For the curious, here is my expanded first draft; I should say emphatically that my coauthors on that paper are not responsible for what’s said here, this is all BbtP baby. As you can see, a lot was cut out to fit in the space provided, and to not come off as some guy ranting on a blog. Fortunately, this is the perfect venue for an article that comes off as a guy ranting on a blog! For those who have no idea what this is a first draft to, don’t worry about it, this stands on its own.
In Science (vol 327, p931, 19 Feb 2010) a little article came out describing the difficulties researchers in China were having with their studies of functional magnetic resonance imaging in children specifically to get healthy children to volunteer for a scan to act as controls for their patient group.
Intending to enroll children in the study, university students last December handed out fliers at a primary school. But they came away empty-handed: Parents were worried that MRI scans might harm their children. […] parents are reluctant to expose children to strong magnetic fields.
After 3 decades in the clinic, MRI is considered safer than x-ray scans and proton emission tomography, says physicist Yihong Yang, chief of the MRI physics section at the National Institute on Drug Abuse in Baltimore, Maryland. The main danger is for people with a pacemaker or other metal in their bodies. “Millions of people have been examined with MRI so far; thus it seems now very unlikely that there would be a side effect,” says Arno Villringer, director of cognitive neurology at the Max Planck Institute for Human Cognitive and Brain Sciences in Leipzig, Germany. […] Recruitment goes more smoothly in the United States, where “many parents will allow their children to take the test…”
There is some balance though:
…that reassurance cuts little ice with many parents—and some scientists. “I would not dare to allow my children to be tested by MRI,” says radiologist Han Hongbin of Peking University Third Hospital. “Nobody can ensure that there is no potential danger,” such as during nonroutine MRI scans that use extremely powerful magnetic fields, he says.
This is a very tricky ethical dilemma, for a number of reasons.
Stating that “millions of people have been examined with MRI so far; thus it seems now very unlikely that there would be a side effect,” is an irresponsible statement to make in the face of empirical evidence to the contrary. We know that there are deterministic effects of the fields used in MRI: strong, rapidly changing gradient fields can stimulate nerves to fire; radiofrequencies can cause heating; strong static fields can create projectiles out of metal objects. There are safety limits and procedures in place to attempt to control these effects in MRI scans, but there may still be stochastic effects present. Indeed, it is well known that at high fields some people may experience nausea/vertigo, or a metallic taste. Several studies have examined the effects of the time-changing fields. These effects are not necessarily negative: Rohan et al.  found that a particular MRI scan could improve mood in bipolar patients, and the undesireable effects are subtle and fairly easily managed. The effects are also not necessarily chronic, but the possibility that more exist that we have not yet discovered must be accounted for, particularly when imaging subjects that will have no medical benefit from the scan.
This becomes a particularly touchy issue when children become involved. There are many processes that affect developing organisms more severely than fully matured ones. Some early research into the effects of electromagnetic fields suggests that this may be the case for something like an MRI as well. The fact that millions of adults have been scanned over the past ~30 years may be cold comfort to a parent volunteering their child.
For some biological effects there may be a long latency time as well, before effects can be observed. Solid tumor formation, for example, may take over a decade to manifest after ionizing radiation exposure. Good, long-term epidemiological studies will be needed in the future, alongside basic experimental studies on animals and cell cultures, to arrive at a good answer to the question of whether there are biological or behavioural effects of MRIs, and what they might be. Indeed, it is well known that there is a slightly increased risk of cancer formation from the ionizing radiation present in an x-ray/CT scan. However, if we were at the same point in ionizing radiation-based imaging as we are with MRI — millions of patients scanned, but little long-term follow-up looking specifically for effects — we might not be able to find that risk! Absence of evidence is not proof of the absence of risk, and it is widely accepted that there are small, but non-zero risks associated with CT. Accordingly, it has been appropriate to adopt the precautionary approach, and apply the dictum of ALARA – as low as reasonably achievable – to ionizing radiation dose.
I’m not in any way trying to be fear-mongering: fuller understanding of the processes is needed to mitigate these potential risks. Gadolinium-DTPA was for a long time used as a contrast agent and was understood to be safe. Recently, it has been determined that it can have severe side-effects in some people. These risks can be managed, for example through kidney function tests to determine who may be at risk. Any behavioural or biological effects of MRIs may likewise be managed in the future once they are better understood.
Hoping that there are no effects, without rigorously pursuing the matter in a scientific manner, will not be beneficial in the long run. More research on the potential biological effects of MRI, especially in children, is needed.
We can’t say for sure yet, and will indeed never be able to say with full confidence that there are no biological effects of the fields used in MRI. At this point we can probably agree that any effects, in adults, are not so severe as to counterbalance the usefulness of MRI as a diagnostic and research tool. As more research is done, we will either uncover what the effects are exactly, or continue to narrow down the maximum effect that we could discover with the population tested. With children, however, the studies have not been done, and our confidence that MRIs have no effect is not particularly strong.
And note that I say this as someone who depends very much on having volunteers step up to get an MRI for research purposes, and as someone who’s had a dozen MRIs himself. For an adult, having an MRI is safer than going for a short cruise in your car, a risk that people take for trivial benefits on a regular basis, so I completely understand when someone steps up to help science (and make a few bucks themselves). For children though, we just don’t know. Ultimately it’s the parents’ decision, but informed consent means that the researchers should try to educate them about both sides: the benefits and research aims as well as the potential, unknown risks.
A study is going to be better scientifically if it uses healthy controls to compare to a patient group (in this case, ADHD), rather than a different patient group (such as epilepsy) where the region of the brain being studied is likely not affected. However, that has to be balanced with the ethical realities of doing research.