Objective
We investigated the extent to which ward-level leadership quality was associated with prospective low-back pain among eldercare workers, and how this association was mediated by observed resident handlings.

Methods
530 Danish eldercare workers, employed in 121 wards, distributed across 20 nursing homes were evaluated. At baseline, leadership quality was measured using the Copenhagen Psychosocial Questionnaire, and resident handlings [handlings per shift, handlings not using assistive devices, handlings done alone, interruptions to handlings, impediments to handlings] were assessed using observations. Frequency and intensity of low-back pain was assessed monthly during the following year. All variables were averaged for each ward. We used ordinary least squares regressions to examine direct effects of leadership on low-back pain and indirect effects through handlings, using PROCESS-macro for SPSS.

Results
After adjustments for low-back pain at baseline, type of ward, staff ratio (i.e., number of workers divided by number of residents) and proportion of devices not in place, leadership quality showed no effect on prospective low-back pain frequency (β = 0.01 [− 0.05:0.07]) and a small beneficial effect on pain intensity (β = − 0.02 [− 0.04:0.00]). Resident handlings did not mediate the association between leadership quality and frequency or intensity of low-back pain.

Conclusions
Good leadership quality was associated with a small decrease in prospective low-back pain intensity, but resident handlings did not seem to play a mediating role, although better ward-level leadership quality contributed to fewer workplace-observed resident handlings without assistance. Potentially, organizational factors, such as type of ward and staff ratio, may have a greater influence on handlings and low-back pain than leadership quality per se among eldercare workers.