Clinical studies have reported that sleep disorders in perioperative patients exacerbate postoperative pain (1–3). However, well-evidenced treatments for this issue have not been established. While basic research is essential for developing novel treatments grounded in scientific mechanisms, traditional animal models of sleep disturbance used in such studies often fail to replicate the sleep disorders experienced by humans in modern society. The conventional animal models rely on physical stimuli or acute strong stress, such as single platform-on-water, forced movement, mechanical stimulation, and fear-conditioning, to interrupt sleep (4–6), which does not accurately mimic the more subtle sleep disorders associated with modern stressors.

To address these limitations, we utilized the perpetual avoidance of water on a wheel (PAWW) model. In this model, mice are housed in a running wheel set above shallow water, promoting chronic voluntary movement and disrupting their sleep–wake cycle without acute physical stress (7). Although the PAWW model induces chronic stress, its primary feature is disruption of the normal sleep–wake cycle via increased spontaneous activity. In this study, we focused on the sleep impairment aspect of the PAWW model as it pertains to postoperative pain recovery. While we acknowledge that stress and sleep disturbances are closely interrelated, our experimental design aimed to investigate the effects of stress-induced sleep disruption on pain recovery, rather than measuring stress responses.

Pain and sleep are interlinked through common brain regions such as the frontal cortex, hypothalamus, and brainstem (8, 9), and via shared neurotransmitters, including opioids (10–13), monoamines (14–17), and orexin (18, 19). Dysregulation of these pathways contributes to both sleep disturbances and pain sensitization. Accordingly, pharmacologic targets such as antidepressants (20, 21), orexin receptor antagonists (22), kappa opioid receptor (KOR) antagonists (8, 9), and melatonin receptor agonists (23) are promising candidates for addressing both conditions.

Recent evidence has highlighted the role of dynorphin/KOR signaling in the anterior cingulate cortex and hypothalamus is implicated in sleep fragmentation and emotional dysregulation. Systemic KOR antagonism restores normal sleep in chronic pain models without altering baseline sleep in sham animals (8, 9). Moreover KOR signaling has also been reported to induce mechanical hypersensitivity in rodent pain models (24, 25). Although the underlying mechanisms of these effects remain unclear, they might involve interactions between the anterior cingulate cortex, hypothalamic orexin neurons, and ascending monoaminergic systems in the brain stem (14, 15, 26).

In the present study,we investigated the effect of nor-binaltorphimine (nor-BNI), a selective and long-acting KOR antagonist, on postoperative pain recovery in mice experiencing sleep disturbance induced by PAWW housing.

The results of the various experimental conditions adopted to assess the effects of sleep deprivation, pain and nor-BNI treatment on postoperative pain recovery are shown below. All data met the assumption of normality (P > 0.05), as assessed by the Kolmogorov–Smirnov test.

In experiment 1 (Figure 1A), during the light phase, one-way ANOVA revealed a significant group effect [F(2,28) = 19.60, P < 0.00001, η² = 0.58]. post hoc Bonferroni tests showed that the number of wheel rotations in the PAWW cage group (n = 9) was significantly higher (720 ± 135.7 rotations) than that in the regular cage group (n = 7, 11 ± 17.1 rotations; P < 0.001) (Figure 2A). Additionally, the PAWW cage group exhibited a significant increase in light phase rotations compared to baseline (192 ± 39.7 rotations; P < 0.001) (Figure 2A). During the dark phase, one-way ANOVA revealed a significant group effect [F(2,28) = 7.37, P < 0.01, η² = 0.35]. post hoc Bonferroni tests showed that the number of wheel rotations in the PAWW cage group (10,611 ± 1,801 rotations) was significantly lower compared to the regular cage group (16,564 ± 1,141 rotations; P < 0.05) (Figure 2B). Similarly, the number of dark phase rotations in the PAWW cage group also showed a significant decrease compared to baseline (17,403 ± 1,058 rotations; P < 0.01) (Figure 2B).

In experiment 2 (Figure 1B), in the sleep percentage evaluation by EEG/EMG, a repeated-measures two-way ANOVA revealed a significant effect of the within-group comparisons [F(1, 6) = 237.81, P < 0.0001, η² = 0.86], and the effect of the between-group comparisons was also significant [F(1, 6) = 14.11, P < 0.01, η² = 0.080], but the interaction between phase and group was not significant [F(1, 6) = 1.36, P = 0.29, η² = 0.0049]. post hoc Bonferroni tests revealed that the PAWW cage group (n = 4) exhibited a reduction in total sleep time (60.9 ± 1.4%) during the light phase, compared to the regular cage group (n = 4, 75.8 ± 5.2%; P < 0.01) (Figure 3A). There was no significant difference between the PAWW cage group (24.8 ± 1.7%) and the regular cage group (33.8 ± 1.0%) in total sleep time during the dark phase.

