J Neurophysiol (April 1, 2003). 10.1152/jn.00915.2002
Submitted on Submitted 11 October 2002; accepted in final form 7 December 2002

Millivolt-Scale DC Shifts in the Human Scalp EEG: Evidence for a Nonneuronal Generator

 ¹Department of Biosciences, University of Helsinki, 00014; and  ²Department of Child Neurology, Hospital for Children and Adolescents, University Hospital of Helsinki, 00029, Finland

Voipio, Juha, Pekka Tallgren, Erkki Heinonen, Sampsa Vanhatalo, and Kai Kaila. Millivolt-Scale DC Shifts in the Human Scalp EEG: Evidence for a Nonneuronal Generator. J. Neurophysiol. 89: 2208-2214, 2003. Slow shifts in the human scalp-recorded EEG, including those related to changes in brain CO₂ levels, have been generally assumed to result from changes in the level of tonic excitation of apical dendrites of cortical pyramidal neurons. We readdressed this issue using DC-EEG shifts elicited in healthy adult subjects by hypo- or hypercapnia. A 3-min period of hyperventilation resulted in a prompt negative shift with a rate of up to 10 µV/s at the vertex (Cz) and an extremely steep dependence (up to 100 µV/mmHg) on the end-tidal Pco₂. This shift had a maximum of up to 2 mV at Cz versus the temporal derivations (T3/T4). Hyperventilation-like breathing of 5% CO₂-95% O₂, which does not lead to a significant hypocapnia, resulted in a near-complete block of the negative DC shift at Cz. Hypoventilation, or breathing 5% CO₂ in air at normal respiratory rate, induced a positive shift. The high amplitude of the voltage gradients on the scalp induced by hyperventilation is not consistent with a neuronal origin. Instead, the present data suggest that they are generated by extracortical volume currents driven by a Pco₂-dependent potential difference across epithelia separating the cerebrospinal fluid and blood. Since changes in respiratory patterns and, hence, in the level of brain Pco₂, are likely to occur under a number of experimental conditions in which slow EEG responses have been reported (e.g., attention shifts, preparatory states, epileptic seizures, and hypoxic episodes), the present results call for a thorough reexamination of the mechanisms underlying scalp-recorded DC-EEG responses.