Reply to Khatoun et al.: Speculation about brain stimulation must be constrained by observation

Khatoun et al. (1) are concerned that our paper (2) did not “establish the field strength required for tACS [transcranial alternating current stimulation] to cause neural entrainment.” This is true, but in our defense there is no evidence that any such threshold exists: The effects of electric fields on neuronal membrane potentials have been shown to be continuous and linearly related to the electric field strength (3, 4); a similar dose–response relationship is seen in the spiking output (5). Nevertheless, the issue of a minimum field strength forms the basis for Khatoun et al.’s (1) comment.

The specific concern Khatoun et al. (1) raise is one that we address extensively in our paper. In short, they argue that the field strength near the site of our recordings (in the hippocampus and basal ganglia) might have been below the threshold required for tACS to entrain neurons. In that case, the neural effects we observed could only have resulted from indirect entrainment via some other brain area that was exposed to a stronger field. Based on our control experiments (2), we considered this possibility to be unsatisfactory because it required a number of unwarranted assumptions. Namely, one has to imagine that the brain regions providing indirect stimulation project exclusively to our recordings sites, so that the entrainment becomes undetectable when the stimulating electrodes are moved or data are instead collected from another brain region. Furthermore, one would have to suppose that the hypothetical indirect pathway was capable of maintaining the frequency-specific stimulation effects we observed and was insensitive to retinal (2) and somatosensory (6) inputs, which did not affect entrainment in our experiments. This combination of properties would entail a strange coincidence, since the positioning of the electrodes on the scalp was not determined by anatomical proximity to the recording sites and was in fact different for the 2 animals.

In any event, the point is ultimately moot, since Khatoun et al. (1) suggest a simple, empirical test of their claim: Entrainment should disappear at weaker stimulation amplitudes, when indirect effects are diminished. We recently performed exactly this experiment, recording from the same sites as in our previous paper, with the same stimulating montages, but using half the stimulation amplitude: 1 mA instead of 2 mA (6). Moreover, another group has recently studied neural entrainment in monkeys when the stimulating current was halved again to 0.5 mA (5). In all cases, the entrainment persisted at these weaker amplitudes. Moreover, these experiments measured the electric fields in each monkey and found them to be similar to those measured in human cortex (7) and predicted for human deep-brain areas (8). Thus, it is not true that reduced stimulation amplitude is necessary to “make a realistic approximation of human tACS,” and Khatoun et al.’s (1) conjecture is ultimately refuted by experimental data.