Disturbances in slow-wave sleep are induced by models of bilateral inflammation, neuropathic, and postoperative pain, but not osteoarthritic pain in rats

Preclinical assessment of pain has typically relied on measuring animal responses to evoked stimulation. Because of inherent limitations of these assays, there is a need to develop measures of animal pain/discomfort that are objective, not experimentally evoked, and mimic the human condition. Patients with chronic pain manifest a variety of co-morbidities, one of which is disturbances in sleep. We used electroencephalography to objectively assess 4 rat models of pain (inflammatory/complete Freund's adjuvant [CFA], neuropathic/chronic constriction injury [CCI], postoperative/skin incision, osteoarthritis/monosodium iodoacetate [MIA]) for the occurrence of sleep disturbances. Four different measures of slow-wave sleep (SWS) were examined: amplitude of 1- to 4-Hz waves, total time spent in SWS, time spent in SWS-1, and time spent in SWS-2. Bilateral injuries were more likely to induce a sleep disturbance than unilateral injuries in the CFA, CCI, and skin incision assays. Sleep disturbances occurred in the deeper stage of SWS, as the amplitude of 1- to 4-Hz waves and time spent in SWS-2 were significantly decreased in all models except the osteoarthritis model. Sleep disturbances lasted for approximately 3 to 14days, depending on the model, and were resolved despite continued hypersensitivity to evoked stimulation. Morphine, gabapentin, diclofenac, and ABT-102 (TRPV1 antagonist) all improved sleep in the bilateral CFA assay at doses that did not significantly alter SWS in uninjured rats. Preclinical assessment of compounds should follow the path of clinical studies and take into account diverse aspects of the "pain condition." This would include evaluating nociceptive thresholds as well as other endpoints, such as cognition and sleep, that may be affected by the pathological state.