Hi all. After reviewing the paper, I believe there are some key factors that could help to explain the similarity between the results of the sham group and the treatment group.
– Sham TMS is not ‘nothing’:
In well-designed TMS studies, a number of sham techniques may be used, but all of them involve introducing a low level of pulsed magnetic field to the brain. This is to ensure that a ‘TMS-like’ sensation is still achieved in the sham group. The key difference between sham TMS and treatment TMS lies in the strength of the magnetic field, with sham techniques endeavouring to introduce a lower-strength magnetic field than the treatment TMS technique.
– The TMS coil in this study:
This study used a MagVenture ‘Cool B-65 A/P’ coil, which has an internal mechanism for generating a sham magnetic pulse. The coil has two sides – ‘A’ for ‘Active’, ‘P’ for ‘Placebo’. The coil is flipped between the ‘A’ side during treatment pulses, and the ‘P’ side during sham pulses. The strength of the P-field in this particular coil has been measured at 7.72% of the maximum field strength of the A-field, while the field shape remains comparable for both A and P fields [1].
– Motor Threshold variability between individuals during treatment pulses, and non-variability of the P-field during sham pulses:
Motor Threshold (MT) refers to the level of magnetic pulse strength required to elicit a motor response (for example: a twitching of the finger). Once this MT has been established, it is used as an approximate guide to the appropriate pulse strength delivered to the treatment site of the brain, which in this case was the left pre-frontal cortex.
The intended intensity of treatment pulses in this study was stated as ‘120% of MT’, however, it is possible that not all participants received this high level of treatment pulse intensity, as individual tolerance to pulse strength varies. Whilst the intensity of A-field pulses in the treatment group typically varies between individuals, the intensity of the P-field pulses in the sham group does not, as an individual’s response to the sham TMS is assumed to be negligible.
– Non-linear responses and plateau effects:
An individual’s brain may not necessarily respond linearly to an increased pulse strength, as the brain has natural defence mechanisms to avoid over-stimulation. At low-to-mid pulse strengths, the ‘stimulus-response curve’ may be steeper than at higher pulse strengths [2, 3], due to these natural defence mechanisms. This may mean that at the lower end of the stimulus-response curve, the response is exaggerated, whilst at the higher end of the stimulus-response curve, the response flattens out, or ‘plateaus’. In practical terms, this may mean that the difference in clinical outcome between pulses delivered at, say, 110% and 120% of MT is negligible, while the difference in clinical outcome between pulses delivered at, say, 50% and 60% of MT is significant.
– Putting all of this together:
Sham TMS is not a placebo in the traditional medical sense, but a weaker magnetic field with the potential to induce electrical stimulation within the brain. The P-field was measured at 7.72% of the A-field in the same coil. This is not a nil magnetic field in the sham group, and may well be contributing to neuronal stimulation.
In the treatment group, the intensity of pulses was variable because it depended on individual participants’ MT. In the sham group, the intensity of pulses was fixed, meaning every participant in the sham group received the same level of magnetic pulse stimulation, i.e., 7.72% of the A-field maximum. This may have contributed to the consistency of depression-lifting results seen in the sham group.
The non-linear response of the brain to stimuli strength may have created a ‘diminishing returns’ effect with increased pulse intensity in the treatment (A-field) group. This effect would not have been present in the sham (P-field) group, as the sham-pulse intensity remained constant and low throughout the study.
When these factors are combined, it becomes easier to understand why the outcomes in the sham group were closer to those in the treatment group than might have been expected. The low-level stimulation from the sham coil, the variability in treatment intensity, and the non-linear stimulus-response curve all contribute to the complexity of the overall picture in this study, and help to explain the counter-intuitive results seen here, in which the sham group’s responses approached those of the treatment group.
*[3] is included for illustration purposes only, as it clearly shows the kinds of non-linear stimulus-response curves and plateau effects I am referring to above.
Note:
*While the P-field strength is measured as 7.72% of the A-field’s maximum output, this percentage does not directly correspond to the Motor Threshold (MT) percentages used to determine individual treatment intensity. The 7.72% refers to the coil’s maximum capacity rather than a personalised stimulation level based on MT.
Hi all. After reviewing the paper, I believe there are some key factors that could help to explain the similarity between the results of the sham group and the treatment group.
