Placebo Groups Point to a Nocebo Effect Behind Nintedanib’s Stomach Side Effects

Nintedanib, an antifibrotic drug used for several forms of interstitial lung disease, is known to be associated with gastrointestinal adverse events including diarrhoea, nausea and vomiting. A new analysis asks how many of those complaints may have arisen not from the drug itself but from expectation. By examining what happened in the placebo groups of four major clinical trials after patients stopped taking their placebo, researchers found that gastrointestinal event rates fell after discontinuation, suggesting the possibility of a nocebo effect, an unfavourable effect that may be attributed to expectation rather than pharmacology.

The study, published as an open-access article in Arthritis Research & Therapy on 23 October 2026, drew on data from the placebo arms of the TOMORROW and INPULSIS trials in idiopathic pulmonary fibrosis, the SENSCIS trial in systemic sclerosis-associated interstitial lung disease, and the INBUILD trial in progressive fibrosing interstitial lung diseases other than idiopathic pulmonary fibrosis. Because these patients received no active drug, gastrointestinal events they reported cannot be direct pharmacological effects of nintedanib. The investigators compared rates of diarrhoea, nausea and vomiting during the on-treatment period, when patients were still taking placebo, with rates in the post-discontinuation period.

The pattern was consistent in direction across all four trials, with one exception. In TOMORROW, rates per 100 patient-years of diarrhoea, nausea and vomiting during on-treatment placebo were 19.8, 11.4 and 5.5 respectively, and all three dropped to zero in the post-discontinuation period. In INPULSIS, the corresponding on-treatment rates were 23.1, 7.6 and 2.9 per 100 patient-years, falling to 2.4, 2.4 and 0 after discontinuation.

The two later trials showed the same directional trend, although with more residual events after placebo stopped. In SENSCIS, which enrolled patients with systemic sclerosis-associated interstitial lung disease, on-treatment rates per 100 patient-years were 33.1 for diarrhoea, 11.8 for nausea and 9.0 for vomiting; after discontinuation these fell to 19.1, 2.4 and 7.4. In INBUILD, which studied progressive fibrosing interstitial lung diseases other than idiopathic pulmonary fibrosis, the on-treatment rates were 23.3, 7.7 and 3.6 per 100 patient-years, shifting to 2.2, 9.5 and 0 in the post-discontinuation period. Nausea was the one exception to the decline, rising from 7.7 to 9.5 per 100 patient-years after placebo stopped, a reminder that the signal is not perfectly uniform across every event type and every trial.

The authors conclude that in trials of nintedanib in patients with interstitial lung diseases, the fall in gastrointestinal adverse event rates after placebo was discontinued suggests the possibility of a nocebo effect.

The design of the analysis is what gives it weight. If gastrointestinal complaints in the placebo arm occur while patients are on treatment and fall when treatment stops, that temporal pattern is difficult to explain by the underlying lung disease alone and points toward expectation-driven symptoms.

The analysis also carries a caution for interpreting safety data across the spectrum of fibrotic lung disease. The four trials covered distinct populations, and the magnitude of the post-discontinuation decline varied among them. SENSCIS retained a diarrhoea rate of 19.1 per 100 patient-years after placebo stopped, the highest residual rate of the four trials, while TOMORROW recorded no gastrointestinal events at all in the post-discontinuation period.

The underlying trials were prospectively registered at ClinicalTrials.gov, under identifiers NCT00514683 for TOMORROW, NCT01335464 and NCT01335477 for INPULSIS, NCT02597933 for SENSCIS and NCT02999178 for INBUILD. The article itself was released early as a peer-reviewed, accepted, citable version carrying a permanent DOI, subject to further edits before the final Version of Record.

For clinicians, the message is not that nintedanib’s gastrointestinal toxicity is imaginary, since the drug’s association with diarrhoea, nausea and vomiting is well established, but that expectation itself appears to shape how much of that toxicity patients experience and report.

Subject of Research: Nocebo contribution to gastrointestinal adverse events reported in placebo groups of nintedanib trials in interstitial lung diseases

Article Title: Nocebo effect of nintedanib on gastrointestinal events in interstitial lung diseases

Article References: Distler, O., Menke, L., Highland, K. B., Azuma, A., Kulkarni, T., Wuyts, W. A., & Sfikakis, P. P. (2026). Nocebo effect of nintedanib on gastrointestinal events in interstitial lung diseases. Arthritis Research & Therapy. https://doi.org/10.1186/s13075-026-03934-3

Image Credits: AI Generated

DOI: 10.1186/s13075-026-03934-3

Keywords: nintedanib, nocebo effect, interstitial lung disease, idiopathic pulmonary fibrosis, gastrointestinal adverse events, placebo, diarrhoea, nausea, vomiting, systemic sclerosis, clinical trials, progressive fibrosing ILD