Conflicts of interest: who paid, and why it matters

The Cochrane review Industry sponsorship and research outcome, in its current version, updated in 2017, included 75 papers that compared drug and device studies paid for by the manufacturer with studies paid for by others.¹

• Industry-sponsored studies more often had favourable efficacy results (risk ratio 1.27, 95% confidence interval 1.17 to 1.37; moderate-quality evidence) and favourable conclusions (risk ratio 1.34, 1.19 to 1.51; low-quality evidence).
• They did not differ in the standard risk-of-bias items for how groups were formed, followed and reported, and they more often had a low risk of bias from blinding.
• Their conclusions agreed less often with their own results (risk ratio 0.83, from six papers).
• The authors concluded that there is “an industry bias that cannot be explained by standard ‘Risk of bias’ assessments”.

The same pattern shows up in opinion. Of 202 articles commenting on the heart risk of the diabetes drug rosiglitazone, only about half carried a conflict-of-interest statement, and authors with a favourable view were more likely to have financial ties with its makers (rate ratio 4.29). The authors note that this does not prove the ties caused the views.²

None of this means company research is worthless. Clinical research is increasingly paid for by the companies that make the products.¹ It means a company’s result needs a second, independent look before it can be taken as settled; the semaglutide example at the end of this page shows what that looks like.

Sponsored studies. Medical journals use a shared form on which authors declare their financial and other relationships.³ The trial registry entry names the sponsor and the outcomes planned in advance, so a published paper can be compared with its plan (see how to read a study).⁴ A missing statement is itself information: in the rosiglitazone debate, almost half of the articles had none.²

Reviews and “research” pages. A narrative review chooses its own studies. When the author has something to sell, three patterns are worth watching for: animal results described as if they were shown in people, positive studies cited and negative ones left out, and reviews cited as if they were new evidence (see “When reviews echo reviews” below).^6,7 The same applies to the “research” or “science” pages on websites that sell peptides. They may cite real studies, accurately, and still mislead by what they leave out.

Sellers’ lab reports. A certificate of analysis supplied by a seller is also a claim by an interested party. In one study of products sold online without a prescription, the sellers’ documents claimed a purity of at least 99%. Independent testing of the three vials that arrived found that semaglutide made up only about 8% to 14% of the powder, and that the amount of semaglutide was 29% to 39% above the amount on the label.⁸ What a certificate can and cannot show is explained on product risk.

In 2011 a survey of corresponding authors of 2008 articles in six high-impact medical journals found honorary or ghost authors, or both, in 21.0% of articles: 17.6% had an honorary author and 7.9% a ghost author, a fall from 1996 for ghost authorship.⁹

Court documents from litigation over the painkiller rofecoxib showed how this can work. Company employees, alone or with medical publishing companies, prepared trial manuscripts and then recruited academic investigators as authors, often in first and second place. Marketing staff planned review articles, contracted publishing companies to write them and recruited academic authors, who were often offered payment. Of 72 related review articles, only half disclosed the company’s sponsorship or the author’s payment.⁵

The comparison in the table comes from a study of 93 potential predatory journals, 99 legitimate open-access journals and 100 subscription journals, which identified 13 characteristics that may help tell them apart. Subscription journals showed even fewer of these signs, for example spelling errors on 3% of homepages, and charged a median fee of US$3,000.¹⁰ Scholars and publishers from ten countries later agreed a shared definition of predatory publishing.¹¹

Practical checks: is the journal indexed in MEDLINE (search it on PubMed), does it name an editor you can find elsewhere, and does it describe its peer review? A paper in a legitimate journal can still be wrong; how to read a study applies to every paper.

• Follow the citation back. If a review says “BPC-157 heals tendons”, find the study it cites. Is it in rats or people? From which group?
• Count primary studies, not citations. A claim with fifty citations may rest on one experiment.
• Watch for reviews of the same group. When most reviews on a compound are written by, or depend on, the group that did most of the original work, they echo it; they do not confirm it.
• Look for the disagreements. A fair review cites the studies that found nothing, and says why it weighs them as it does.

BPC-157 is a live example. Between 2024 and 2026, most new BPC-157 records in PubMed (34 of 52) did not list the leader of the Zagreb group that did most of the original work as an author. But more than half of those were tagged as reviews, comments, letters or editorials, and because about 85 in 100 primary studies come from Zagreb, any review of this literature rests mostly on that one group’s data.¹² A review that cites a study is an echo of it, not a second test (see the Croatian lab).

Payments to doctors. Among 279,669 US doctors who prescribed in four common classes of medicine, there were 63,524 company payments linked to the four promoted brand-name drugs between August and December 2013; 95% were meals, with a mean value under US$20. Doctors who received one sponsored meal had higher odds of prescribing the promoted drug over alternatives (odds ratios from 1.18 to 2.18 for the four drugs), and more or costlier meals went with higher rates. The authors stress that this shows an association, not cause and effect.¹³ In the United States such payments are published: the Open Payments database lists payments that drug and device companies make to doctors.¹⁵

Influencers. A preregistered review of research on prescription-drug promotion by social-media influencers, published in 2026, found three recurring problems: misinformation, because influencers often lack expertise and their audiences often have low health literacy; weak oversight and inconsistent disclosure of payment; and personal, emotional stories that blur the line between testimony and paid promotion. The evidence base was small, 12 articles.¹⁴

For peptides sold outside pharmacies the same questions apply, with one more: does the person recommending a product earn from its sale, through a discount code, an affiliate link or their own shop? If so, treat what they say as advertising, however sincere it sounds.

Semaglutide. Every pivotal trial was funded by the manufacturer, with many company authors. The main recent safety questions, NAION, thyroid cancer and suicidal thoughts, were examined by independent academic groups and by regulators. The EU safety committee concluded that NAION is a very rare side effect; a large Scandinavian study of GLP-1 medicines as a group found no substantial rise in thyroid cancer over about four years; and the committee found that the evidence does not support a causal link with suicidal thoughts.^16,17,18,19 Company funding did not make the results false; independent checking is what lets them be relied on, and it can find what the company trials missed. Its evidence grade is “broadly replicated” for independence (see the semaglutide page).

BPC-157. Most of the published research comes from one group in Zagreb, and its leader is named as an inventor on a 1998 international patent application for salts of the peptide.^12,20 That does not show wrongdoing, and we found no sign of any. It means the findings need what they mostly lack: repeats by other laboratories. Its grade for independence is “single group dominant” (see the Croatian lab).

The same rule applies to this site. We sell nothing and take no money from anyone who makes or sells these products; how we are funded is set out on funding.