In the evaluation of the REM sleep percentage, a repeated-measures two-way ANOVA revealed that there were no significant effects of either phase [F(1, 6) = 2.00, P = 0.21, η² = 0.019] or group [F(1, 6) = 3.31, P = 0.12, η² = 0.33]. The interaction between phase and group was also not significant [F(1, 6) = 0.52, P = 0.50, η² = 0.0049]. post hoc Bonferroni tests revealed that the PAWW cage group also showed a lower percentage of REM sleep during the light phase (4.7 ± 0.8%) compared to the regular cage group (8.9 ± 2.1%), although this difference was not statistically significant (Figure 3B). During the dark phase, the percentage of REM sleep was 4.3 ± 0.5% in the PAWW cage group and 7.7 ± 1.9% in the regular cage group, with no significant difference between them.

In Experiment 3 (Figure 1C), following the plantar incision, pain behavior was assessed using the von Frey test. A repeated-measures two-way ANOVA revealed there were significant effects of groups [F(3, 28) = 55.64, p < 0.0001, η² = 0.35], and days [F(7, 196) = 12.90, p < 0.0001, η² = 0.15]. A significant interaction effect between days and groups was also observed [F(21, 196) = 6.80, p < 0.0001, η² = 0.20]. post hoc Bonferroni tests revealed that in the regular cage group (n = 8), the pain threshold in the ipsilateral hind paw decreased until post-surgery day 1, followed by recovery starting on day 3. In contrast, the PAWW cage group (n = 8) exhibited a prolonged decrease in pain threshold, with no statistically significant difference between the ipsilateral and contralateral hind paws two weeks after surgery (Figure 4A, Supplementary Table S1A).

In Experiment 4, nor-BNI was administered at three time points, followed by von Frey tests to evaluate the course of postoperative pain recovery (Figure 1D). A repeated-measures two-way ANOVA revealed there were significant effects of groups [F(3, 28) = 17.22, p < 0.0001, η² = 0.14], and days [F(7, 196) = 18.41, p < 0.0001, η² = 0.21]. A significant interaction effect between days and groups was also observed [F(21, 196) = 6.95, p < 0.0001, η² = 0.24]. post hoc Bonferroni tests revealed that in the regular cage group (n = 8), the pain threshold in the ipsilateral hind paw decreased until post-surgery day 3, followed by recovery starting on day 5. Similarly, the PAWW cage group (n = 8) exhibited a decrease in pain threshold up to day 3, with recovery beginning on day 5 (Figure 4B, Supplementary Table S1B).

Our findings demonstrate that housing mice in a PAWW cage induces sleep disturbances and delays recovery from postoperative pain following a plantar incision. Notably, nor-BNI, a KOR antagonist, alleviated the delayed recovery of the mechanical pain threshold.

In our study, PAWW housing significantly increased spontaneous activity during the light phase and decreased activity during the dark phase, disrupting circadian locomotor rhythms consistent with previous reports (7). EEG recordings confirmed a significant reduction in total sleep time during the light phase, with a decreasing trend also observed during the dark phase. The mismatch observed during the dark phase, where physical activity is reduced but sleep also tends to decrease, is not merely a sign of hyperactivity, but rather reflects a broader dysregulation of sleep homeostasis. Similar patterns have been reported in human studies, in which chronic stress leads to reductions in both sleep and physical activity (31).

Compared to traditional sleep deprivation models, PAWW offers a more physiological approach to mimicking chronic sleep disturbance (Table 1). The single platform-on-water model and bar rotation method effectively induce sleep deprivation, but cause excessive acute stress and severe physical impairment, limiting their feasibility in long-term studies (4–6, 32–34). The social defeat model induces fragmented sleep and decreased pain thresholds, but has low reproducibility for sleep disturbances and primarily relies on short-term acute stress exposure (35, 36). The PAWW model, in contrast, induces chronic sleep-wake cycle disruptions without excessive acute stress or severe physical distress. It effectively mimics the circadian rhythm dysregulation seen in human sleep disorders, with characteristic alterations in corticosterone secretion rhythms and weight loss despite increased appetite (7).

Poor sleep quality is a well-established risk factor for hyperalgesia and spontaneous pain (37–39). Several mechanisms may underlie this relationship, including opioid system dysfunction (12, 13), altered orexin signaling (18, 19), monoaminergic imbalance (serotonin, noradrenaline, dopamine) (16, 17), and immune and endocrine systems (40). Sleep disruption might delay tissue recovery, since sleep-wake cycles are essential regulators of immune function and tissue repair (41). In our study, PAWW-induced sleep disturbances prolonged postoperative pain recovery, further supporting this link. However, since sleep and stress are interdependent, their relative contributions remain uncertain. While measuring corticosterone levels could have provided insights into stress involvement, we opted not to include this measurement in this study, in order to minimize additional handling stress that could have confounded the results. Future research should include stress biomarkers (e.g., corticosterone, pro-inflammatory cytokines) to clarify the individual contributions of sleep disruption and stress-induced hyperalgesia. Exploring dual-targeted interventions may offer new therapeutic insights.

