We appreciate the interest in, and welcome the critical reading by, Choi et al.  of our paper that first appeared in the online version of this Journal 2 years ago (December 2011) . Our analyses were based on studies from the IPD-Work collaboration which has generated important new evidence on the association between job strain, risk factors and disease end-points, such as coronary heart disease and cancer [2-9].
According to our latest count, Choi et al. have now published at least eight letters to the editors of various journals, all of which are predicated upon the effects of job strain on various health outcomes being greater than we have empirically demonstrated them to be using the largest individual participant meta-analyses conducted to date. Rather than providing empirical evidence to justify their argument, they speculate about biases that could have led to an underestimation of effect estimates.
It is well known that all research based on observational data is prone to various biases which may not only lead to an underestimation of the reported associations but may also artificially inflate them. Our studies are no exception. This is a point we feel we have made in the papers themselves, in invited commentaries [10, 11], and indeed, in prior responses to correspondence from Choi et al. [12-14].