Tetraplegia is associated with increased hypoxic ventilatory response during nonrapid eye movement sleep

Abstract

People with cervical spinal cord injury (SCI) are likely to experience chronic intermittent hypoxia while sleeping. The physiological effects of intermittent hypoxia on the respiratory system during spontaneous sleep in individuals with chronic cervical SCI are unknown. We hypothesized that individuals with cervical SCI would demonstrate higher short- and long-term ventilatory responses to acute intermittent hypoxia (AIH) exposure than individuals with thoracic SCI during sleep. Twenty participants (10 with cervical SCI [9 male] and 10 with thoracic SCI [6 male]) underwent an AIH and sham protocol during sleep. During the AIH protocol, each participant experienced 15 episodes of isocapnic hypoxia using mixed gases of 100% nitrogen (N₂ ) and 40% carbon dioxide (CO₂ ) to achieve an oxygen saturation of less than 90%. This was followed by two breaths of 100% oxygen (O₂ ). Measurements were collected before, during, and 40 min after the AIH protocol to obtain ventilatory data. During the sham protocol, participants breathed room air for the same amount of time that elapsed during the AIH protocol and at approximately the same time of night. Hypoxic ventilatory response (HVR) during the AIH protocol was significantly higher in participants with cervical SCI than those with thoracic SCI. There was no significant difference in minute ventilation (V_.E. ), tidal volume (V_.T. ), or respiratory frequency (f) during the recovery period after AIH in cervical SCI compared to thoracic SCI groups. Individuals with cervical SCI demonstrated a significant short-term increase in HVR compared to thoracic SCI. However, there was no evidence of ventilatory long-term facilitation following AIH in either group.

Keywords: acute intermittent hypoxia; hypoxia; long-term facilitation; plasticity; spinal cord injury.

FIGURE 1

A diagram of the study protocol.

FIGURE 2

A representative polygraph recording of the intermittent hypoxia protocol illustrates respiratory changes and demonstrates that isocapnia was maintained during the AIH episodes. This tracing is from a female participant with a thoracic spinal cord injury. Her apnea‐hypopnea index was 8.6, and her oxygen desaturation index was 3.0.

FIGURE 3

Individual ventilatory parameters across study conditions (pre‐hypoxia, during acute intermittent hypoxia [AIH], and post‐hypoxia), presented as absolute values, in participants with cervical and thoracic SCI. (a) Minute ventilation (V _.E.), (b) tidal volume (V _.T.), (c) frequency (f), (d) oxygen saturation (SaO₂), and (e) end‐title CO₂ (P_ETCO₂). Data are presented individually and with the group mean ± SD. A 2 × 2 repeated measures ANOVA in SPSS was used to compare differences within and between groups. The variables in this model included study conditions (baseline, AIH, recovery) and cervical and thoracic SCI groups. *Significantly different compared with baseline (p < 0.001). SCI, spinal cord injury.

FIGURE 4

Individual ventilatory parameters across sham study conditions (baseline and recovery), presented as absolute values, in participants with cervical and thoracic SCI. (a) Minute ventilation (V _.E.), (b) tidal volume (V _.T.), (c) frequency (f), (d) oxygen saturation (SaO₂), and (e) end‐title CO₂ (P_ETCO₂). Data are presented individually and with the group mean ± SD. A 2 × 2 repeated measures ANOVA in SPSS was used to compare differences within and between groups. The factors used in the model included study condition (baseline and recovery) and cervical and thoracic SCI groups. No significance or interactions were found between or within groups. SCI, spinal cord injury.

FIGURE 5

Individual data to illustrate hypoxic ventilatory response (HVR) during acute intermittent hypoxia episodes in participants with cervical (n = 10) and thoracic (n = 8) SCI. (a) Total HVR, (b) early HVR episodes, and (c) late HVR episodes. Data are presented as the mean ± SD. Mixed linear modeling in SAS was used. The variables used in this model included early and late HVR and cervical and thoracic SCI groups. *Significantly different compared with cervical p < 0.05. SCI, spinal cord injury.

FIGURE 6

Ventilatory changes presented as the average (%) change from baseline in (a) minute ventilation (V _.E.), (b) tidal volume (V _.T.), and (c) frequency (f) for individual participants grouped by injury level (cervical or thoracic SCI) and recovery from AIH or corresponding sham protocol. Individual data are presented for each participant with the group mean ± SD. Mixed linear modeling in SAS was used. The variables used in the model included recovery from AIH versus corresponding sham exposure and cervical and thoracic SCI groups (within‐subjects, between groups, and within‐subjects by between‐group interaction). No statistically significant differences were found. AIH, acute intermittent hypoxia; SCI, spinal cord injury.