– Sham TMS is not ‘nothing’:
In well-designed TMS studies, a number of sham techniques may be used, but all of them involve introducing a low level of pulsed magnetic field to the brain. This is to ensure that a ‘TMS-like’ sensation is still achieved in the sham group. The key difference between sham TMS and treatment TMS lies in the strength of the magnetic field, with sham techniques endeavouring to introduce a lower-strength magnetic field than the treatment TMS technique.
– The TMS coil in this study:
This study used a MagVenture ‘Cool B-65 A/P’ coil, which has an internal mechanism for generating a sham magnetic pulse. The coil has two sides – ‘A’ for ‘Active’, ‘P’ for ‘Placebo’. The coil is flipped between the ‘A’ side during treatment pulses, and the ‘P’ side during sham pulses. The strength of the P-field in this particular coil has been measured at 7.72% of the maximum field strength of the A-field, while the field shape remains comparable for both A and P fields [1].
– Motor Threshold variability between individuals during treatment pulses, and non-variability of the P-field during sham pulses:
Motor Threshold (MT) refers to the level of magnetic pulse strength required to elicit a motor response (for example: a twitching of the finger). Once this MT has been established, it is used as an approximate guide to the appropriate pulse strength delivered to the treatment site of the brain, which in this case was the left pre-frontal cortex.
The intended intensity of treatment pulses in this study was stated as ‘120% of MT’, however, it is possible that not all participants received this high level of treatment pulse intensity, as individual tolerance to pulse strength varies. Whilst the intensity of A-field pulses in the treatment group typically varies between individuals, the intensity of the P-field pulses in the sham group does not, as an individual’s response to the sham TMS is assumed to be negligible.
– Non-linear responses and plateau effects:
An individual’s brain may not necessarily respond linearly to an increased pulse strength, as the brain has natural defence mechanisms to avoid over-stimulation. At low-to-mid pulse strengths, the ‘stimulus-response curve’ may be steeper than at higher pulse strengths [2, 3], due to these natural defence mechanisms. This may mean that at the lower end of the stimulus-response curve, the response is exaggerated, whilst at the higher end of the stimulus-response curve, the response flattens out, or ‘plateaus’. In practical terms, this may mean that the difference in clinical outcome between pulses delivered at, say, 110% and 120% of MT is negligible, while the difference in clinical outcome between pulses delivered at, say, 50% and 60% of MT is significant.
– Putting all of this together:
Sham TMS is not a placebo in the traditional medical sense, but a weaker magnetic field with the potential to induce electrical stimulation within the brain. The P-field was measured at 7.72% of the A-field in the same coil. This is not a nil magnetic field in the sham group, and may well be contributing to neuronal stimulation.
In the treatment group, the intensity of pulses was variable because it depended on individual participants’ MT. In the sham group, the intensity of pulses was fixed, meaning every participant in the sham group received the same level of magnetic pulse stimulation, i.e., 7.72% of the A-field maximum. This may have contributed to the consistency of depression-lifting results seen in the sham group.
The non-linear response of the brain to stimuli strength may have created a ‘diminishing returns’ effect with increased pulse intensity in the treatment (A-field) group. This effect would not have been present in the sham (P-field) group, as the sham-pulse intensity remained constant and low throughout the study.
When these factors are combined, it becomes easier to understand why the outcomes in the sham group were closer to those in the treatment group than might have been expected. The low-level stimulation from the sham coil, the variability in treatment intensity, and the non-linear stimulus-response curve all contribute to the complexity of the overall picture in this study, and help to explain the counter-intuitive results seen here, in which the sham group’s responses approached those of the treatment group.
References:
[1]https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6125221/
[2]https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2022.867091/full
[3]https://jneuroengrehab.biomedcentral.com/articles/10.1186/1743-0003-11-40
*[3] is included for illustration purposes only, as it clearly shows the kinds of non-linear stimulus-response curves and plateau effects I am referring to above.
Note:
*While the P-field strength is measured as 7.72% of the A-field’s maximum output, this percentage does not directly correspond to the Motor Threshold (MT) percentages used to determine individual treatment intensity. The 7.72% refers to the coil’s maximum capacity rather than a personalised stimulation level based on MT.
Report comment