The results of Taylor et al's study ^{Reference Taylor, Firbank, Barnett, Pearce, Livingstone and Mosimann1} are intriguing, shedding light on the pathogenesis of visual hallucinations in dementia with Lewy bodies.

However, I have some concerns about its methodology. The authors did not adopt the rather restrictive (and currently used) definition of phosphene threshold (i.e. the lowest stimulus intensity required to elicit phosphenes in 50% of trials), but used a much lower value (25%) to minimise the number of participants who might not respond. Moreover, to ensure inclusion of all individuals in analyses, participants who did not report phosphenes up to 100% stimulator output were arbitrated a phosphene threshold of 101%. The authors therefore assumed that not reporting phosphenes meant having a threshold above 100% because of an insufficient magnetic field strength from the stimulator to induce phosphenes in these individuals. However, as far as I know, to date there is no evidence definitely demonstrating such an assumption.

As a matter of fact, in most published studies of phosphene thresholds a certain number of participants do not experience phosphenes even with a maximum stimulator output. There are some reasons which may (partially) explain such a phenomenon.

First, it is possible that owing to methodological difficulties in mapping phosphene thresholds over each square millimetre of the occipital skull, the correct point for stimulation may not be identified in each participant.

Second, unlike primary motor cortex, primary visual cortex (calcarine fissure) is deeply located, lying in the mid-sagittal plane, so that the magnetic field strength applied over the entire skull may be insufficient to reach and stimulate the visual cortex. Regarding this aspect, it is noteworthy to consider that Taylor et al used a figure-of-eight coil (and not a circular one), which, although it is much more selective and has a higher spatial accuracy, stimulates a smaller cortical area, ^{Reference Cohen, Roth, Nilsson, Dang, Panizza and Bandinelli2,Reference Hallett3} and may generate, at least theoretically, a weaker electric current, resulting in a lower probability of evoking phosphenes.

Finally, as the authors stated, every millimetre the surface cortex is away from the stimulating coil, approximately an additional 3% of the maximum power output is required to induce an equivalent level of brain stimulation at the motor cortex (although no similar data on visual cortex stimulation are available in the literature). Such an aspect needs to be taken into account not only with regard to occipital cortical atrophy in affected patients compared with healthy controls, but also with regard to the fact that, because the lower portion of the visual cortex representing the upper visual field is farther from the scalp (as observed in magnetic resonance imaging), it is more difficult to elicit phosphenes with transcranial magnetic stimulation in the upper than in the lower visual field. ^{Reference Tani, Hirata, Motoki, Saitoh, Yanagisawa and Goto4} Although in the study an adjusted phosphene threshold ratio was performed to account for possible group differences in atrophy, it is not clear whether other aspects (anatomical differences in skull thickness and portion of visual cortex stimulated) were considered.

In the light of the above, I think that the authors should have performed a statistical analysis of phosphene threshold including only those participants in whom phosphenes were actually induced.