“[This study presents] the possibility that nocturnal hypoxemia and sleep fragmentation are distinct sleep-disordered breathing phenotypes that may exert potentially differential effects on pain sensitivity and symptom expression.”
THE bidirectional interaction between sleep and pain is now a well-recognized clinical phenomenon supported by experimental studies and large-scale epidemiological investigations, which convincingly demonstrate that sleep disruption induces hyperalgesia and that poor sleep is a risk factor for the development of chronic pain, for example, see studies by Smith et al., Roehrs et al., and Gupta et al.1–3 Despite decades of work demonstrating that pain and sleep systems mutually influence each other, the mechanisms accounting for these associations remain largely a mystery. Doufas et al.’s naturalistic study4 in the Journal, reporting that nocturnal hypoxemia is associated with sleep-related pain, highlights the possibility that brainstem regions regulating both sleep and respiration may provide an underappreciated, regional nexus to guide future experimental investigations seeking to unravel the mechanisms underlying sleep–pain interactions.
In the Doufas et al.’s study4 published in the Journal, nocturnal hypoxemia was found to be correlated with self-reported morning headache, headache disrupting sleep, and chest pain while in bed. Importantly, these effects were independent of sleep fragmentation, which has been shown in numerous experiments to induce hyperalgesia, for example, see study by Smith.5 The Doufas et al.’s findings4 are consistent with recent work demonstrating that positive airway pressure therapy decreases laboratory measures of pain sensitivity in pain-free patients with severe sleep apnea,6 as well as, a study demonstrating that low nocturnal oxygen saturation levels were associated with the presence of morning headache in patients with moderate to severe sleep apnea.7
Intriguingly and perhaps even critically important, the findings seem to conflict with work demonstrating that patients with chronic temporomandibular joint disorder pain and mild sleep apnea have diminished laboratory pain sensitivity compared those with temporomandibular joint disorder without sleep apnea.8 Another study on patients with suspected sleep-disordered breathing similarly found that those meeting formal criteria for a headache disorder exhibited higher mean nocturnal Spo2 values and lower apnea hypopnea indices compared with patients not meeting criteria for headache disorder.9 Even recent work by the authors of the current article in the Journal reported that hypoxemia during sleep was associated with a decreased sensitivity to cold pain during remifentanil infusion in patients at risk for sleep apnea.10 This study also found that insulin growth factor binding protein-1 levels, a marker of hypoxia, were associated with decreased baseline pain sensitivity. The reasons for the seemingly conflicting findings are unclear and will require subsequent research, but perhaps the extant data suggest the possibility that pain chronicity and/or sleep-disordered breathing severity may moderate the direction of the association between sleep-related hypoxemia and pain.
With respect to chronic pain, this type of effect has precedent, as pain chronicity has similarly been found to moderate the association between blood pressure and pain sensitivity.11 Specifically, resting blood pressure is inversely associated with pain sensitivity in individuals free of chronic pain, but this relationship is absent in patients with chronic pain. This suggests a dysfunction of these interacting systems as pain becomes chronic. It should be noted, however, that pain duration was not reported in the study by Doufas et al.4 in the Journal, and therefore, it is unclear whether pain in this sample might best be classified as acute or chronic. Moreover, the pain reported is episodic and highly specific to the sleep period, making it rather atypical. It is also unclear from the previous sleep apnea studies reporting a positive association between sleep-disordered breathing and pain or hyperalgesia whether patients with other episodic and/or persistent chronic pain were included in the samples. We suspect that the samples were not comprised primarily of patients with chronic pain, because they were not preselected for specific chronic pain disorders. One interesting possibility to consider regarding the apparent reversal of the association between hypoxemia and pain sensitivity in chronic pain samples is that this reversal might serve as a potential biomarker for the transition between acute and chronic pain. Testing this hypothesis, of course, would require longitudinal studies measuring sleep in acute pain patients as they transition to chronic pain disorders.
In addition to the possibility that the association between Sao2 and pain sensitivity might be altered as a function of pain chronicity, the severity of sleep-disordered breathing might also moderate the relationship. All three studies8–10 demonstrating a positive association between Sao2 during sleep and pain or pain sensitivity involved samples of minimal to mild sleep-disordered breathing, whereas studies reporting an inverse association between nocturnal Sao2 and clinical pain or pain sensitivity included patients with moderate to severe apnea or hypopnea indices. The implications of findings by Doufas et al.4 published in the Journal, therefore, would be enhanced by determining whether a subset of subjects reporting at least some pain and only mild or no sleep apnea demonstrated positive correlations between pain frequency and Sao2. Interestingly, recent longitudinal work has found symptoms suggestive of sleep apnea, for example, loud snoring and daytime sleepiness predicted the development of chronic temporomandibular joint pain disorder.12 Unfortunately, this study did not measure the severity of sleep-disordered breathing or associations between sleep-related hypoxemia and pain sensitivity, but it does underscore the importance of considering nocturnal hypoxemia as a potential mechanism for neuroplastic changes that lead to the development of persistent pain.
Aside from contributing to an emerging literature linking pain to sleep-related hypoxemia, one of the most important contributions of the Doufas et al.’s study4 in the Journal is the emphasis on the possibility that nocturnal hypoxemia and sleep fragmentation are distinct sleep-disordered breathing phenotypes that may exert potentially differential effects on pain sensitivity and symptom expression. Researchers studying sleep–pain interactions have rarely considered this possibility. To disentangle the effects of sleep fragmentation from the effects of sleep-related hypoxemia and inflammation, Doufas et al. studied the effects of nocturnal Sao2 indices on clinical pain report while controlling for six polysomnographic parameters such as: (1) a respiratory disturbance index associated with no change in Sao2, (2) overall arousal index, (3) % Rapid Eye Movement sleep, (4) % Stage 1, (5) % Stage 2, and (6) % Stage 3. They also controlled for inflammatory markers, using a data-reducing principal components analysis. Methodologically, the data reduction procedure may be a quite useful approach for subsequent studies involving models with many potential multiple confounders. It is unclear, however, why total sleep time, wake after sleep onset time, and sleep efficiency were not included in the models. Many studies demonstrate that these traditional sleep continuity variables are robustly linked with hyperalgesia, and perhaps to a greater extent than brief electroencephalographically recorded arousals without frank wakefulness.13,14 Future studies attempting to tease apart the unique effects of hypoxemia from sleep continuity should include measures of fragmentation (arousals) and sleep duration (total sleep time), wake after sleep onset time, and/or sleep efficiency.
Although the reasons for the seemingly conflicting findings between sleep-related oxygen saturation and pain are unclear and perplexing at present, the answers may lie in brain stem regions, such as the nucleus tractus solitarius, that regulate both nociception and breathing, for example, see study by Lewis.15 Future work in this regard is needed and likely to inform the pathophysiology of both chronic pain and sleep-disordered breathing. Continued study on these relationships will also contribute to a better understanding of the complex symptomatology and morbidity associated with these commonly cooccurring disorders.