Background Organisational injustice may affect sleep quality, but previous empirical studies have not examined whether improvement in the psychosocial risk factors is associated with a beneficial change in sleep.

Aim We examined the onset and removal of the exposure to organisational injustice (relational and procedural) in relation to subsequent changes in insomnia symptoms.

Methods Participants were 24,287 Finnish public sector employees with data from three consecutive surveys between 2000 and 2012, contributing to a total of 36,189 observations. We applied a “pseudo-trial design” to determine temporality between deterioration in relational and procedural justice over the first two surveys, and onset of insomnia symptoms in the third survey. Similarly, we examined, if the improvement in the psychosocial exposures was associated with subsequent remission of insomnia symptoms. This design allowed us to examine whether removal of relational and procedural injustice was associated with better sleep in the third survey. Several social and health-related covariates were controlled for in generalised estimating equation based models, allowing to account for within person correlation.

Results Onset of relational injustice was associated with subsequent insomnia symptoms (OR 1.15; 95% CI: 1.02–1.30), whereas no association was found for procedural injustice (OR 1.08; 95% CI: 0.95–1.22), after adjustments for sociodemographic factors, health behaviours, comorbid conditions, shift work and apnoea. Removal of relational injustice was associated with better sleep after full adjustments (OR 0.83; 95% CI: 0.71–0.96). Removal of procedural injustice did not improve sleep (OR 0.91; 95% CI: 0.77–1.06).

Conclusions These results suggest that relational injustice can worsen sleep quality, while removal of relational injustice appears to improve sleep.