Wednesday, May 15, 2019

Human Magnetoreception Research Resumes After 30 Years of Neglect

Robin Baker’s human magnetoreception experiments in the late 1970’s and 80’s were pioneering efforts that led to—30 years of basically doing nothing. Failure to replicate his results was the reason for this, but failure to replicate is common in neuroscience. The brain is complicated, and we don’t have a good understanding of it yet. Why failure to replicate led to scientists abandoning human magnetoreception, but not abandoning other fields like brain imaging, search for chemical imbalances, etc. probably requires a psychological explanation along the lines of Thomas Kuhn’s classic The Structure of Scientific Revolutions.

While scientists abandoned the study of human magnetoreception, a small number of scientists continued studying nonhuman animal magnetoreception, making important contributions in experimental technique and understanding of biophysical mechanisms. These contributions, which were yet to be discovered at the time of Robin Baker’s pioneering experiments, along with general technological improvements, have led to much better experimental control than was possible during Robin Baker’s time. I’m pleased to be able to report two human magnetoreception experiments that make use of these scientific and technological advances, and have been published in science journals this year.

The first experiment, Chae et al., was done by a research group in South Korea. They did a modification of Robin Baker’s spinning chair experiment. In Baker’s original experiment, blindfolded subjects were spun around in random directions in a chair, and asked to say the compass direction they were facing. In this new experiment, the researchers created an artificial magnetic field using Helmholtz coils, and shielded subjects from external EMF’s using a Faraday cage. 41 subjects with no physical or mental disorders, ages 19-33 years, approximately evenly divided between men and women, were studied. Subjects were able to rotate their chairs instead of having the chairs rotated for them. They were asked, with eyes closed, to rotate their chair to face magnetic north or east. These magnetic directions were modified by the Helmholtz coils. Adding an operant conditioning component, some subjects were starved, and “rewarded” with candy if they faced the correct direction. The ambient light was experimentally controlled. Some subjects were blindfolded. The study made use of scientific advances in understanding of factors that can affect animal magnetoreception, including that some RF frequency EMF’s can affect magnetoreception, and that some types of magnetoreception requires low wavelength monochromatic visible light (i.e. blue or green, but not red or yellow). The authors concluded that starved men (but not women) significantly oriented toward magnetic north or east. This orientation was maintained under blue light, but not under long wavelength (> 500nm) light.

I have several issues with this experiment. One is that, like with many of the Robin Baker experiments, and unlike with most nonhuman animal experiments, there’s a lot of spread in the data. The highest r value (a measure of the variability, with a higher r meaning less variability) is 0.51, which isn’t very good. To get a visual idea, here’s a reproduction of the main results figure:



The blue dots represent data points, i.e. direction estimates. See how the blue dots in the supposedly significant D and H are spread around in a circle. If the subjects really were able to ascertain their direction, the dots would be congregated either at magnetic north or magnetic east. Results like these are invitations for failure to replicate, as happened with Robin Baker.

If you’re going to test light-dependent magnetoreception, why ask subjects to close their eyes? It drastically reduces the light hitting the retina, and may disable the light-dependent magnetoreceptor. There was no reason in this experiment for subjects to close their eyes, as the magnetic field was modified by the Helmholtz coils, and was invisible to the subjects.

It’s not a good idea to test human magnetoreception in only healthy subjects. Most people who have contacted me regarding purported magnetic sensitivity have some sort of psychological and/or physical problems. Just think about it—if you’re sensitive to magnetic fields, you’re not going to do well in modern society, with all the artificial magnetic fields that we’re forced to live with. You may have done well a long time ago, when we were hunter-gatherers, but not today.

I don’t see how starving humans and conditioning them using food adds anything of value to this experiment. If humans were able to ascertain compass direction, then it shouldn’t matter if they were starving or not.

Did any individuals exhibit special powers of ascertaining compass direction? The results were grouped, so I can’t tell. Anyone studying magnetoreception should look for individual differences and focus on those who are sensitive. This leads to my discussion of the second experiment, which did just that.

This Wang et al. experiment, done by a research group at Caltech headed by longtime magnetoreception researcher Joseph Kirschvink (joined with others at Princeton and the University of Tokyo), has been extensively reported in the science press (see this, and this for good nontechnical summaries). Instead of measuring navigational abilities like Robin Baker or the Chae et al. study mentioned above, this experiment looked at a drop in amplitude of the alpha EEG brainwave (known as alpha-event-related desynchronization, or alpha-ERD) in response to changes to external magnetic fields. It studied 24 adult males and 12 adult females, ages 18-68, “recruited from the Caltech population.” Subjects included people of European, Asian, African, and North American descent. Like with the Chae et al. experiment, a Faraday Cage was used to shield against external EMF’s. This experiment used a nested set of orthogonal, squared Merritt coils to modify the magnetic field surrounding the subject. An EEG was used to measure brainwaves, and current EEG analytical techniques were used to identify patterns. A battery-powered digital conversion unit relayed data over an optical fiber cable to a remote-control room. This room, ~ 20 meters away from the subject, had all power supplies, computers, and monitoring equipment. The paper goes into great detail on the experimental setup, to aid in future replication efforts. Participants sat with eyes closed, in total darkness during the experiments. Tests were run that varied inclination with declination constant, or varied declination with inclination constant. Each run was ~ 7 minutes long, with 8 runs in a ~ 1 hour session. There were sham runs (no changes) interspersed with real runs, and the experiment was conducted double blind. Here's a picture of the experimental setup:



