Circadian rhythms coordinate systemic physiology and are important for skeletal homeostasis. In humans, chronic disruption of circadian rhythms (e.g. shift work) is associated with bone loss and increased fracture risk. However, whether bone loss induced by disruption of circadian rhythms is reversible following restoration of normal circadian cues remains poorly defined.

To investigate this, 8-week-old C57BL/6J male mice were exposed to an established model of chronic circadian rhythm disruption (CCRD). Mice were maintained on either a normal 12:12hr light-dark cycle (non-shifted) or exposed to weekly 12hr phase-shifts (shifted) for either 10 or 22 weeks, equivalent to spending alternate weeks in London and Sydney. An additional cohort of mice was exposed to CCRD for 10 weeks and then returned to the normal 12:12hr light-dark cycle for the remaining 12 weeks of the experiment (recovery group). At endpoint, tibiae were harvested, scanned by micro-CT and analysed for trabecular and cortical bone loss.

CCRD caused pronounced bone loss in the tibiae of all shifted groups compared to age-matched non-shifted mice. Trabecular bone volume fraction (BV/TV) was reduced by 32% in shifted mice after 10 weeks of CCRD relative to non‑shifted controls (P<0.0001). Consistent with physiological ageing, non‑shifted mice also exhibited a reduction in trabecular bone volume between 10 and 22 weeks of the experiment (P=0.0004). Mice subjected to 10 weeks of CCRD followed by restoration of normal light-dark cycles for 12 weeks (recovery group) displayed trabecular bone volume comparable to age‑matched non‑shifted controls (BV/TV: 14% non-shifted vs 13% recovery, p=0.3213). However, mice that were shifted for 22 weeks displayed a 20% decrease in bone volume compared to age-matched non-shifted mice (14% non-shifted vs 11% shifted, p=0.0026). Cortical area fraction (Ct.Ar/T.Ar) was reduced by 11% after 10 weeks of CCRD relative to non‑shifted controls (P=0.0006). However, this difference was no longer evident at 22 weeks as cortical area fraction was comparable among non-shifted, shifted and recovery groups. This suggests trabecular but not cortical bone has sustained sensitivity to CCRD.

We conclude that trabecular bone loss induced by CCRD is partly reversible following realignment of circadian rhythmicity, offering a potential strategy to mitigate skeletal decline.