The time course of ineffective sham-blinding during low-intensity (1 mA) transcranial direct current stimulation

Studies using transcranial direct current stimulation (tDCS) typically compare an active protocol relative to a shorter sham (placebo) protocol. Both protocols are presumed to be perceptually identical on the scalp, and thus represent an effective method of delivering double-blinded experimental designs. However, participants often show above-chance accuracy when asked which condition involved active/sham retrospectively. We assessed the time course of sham-blinding during active and sham tDCS. We predicted that participants would be aware that the current is switched on for longer in the active versus sham protocol. Thirty-two adults were tested in a preregistered, double-blinded, within-subjects design. A forced-choice reaction time task was undertaken before, during and after active (10 min 1 mA) and sham (20 s 1 mA) tDCS. The anode was placed over the left primary motor cortex (C3) to target the right hand, and the cathode on the right forehead. Two probe questions were asked every 30 s: "Is the stimulation on?" and "How sure are you?". Distinct periods of non-overlapping confidence intervals were identified between conditions, totalling 5 min (57.1% of the total difference in stimulation time). These began immediately after sham ramp-down and lasted until the active protocol had ended. We therefore show a failure of placebo control during 1 mA tDCS. These results highlight the need to develop more effective methods of sham-blinding during transcranial electrical stimulation protocols, even when delivered at low-intensity current strengths.