After publication of this manuscript, the authors became aware of an error in the coding used to generate the analytic dataset. This erroneously excluded, from the cohort at baseline, all individuals who died at any time during follow-up. Consequently, the deceased did not contribute the risk-time they should have contributed up to their date of death, including risk-time before any diagnosis of autoimmune rheumatic diseases (RA, SS, SLE) occurring prior to death. Once identified, the authors corrected the coding and reran all analyses. They then systematically re-checked all reported estimates throughout the manuscript and updated any values that changed accordingly. Overall, the rerun analyses are consistent with the published version. The direction of associations and their interpretations remain the same. The main conclusions of the article are unaffected. For a sensitivity analysis of full follow-up since 1979, the results of the reanalysis were similar to those of the main analyses, but for the estimates regarding accumulated work life exposure, where some estimates either became or lost statistical significance. The paper has been corrected and is provided here in full along with revised supplementary material.