Conflict of Interest
Conflict of Interest. A financial, professional or personal tie that gives the author of a claim a stake in what the claim concludes.
A stake in the answer
A conflict of interest exists when the person or organisation making a claim also stands to gain or lose depending on what the claim says. It is a description of a situation, not an accusation. Plenty of conflicted research is careful and correct. The reason to record conflicts is that, across many studies, they are associated with a predictable lean, and a reader deserves to know when that lean might be present.
The forms it takes around calorie apps
- Funding. The app’s developer pays for, commissions or runs a study of its own product.
- Employment and authorship. Authors work for, consult for, advise or hold shares in the developer.
- Commercial arrangements. A review site earns affiliate commissions when readers subscribe, sells advertising to app makers, or receives free premium accounts.
- Professional ties. A clinician or coach endorses an app they helped design, or one whose maker sponsors their work.
- Personal ties. Family or close relationships with people at the company.
Not all of these are equal. A free premium account given so that a reviewer can test paid features is a small, common and easily disclosed arrangement. Equity in the company whose product is being ranked is a large one.
What the evidence says about funding
Lundh and colleagues pooled studies of drugs and medical devices for the Cochrane Collaboration’s methodology series. Trials paid for by manufacturers were more likely than others to report results and conclusions that favoured the product, and the usual checklists for study flaws could not account for the whole gap. Nothing in that finding requires misconduct. Choices about which comparison to run, which test set to use and which outcome to highlight can all tilt a result while every individual step is defensible.
The same logic applies to calorie trackers. A developer choosing the meals for its own accuracy test will, quite reasonably, choose meals it understands well. An outside group choosing meals to stress-test the app will choose differently.
Worked example: reading a disclosure
Imagine an accuracy study of a calorie app with this statement at the end: “Funded by the app’s developer. Two of four authors are employees. The meal set was selected by the developer’s nutrition team.”
A careful reader would note three separate ties, funding, employment and control of the test set, and would treat the headline figure as plausible but unconfirmed until a group without those ties reports something similar. Compare: “Funded by a university grant. No author has a relationship with any app developer. Meals were selected before the apps were chosen.” Same method, very different weight.
How it is used in our grading
When we grade evidence (see evidence grade), who paid and who wrote the study are weighed under risk of bias. We still cite a study run by an app’s maker if nothing better exists, but it cannot earn a top grade on its own; that takes a matching result from people without the same ties, as the replicability entry explains. Vendor-funded studies are still cited where they are the only available evidence, but their grade is capped at low under the v3.2 framework.
The publication’s editorial team has declared their own conflicts on each author profile on the about page. No member of the editorial team has financial relationships with calorie-tracking-app vendors, no member holds equity in app makers or developers, and no member has advisory positions with companies whose products are reviewed. Income derives from this publication and from independent academic consulting unrelated to consumer software. See also our no-affiliate disclosure.
If you believe a conflict has been missed, in our work or in a study we cite, write to editor@whatsthebestcalorietracking.app.