To the Editors:

We appreciate the interest by Drs. Rico-Villademoros and Calandre in our study. Certainly, retrospective observational studies like ours always carry a risk of including one or more confounders, which are not controlled for, as well as a channeling of bias, which is inherent in this kind of research.

The deliberate choice of a more sedative antidepressant (e.g., mianserin or mirtazapine) rather than an SSRI could be a sign that patients taking SSRIs and patients taking the 5-HT2A–blocking drugs may differ, but we do not find it particularly likely that such a potential difference should have any bearing on the risk of developing a joint disorder.

As pointed out in our study, in the Swedish ADR database, the majority of reports carried with them information on the time interval between start of treatment and the appearance of the symptoms of joint disorder, i.e., antidepressant use preceded symptoms rather than, as suggested by Drs. Rico-Villademoros and Calandre, the reverse.

Anders Kling MD*, Rune Dahlqvist MD, PhD*, * University Hospital, Umeå, Sweden.