BACKGROUND: Research on health effects of shift work has especially focused on somatic diseases, such as breast cancer and cardiometabolic disease, while less attention has been given to the association between shift work and mental health.

METHODS: We used information on 19 964 female nurses (≥44 years) from the Danish Nurse Cohort, who reported current work schedule (day, evening, night, or rotating) at recruitment (1993/1999). In 5102 nurses who participated in both cohort waves, we defined persistent night shift work as working night shift in 1993 and 1999. We used Cox regression models to calculate hazard ratios (HRs) and 95% confidence intervals (CIs), adjusting for relevant confounders. Through linkage of cohort participants to national registers, we defined incidence of mood and neurotic disorders as first hospital contact or redeemed prescription until November 2018.

RESULTS: We found association between night shift work with mood disorders (HR = 1.31; 95%CI = 1.17-1.47) and neurotic disorders (1.29; 1.17-1.42), compared to day work. Associations were enhanced in nurses with persistent night shift work (1.85; 1.43-2.39 and 1.62; 1.26-2.09 for mood and neurotic disorders, respectively) and in nurses with specialist confirmed mood (1.69; 1.24-2.29) and neurotic (1.72; 1.22-2.44) disorders. Nurses with preexisting psychiatric disorders and full-time work seemed most susceptible.

CONCLUSIONS: Night shift work is associated with increased risk of major psychiatric disorders. The novel suggestive findings of vulnerable groups, including nurses with a history of psychiatric disorders and full-time workers, are based on a limited number of cases, and further research is needed to confirm the results.

To compare changes in low-density lipoprotein cholesterol and other lipids in patients with rheumatoid arthritis (RA) randomised to a 1-year treat-to-target strategy with either adalimumab plus methotrexate or placebo plus methotrexate. Prespecified secondary analyses from the OPERA trial, where 180 early and treatment-naïve RA patients received methotrexate 20 mg once weekly in combination with either placebo or subcutaneous adalimumab 40 mg every other week. Serum lipid levels were measured at baseline and after 1 year. Changes in lipid levels were analysed using mixed linear models based on the intention-to-treat (ITT) population. Overall, 174 patients were included in the ITT population (adalimumab plus methotrexate n = 86; placebo plus methotrexate n = 88). Differences between changes in lipid levels were low-density lipoprotein cholesterol 0.18 mmol/l [95% CI - 0.05 to 0.42], total cholesterol 0.27 mmol/l [- 0.002 to 0.54], high-density lipoprotein cholesterol 0.05 mmol/l [- 0.06 to 0.15], triglycerides 0.11 mmol/l [- 0.08 to 0.29], very-low-density lipoprotein cholesterol 0.03 mmol/l [- 0.05 to 0.12], and non-high-density lipoprotein cholesterol 0.22 mmol/l [- 0.02 to 0.46]. In early RA patients treated to tight control of inflammation over a period of 1 year with either adalimumab plus methotrexate or placebo plus methotrexate, changes in lipid levels were similar. Trial registration number: NCT00660647.

OBJECTIVE: To study the effects of supervised training in adults with subacromial pain syndrome.

DATA SOURCES: Embase, MEDLINE, Cochrane Library, Cumulative Index to Nursing and Allied Health, and Physiotherapy Evidence Database were searched from inception to March 2020.

STUDY SELECTION: Independent reviewers selected randomized controlled trials comparing supervised training with (1) no training or (2) self-training in adults with subacromial pain syndrome lasting for at least 1 month. Critical outcomes were shoulder pain, function, and patient-perceived effect. Important outcomes included other potential benefits and adverse events at 3-month follow-up.

DATA EXTRACTION: Two independent reviewers extracted data for the meta-analysis. Risk of bias was assessed using the Cochrane Risk of Bias tool 1, and certainty of evidence was evaluated using the Grades of Recommendation Assessment, Development, and Evaluation (GRADE).

DATA SYNTHESIS: Ten studies (n=597, 43% female) were included. Supervised training resulted in larger improvements than no training on pain (at rest: n=286; mean difference [MD], 1.68; 95% confidence interval [CI], 0.31-3.06 on 0-10 scale; during movement: n=353; MD, 1.84; 95% CI,0.91-2.76), function (n=396; standardized MD, 0.30; 95% CI, 0.07-0.52), and patient-perceived effect (n=118; risk ratio, 1.43; 95% CI, 0.87-2.34). Supervised training had potential benefits regarding quality of life, return to work, dropout, and training adherence, albeit more patients reported mild, transient pain after training. Supervised training and self-training showed equal improvements on pain (n=44) and function (n=76), with no data describing patient-perceived effect. Certainty of evidence was low for critical outcomes and low-moderate for other outcomes.

CONCLUSIONS: Supervised training might be superior to no training and equally effective as self-training on critical and important outcomes. Based on low-moderate certainty of evidence, these findings support a weak recommendation for supervised training in adults with subacromial pain syndrome.