The study reported alpha-ERD in 4 out of the original 36 participants (11%) that remained stable over time. The alpha-ERD occurred when inclination changed (i.e. from upward to downward, or vice versa). The alpha-ERD also occurred when declination changed counterclockwise, but only when the magnetic vector was pointed downward, as it does naturally in the Northern Hemisphere. There was no alpha-ERD when declination changed but the magnetic vector pointed upward, as it does naturally in the Southern Hemisphere. This asymmetry, along with other analysis, was used to rule out potential biophysical mechanisms such as the quantum compass and induction. The authors suggest magnetite as a likely biophysical magnetoreceptive mechanism, although this experiment wasn’t designed to prove this. None of the participants in the study could consciously distinguish between different magnetic field conditions.

This study needed a lot more information about the participants, especially the ones who had the strong responses. We are given no information about how they were selected. They were from the Caltech population. What does that mean? Were they students, professors, employees, or a combination of the three? The fact that someone is at Caltech at the time of the experiment tells me nothing of their background. They could have lived all their life in Sydney, and started at Caltech a month or two before the study. In that case, I’d expect their magnetoreceptor to be tuned to the Southern Hemisphere. Future human magnetoreception studies should have a detailed history of where and when a person lived. Many people move around a lot in their childhood, including from Northern to Southern Hemisphere, and vice versa. We also need to know where they spent their adult life. If someone who spent their childhood in Cape Town but has lived in LA for 6 years, and has alpha-ERD responses at Caltech with downward inclination like in the Northern Hemisphere, then that would indicate that this response is capable of adaptation after childhood.

There should also be personality and clinical testing of participants to determine if the responders had any noticeable differences from the non-responders. That would aid in identifying other responders.

One flaw in this study is that it groups statistics between responders and non-responders. They did find significant ANOVA results, but what if there were only one or two responders instead of four? That’s a potential problem for replication. If they need more subjects for statistical power, they should find more responders, and group the responders together.

While the authors’ exclusion of the quantum compass (i.e. radical pair) biophysical process makes sense based on their data, it must be remembered that the experiment was done with eyes closed, in total darkness. The quantum compass is a light-dependent magnetoreceptive process. It’s possible that humans have this quantum compass, which was turned off under these experimental conditions.
The alpha rhythm is an awake, resting rhythm. The conditions of eyes closed and total darkness is somewhere in between normal waking behavior and sleep. Studies of normal alert waking behavior should involve eyes open and lights on. This would activate the light-dependent quantum compass magnetoreceptor, assuming humans have it.

This study is very important in creating a magnetically-controlled condition that utilizes EEG, as this can be applied to sleep research. Ten years ago, I argued in my research paper that any studies on my magnetoreceptive abilities would require experimental control of my sleeping behavior. I couldn’t imagine at the time how this could be accomplished in a magnetically controlled way, especially EEG, which is critical in sleep research. Thanks to the Wang et al. experiment, it does appear now within reach. The kind of experimental controls used in this study can also be used in a sleep experiment. Some of the independent variables in sleep magnetoreception research include inclination and intensity, bed angle, bed time, and ferromagnetic materials near one’s head when sleeping. Bed angle (i.e. angle of the long axis of the bed relative to magnetic north), inclination and intensity can be manipulated by the Merritt coils. With EEG, bed time (i.e. the time when you initially go to sleep) can be precisely determined, along with the progression of sleep stages. Ferromagnetic materials can be introduced in the experimental chamber in a double blind manner to determine the effect on sleep. Dependent variables include subjective sleep quality, psychological state upon awakening (which is heavily dependent on sleep quality), and EEG.

I think from my own experience that human conscious perception of magnetic field changes requires prior sleep under similar conditions to the awake testing ones. That means that the subject will have to spend at least several nights sleeping in the experimental sleep apparatus prior to awake testing. It also means that magnetic changes need to be physically realistic. In the Wang et al. experiment, the inclination sweep was not realistic. It would require a Star Trek-type transporter to move physically that amount of inclination in that short a time. Realistic inclination changes that would maintain conscious magnetoreceptive responses are similar to those occurring while running, biking, or driving (not flying). Declination sweeps, however, are more realistic, as it only requires body rotation. In North America, both inclination and intensity are highly correlated, so physically-realistic changes would require changes in both at the same time, and in the same direction. Lastly, the subjective importance of bed time in my research seems to imply perception of geomagnetic diurnal variation. It’s possible that conscious perception of the artificially-generated magnetic field may require a time-dependent component similar to the natural field.

In summary, human magnetoreception research is back in business. While these two experiments are a start, and need to survive replication, they point the way to further experiments of sleeping and waking behavior. Under the right conditions, these future experiments can verify my hypothesis that some people are sleep sensitive to magnetic fields, that some people can consciously perceive magnetic fields, and this sleep sensitivity and conscious perception may be connected to symptoms of psychiatric disorders.