitative data, describe how interviews were recorded, transcribed, and analysed, and how themes were checked across researchers. Transparency here does not weaken a paper; it is what allows reviewers from different fields to trust a combined analysis they did not themselves conduct.
Ethics in One Health research is layered, because each component carries its own obligations. Animal work requires animal ethics approval and attention to welfare, housing, handling, and humane sampling. Human work requires human research ethics approval, informed consent in a language participants understand, confidentiality for health and occupational information, and particular care where participants are workers whose livelihoods could be affected by findings. Site permissions — from farms, markets, abattoirs, clinics, or local authorities — are separate from ethics approval and should be obtained and recorded.
Consent should cover the cross-sector nature of the study. Participants should understand that their samples or answers may be linked to animal or environmental data at farm or site level, how identifiers will be protected, and whether results will be fed back to them or their community. Where human samples are stored for future testing, that future use needs explicit consent and approval, not an assumption. For animal owners, consent should cover sampling, any follow-up visits, and what happens if a notifiable disease is found — including any reporting obligation that applies in that setting.
Biosafety and data sharing complete the picture. Field and laboratory teams need training, protective equipment, and procedures for handling potentially infectious samples, and these should be described briefly so reviewers can judge safety and feasibility. For data sharing, state what will be shared, in what form, and under what conditions: anonymised datasets, laboratory protocols, and questionnaires can often be shared on request or through a repository, while identifiable human data cannot. A clear data-availability statement, agreed among all partners, avoids disputes after publication and aligns with the expectations of open access publishing. If you need a refresher on how open access licensing works for published articles, our guide to open access publishing explained covers the CC BY 4.0 licence used by GJVAS.
Results in a One Health paper should be organised so that a reader from any discipline can follow the argument without re-reading the methods. Presenting components in the same order as the methods — animal, human, environmental, then linked findings — helps, and the linked findings deserve their own subsection rather than being scattered. Tables that place animal and human results side by side, with the same definitions and time periods, make comparison immediate; figures such as maps of sampling sites, flow diagrams of linked samples, or diagrams of the transmission pathway under study can convey the interface more clearly than paragraphs alone.
Define every term at first use, including terms that are routine in one field but unfamiliar in another. Abbreviations, organism names, resistance categories, and occupational groups should not be assumed knowledge. When reporting statistical results, state the analysis, the comparison, and the uncertainty, and avoid presenting a cross-sector association as proof of transmission. Language matters here: "isolates from animals and people shared the same resistance pattern" is a description; "animals transmitted resistance to people" is a causal claim that requires stronger evidence, and reviewers will hold the paper to the stronger standard if it uses the stronger words.
Limitations should be written across sectors, not only for the author's home discipline. Note where sampling was unmatched, where one component was smaller or less complete, where laboratory methods differed, and where findings apply to one setting rather than generally. A candid limitations section, followed by specific recommendations for each sector and a clear statement of what the next study should do differently, is often what distinguishes a publishable interdisciplinary paper from one that reviewers return for overclaiming. Authors who show they understand the limits on both sides of the interface earn the trust that makes editors willing to send the work to reviewers from both sides.
Before submitting, work through this checklist with co-authors from each discipline. First, does the title and abstract name both sides of the interface and the link between them? Second, is the One Health claim supported by the design, with linked sampling or data, rather than only by the discussion? Third, are all ethics approvals, site permissions, and consent procedures stated, with approval details the journal can verify? Fourth, can a reviewer from each discipline reproduce the methods for their component, and can they see how the components were linked? Fifth, are claims proportionate — associations described as associations, single-site findings presented as single-site findings, and recommendations tied to the evidence actually presented?
