The article admits at the bottom: "The study findings do not prove that volatile pay causes health symptoms, only that there is a correlation between them." Ye

The article admits at the bottom: "The study findings do not prove that volatile pay causes health symptoms, only that there is a correlation between them." Yet, in the body of the article, it says "pay volatility was less harmful to health", the wording of which implicitly asserts causation. Sigh. The null hypothesis is "health problems, and difficulties associated with health problems, cause people to take jobs with more volatile pay". It is plausible that volatile pay causes stress which causes problems, but this doesn't provide any convincing evidence of it. In fact, this part is evidence against it: "Mindfulness [...] has been shown to buffer against stress in other aspects of work and life [...] but it did not help reduce the physical symptoms associated with volatile pay." (Of course, an individual's mindfulness could also influence what jobs they'll accept and keep.) (Another mechanism comes to mind for the "performance bonuses" they mention. That could lead to some people staying up late to finish things, or otherwise deciding to stretch themselves physically for their job. But if you want to study that, you should ask about those behaviors specifically.)