The authors regret to inform that a phrasing error has been discovered in one of the items of the self-report instrument used as primary outcome measure in the trial. Consequently, the item is perceived as one that should be reverse-coded (the Swedish equivalent of the word “not” was omitted), yet the total scores used for analyses and reported in the publication were not adjusted for this error. After discussing with the journal editor, we have decided that the most appropriate adjustment would be to reverse-code the item and calculate new total scores for each individual and time point, since the instrument already features other reverse-coded items. Analyses have been repeated using these new adjusted total scores. As expected (the total score is the sum of 17 items), the impact on individual total scores and group means was negligible. Study findings were robust, with only very minor changes to effect size estimates. Researchers interested in performing meta-analyses on reported means and standard deviations should consult the corresponding author to access a revised version of the original Table 2 containing these. The authors would like to apologise for any inconvenience caused.