Next, check the practical details: are tables and figures readable without the main text, are organism and drug names correct and consistent, are references drawn from both sides of the literature, and does the reference list include the foundational One Health and sector-specific sources a reviewer would expect? Confirm that authorship and contributions reflect every team, that funding and competing interests are declared, and that the data-availability statement matches what partners agreed. Finally, read the manuscript once as a veterinary reviewer and once as a public-health reviewer, asking at each paragraph whether that reader would find the paper useful, credible, and clearly written. If either reading leaves a gap, that gap is where revision is still needed — and closing it before submission is far faster than closing it after peer review. For journal-specific formatting and submission steps, the GJVAS author guide and our animal sciences manuscript guidelines should be checked alongside this article.
| Sector | Focus | Example topic |
|---|---|---|
| Human health | Disease prevention and treatment | Zoonotic disease surveillance |
| Animal health | Livestock and wildlife health | Veterinary vaccination programmes |
| Environment | Ecosystems, water and food safety | Antimicrobial resistance in water |
| Shared policy | Joint action across sectors | Cross-sector outbreak response |
These authoritative external resources explain the standards and tools mentioned in this guide:
What is One Health research in simple terms?
It is research that studies health problems where people, animals, and their shared environment meet — and designs the study across those boundaries instead of looking at one side alone.
Does my study count as One Health if it only samples animals?
Possibly, but the claim needs support. An animal-only study is strongest as One Health work when it is explicitly connected to human or environmental health — for example, sampling animals specifically because of a documented exposure pathway to people, and analysing the results in that context. If the connection is speculative, present the paper as the veterinary study it is.
Which ethics approvals does a One Health study need?
Usually one for each component: animal ethics approval for work with animals, human research ethics approval for questionnaires, interviews, or human samples, and site permissions where relevant. Journals will ask for these statements, and missing approvals delay decisions.
Why do journals encourage interdisciplinary papers?
Because many of the problems journals exist to address — emerging zoonoses, antimicrobial resistance, unsafe food chains — cannot be solved within one discipline. A well-framed interface study is useful to more readers, reviewers, and decision-makers than the same data framed for one field alone.
Can I submit a One Health manuscript to GJVAS?
Yes, if the work falls within the journal's scope: the animal or veterinary component should be substantial, and the human or environmental link should be part of the design rather than an afterthought. Review the journal on our journals page and follow the author guidance before you submit.
How do I choose a study design for a One Health question?
Let the question decide. Use a cross-sectional survey to describe a problem at one point in time, a longitudinal or cohort design to follow change over time, a case-control design for rapid investigation of an uncommon outcome, surveillance for ongoing detection, and an intervention design to test a change. State the design, justify it for the question, and describe sampling and sample-size reasoning so reviewers from each discipline can follow it.
How should I link animal, human, and environmental data?
Choose a linkage unit before fieldwork — farm, household, market, water point, or site — and use it consistently in field forms, laboratory labels, and the dataset. Record place and time identifiers for every component, use a single data dictionary, and pre-specify the cross-sector comparisons you will make. Without a shared identifier, components that were collected together will read as separate studies.
What should an interdisciplinary team agree before fieldwork?
Who leads each component, who holds the combined dataset, how authorship reflects each group's contribution, how samples and data may be reused, and how results will be fed back to participating communities. Agreeing these points in writing before data collection prevents most authorship and data disputes after results arrive.
Can qualitative research be part of a One Health paper?
Yes. Interviews and focus groups with farmers, workers, veterinarians, and clinicians can explain why practices persist, why control measures are or are not adopted, and where guidance conflicts with daily reality. Combined with survey or laboratory data, qualitative findings often explain patterns that numbers alone leave puzzling, provided the qualitative methods are described as rigorously as the quantitative ones.
How do I avoid overclaiming in a One Health manuscript?
Match every claim to the evidence that supports it. Describe associations as associations, do not present a shared pattern as proof of transmission without stronger evidence, and present single-site or single-district findings as such. State limitations for each sector, not only your own discipline, and tie each recommendation to a finding in the paper.
Submit your manuscript ← Back to GAP Insights