These findings may have clinical relevance for perioperative patients, who are highly susceptible to stress and sleep disturbances. Addressing sleep disturbances could be key to improving postoperative recovery. However, currently, only a few well-evidenced interventions have addressed sleep disorders in this context (42). Mice and humans share a 24 h circadian rhythm, and their sleep influences the nervous and endocrine systems similarly in both species (43). However, key differences exist: mice are nocturnal, have shorter sleep cycles (∼10–15 min between NREM and REM sleep), and exhibit polyphasic sleep patterns. These differences highlight the need for further clinical validation when translating findings from rodent models to human patients. Nonetheless, our results strongly suggest that targeting sleep disturbances might be a promising approach to improving postoperative pain recovery.

Nor-BNI is unlikely to have direct and immediate analgesic effects, since no improvement in pain thresholds was observed between 2 and 24 h post-administration. Previous studies also report a delayed onset of nor-BNI's effects (24, 44), likely due to indirect modulation of stress, sleep architecture, and pain processing (24, 25, 44, 45). Nor-BNI may also enhance wound healing by improving sleep and immune responses, both of which are closely linked to pain recovery processes (41). Although nor-BNI improved postoperative pain recovery, its underlying mechanisms remain uncertain. Given prior evidence that KOR antagonists normalize sleep in chronic pain models (8), further EEG analyses post-nor-BNI are needed to clarify whether sleep restoration mediates its effects. While we focused on KOR antagonism, it remains possible that nor-BNI's effects on pain recovery involve interactions with other neurotransmitter systems, including serotonin and dopamine (46, 47), both of which are essential for pain modulation. Tricyclic antidepressants, which enhance monoaminergic transmission, are widely used for chronic pain treatment. Thus, nor-BNI's effects may involve downstream interactions with monoaminergic systems. Future studies should quantify changes in these neurotransmitters following nor-BNI treatment and assess the therapeutic potential of combining KOR antagonists with monoamine modulators.

In addition to central mechanisms, peripheral KOR signaling may also modulate pain. KORs in peripheral sensory neurons have been implicated in nociceptive processing (48). While peripheral KOR agonists suppress inflammation and hypersensitivity, they may also contribute to opioid tolerance and hyperalgesia. Our study did not evaluate local KOR expression or inflammatory markers. Further investigation could elucidate the peripheral contributions of KOR antagonism. Because nor-BNI was administered intraperitoneally, it may have affected both central and peripheral KORs. Clarifying the site-specific contributions will require future studies employing localized drug delivery (e.g., intracerebral injection) or region-specific genetic approaches such as conditional knockouts.

While testing a KOR agonist could have provided further insights, our primary aim was to investigate the efficacy of KOR blockade. Nor-BNI was selected due to its long-lasting action, ensuring sustained receptor inhibition. Previous literature has reported that KOR activation can induce mechanical hypersensitivity (24, 25), and KOR agonists can disrupt sleep (10), suggesting that KOR agonists may exacerbate pain recovery. These findings support the rationale for targeting KOR antagonism in our experimental context. We acknowledge this as an important future direction, and we need to explore the differential effects of KOR agonists vs. antagonists on sleep-related pain outcomes in subsequent studies. Further investigations are required to determine whether the observed effects are specific to KOR blockade or involve additional neurotransmitter systems. Future studies incorporating KOR agonists, monoamine modulators, and glutamate antagonists may clarify the complex mechanisms underlying sleep-related pain enhancement.

The limitations of this study include the short duration of observation for postoperative pain recovery and sleep disturbances. While we demonstrated that sleep disruption delays pain recovery, we could not assess long-term consequences, such as chronic pain development, due to the limited observation period. Future studies with extended monitoring periods are needed to evaluate potential long-term effects. Additionally, although we observed a trend toward REM sleep reduction in the PAWW group, this difference did not reach statistical significance. This suggests that a larger sample size or more precise EEG analysis methods might be necessary for better assessment of REM sleep alterations and their potential impact on pain recovery. Furthermore, while nor-BNI improved postoperative pain recovery, the underlying mechanisms remain unclear. Future studies should include EEG assessments following nor-BNI treatment, along with investigations into other potential mechanisms, such as monoaminergic or neuroimmune interactions, to elucidate the precise role of KOR antagonism in postoperative pain recovery. Additionally, our study utilized only male mice to minimize variability related to the estrous cycle and to maintain consistency with prior PAWW model studies. Future research is warranted to determine whether similar effects are observed in female mice, which may provide further insight into sex-specific mechanisms in sleep–pain interactions and KOR signaling.

In conclusion, our study demonstrates that the PAWW model, which mimics sleep disorders in modern society, delays recovery from postoperative pain. Importantly, our findings suggest that KOR antagonism could be a promising option to prevent or treat prolonged pain induced by sleep disorders in perioperative patients. This potential application of KOR antagonists in perioperative care is the key takeaway of our research.