Respiratory cerebrospinal fluid flow is driven by the thoracic and lumbar spinal pressures

Key points Respiration plays a key role in the circulation of cerebrospinal fluid (CSF) around the central nervous system. During inspiration increased venous return from the cranium is believed to draw CSF rostrally. However, this mechanism does not explain why CSF has also been observed to move caudally during inspiration. We show that during inspiration decreased intrathoracic pressure draws venous blood from the cranium and lumbar spine towards the thorax. We also show that the abdominal pressure was associated with rostral CSF displacement. However, a caudal shift of cervical CSF was seen with low abdominal pressure and comparably negative intrathoracic pressures. These results suggest that the effects of epidural blood flow within the spinal canal need to be considered, as well as the cranial blood volume balance, to understand respiratory-related CSF flow. These results may prove useful for the treatment of CSF obstructive pathology and understanding the behaviour of intrathecal drug injections. It is accepted that during inspiration, cerebrospinal fluid (CSF) flows rostrally to compensate for decreased cranial blood volume, caused by venous drainage due to negative intrathoracic pressure. However, this mechanism does not explain observations of caudal CSF displacement during inspiration. Determining the drivers of respiratory CSF flow is crucial for understanding the pathophysiology of CSF flow disorders. To quantify the influence of respiration on CSF flow, real-time phase-contrast magnetic resonance imaging (MRI) was used to record CSF and blood flow, while healthy subjects (5:5 M:F, 25-50 years) performed either a brief expiratory or inspiratory effort between breaths. Transverse images were taken perpendicular to the spinal canal in the middle of the C3 and L2 vertebrae. The same manoeuvres were then performed after a nasogastric pressure catheter was used to measure the intrathoracic and abdominal pressures. During expiratory-type manoeuvres that elevated abdominal and intrathoracic pressures, epidural blood flow into the spinal canal increased and CSF was displaced rostrally. With inspiratory manoeuvres, the negative intrathoracic pressure drew venous blood from C3 and L2 towards the thoracic spinal canal, and cervical CSF was displaced both rostrally and caudally, despite the increased venous drainage. Regression analysis showed that rostral displacement of CSF at both C3 (adjustedR(2) = 0.53;P < 0.001) and L2 (adjustedR(2) = 0.38;P < 0.001) were associated with the abdominal pressure. However, with low abdominal pressure and comparably negative intrathoracic pressure, cervical CSF flowed caudally. These findings suggest that changes in both the cranial and spinal pressures need to be considered to understand respiratory CSF flow.