Sleep problems in spinal cord injury

Introduction

Spinal cord injury (SCI) is a significant global cause of mortality and morbidity. In 2007, the global incidence was estimated to be 23 cases per million per annum, ranging from 15 cases per million in Australia to 29 cases per million in sub-Saharan Africa ( ).

Most people with SCI sustain their injuries in their second or third decade of life. Because of improvements in both the acute management and long-term supportive care, their life expectancy after injury has increased significantly. As such, secondary diseases and impairments that are a direct consequence of the SCI have significant impacts for many years. Sleep disorders, especially sleep-disordered breathing (SDB), periodic and/or restless leg syndromes (PLM/RLS), and circadian rhythm disorders, are all prevalent disorders after SCI. This chapter focuses on these particular sleep disorders because while insomnia, generalized fatigue, and other sleep-associated disorders affect people living with SCI just as the general population, the increased prevalence and altered pathogenicity due to the spinal lesion per se of SDB, PLMs, and circadian rhythm disorders after SCI deserves special consideration ( ). Population prevalence, known physiological differences in causation, response to therapies, and longer-term consequences of sleep disorders in these sleep disorders after SCI are discussed.

Sleep-disordered breathing

Sleep-disordered breathing is an umbrella term that includes obstructive sleep apnea (OSA), central sleep apnea (CSA), and sleep-related hypoventilation. OSA is repeated partial (narrowing/hypopnea) or complete (occlusion/apnea) closure and interruption to ventilation at the level of the upper airway during sleep. The drive to breath is generally maintained during OSA, whereas CSA is characterized by periods of diminished or absent ventilatory drive with associated reductions in ventilation. Both OSA and CSA result in periodic oxygen desaturation and fragmentation of sleep typically resulting excessive daytime sleepiness. Hypoventilation refers to periods during which arterial carbon dioxide homeostasis is not maintained because minute ventilation inadequately matches metabolism. The consequence of hypoventilation is a raised arterial carbon dioxide. All three disorders may exist separately or together in the same person and to various degrees over, and indeed within, sleep periods ( ). The most common risk factors for OSA in the general population are being male and having a raised body mass index (BMI).

There is very little published research into respiratory sleep disorders in paraplegia. While little data reduces our ability to be certain, it is believed that the pathophysiology, risk factors and prevalence of respiratory sleep disorders in those living with paraplegia is not substantially different to the general population. Conversely, tetraplegia has been studied extensively. Despite substantial overall respiratory muscle weakness, ( ) tetraplegia is not routinely associated with hypoventilation, but rather with OSA ( ). The upper airway muscles derive motor innervation from the cranial nerves, predominantly the hypoglossal, rather than from the spinal segmental nerves. Hence, an injury that disrupts, for example, C6 should not have any effect on the genioglossus. Yet while OSA is highly prevalent ( Fig. 1 ), hypoventilation appears rare at a population level. In a recent large, multi-center randomized controlled trial of continuous positive airway pressure (CPAP) for SDB in acute tetraplegia ( ), only eight of the 212 otherwise eligible participants (3.8%) were excluded due to hypoventilation (PaCO ₂ > 45 mmHg) ( ). Hypoventilation has been reported, particularly at sleep onset in physiological studies ( ) and in association with a reduced ventilatory reserve for carbon dioxide ( ), but there is little evidence for increased hypoventilation prevalence across all those living with tetraplegia. Clinically however, if a raised carbon dioxide is detected in a particular patient, this relatively rare sign demands immediate clinical attention.

Prevalence of sleep-disordered breathing after tetraplegia

The prevalence of disordered breathing during sleep is estimated to be between 9% and 38%, and increases with age ( ). In previous cross-sectional population studies of sleep disorders in patients with tetraplegia, the prevalence of SDB has been estimated to be between 27% and 97% ( ; ). However, a recent meta-analysis of the published population estimates in tetraplegia presented the pooled estimates at various cut-off severities ( ). Twelve articles across 20 years were included, nine of which presented data amenable to meta-analysis (combined n = 630). The reported SDB prevalence rates from the 12 studies ranged from 46% to 97%. Following meta-analysis the mean prevalence of at least mild (apnea hypopnea index AHI ≥ 5 events per hour), moderate (AHI ≥ 15), and severe (AHI ≥ 30) SDB were 83% (95% CI = 73–91), 59% (46–71), and 36% (26–46), respectively. As illustrated in Fig. 1 , the prevalence of SDB after cervical SCI is clearly elevated at all levels of severity compared with general population estimates ( ).

Sleep-disordered breathing is an acute and direct consequence of cervical SCI

As noted above, the upper airway is innervated directly from the brain via the cranial nerves, not by the spinal segmental nerves. Notwithstanding this, an acute cervical spinal cord injury appears to directly compromise upper airway patency during sleep within weeks of the injurious event. Berlowitz et al. studied all patients presenting to the specialized SCI unit in Melbourne, Australia, over an 18-month period and undertook full, bedside polysomnography to determine when SDB appeared after injury ( ). As illustrated in Fig. 2 , approximately 10% of the cohorts were at increased risk of the SDB prior to injury as estimated by a prediction algorithm that incorporates signs and symptoms. Within 2 weeks of injury, the measured prevalence of clinically important disease had increased to 60% and peaked at over 80% 3 months after initial injury. Most people with SDB have an insidious onset of the disorder over a number of years as they age and gain weight. In contrast, tetraplegia appears to be the only model of acute SDB in humans. As such, the pathophysiology in acute tetraplegia is likely different from the general population and is not yet fully understood ( ).