Can CT scans performed on a child's head lower his or her intelligence?

Question

The Daily Mail reports that X-rays and CT scans performed on a child's head could lower their IQ:

The study is the first to suggest that medical X-rays to the head could harm the development of the brain in later life. It found that adults whose brains were exposed to ionising radiation during infancy grew up to be less intelligent with fewer educational achievements than those who had not. The study also has implications for children today, who are increasingly given CT scans to check for injury after a minor head trauma.

I was also told this by a nurse.

Is this correct?

Answer 1

The British Medical Journal published Professor Per Hall's research paper BMJ 2004;328:19 in January 2004.

It claims

Low doses of ionising radiation to the brain in infancy influence cognitive abilities in adulthood.

The NHS advice (last reviewed January 2012) says that a typical CT-scan of a head uses and x-Ray dose of 1.4 millisieverts and that the diagnostics benefits outweigh the risks.

The average annual dose that a person in the UK receives from natural sources is 2.2 millisieverts

However, if you do not have any symptoms, the benefits of having a CT scan may not outweigh the risks, particularly if it leads to further unnecessary testing and added anxiety.

So it seems that other medical experts do not share Professor Per Hall's alarm.

The BMJ published several responses to this paper. For example, one containing

The relationship between head CT and adulthood intellect as established by the authors was based on the fact that some sources reported head CT doses in infants as high as 100 mGy (1, 2). Indeed, in their study, adverse effects on high school attendance was evident for groups at radiation doses higher than 100mGy. No such effects was evidenced in groups with lower doses (1-20 mGy). However, the doses discussed in the paper of Brenner et al. (2) resulted from a national trial in Britain, that taken in account adult scanning protocols. Current, paediatric head CT effective doses as recently measured in a Belgian multicentre study ranged from 0.4 to 2.3 mSv*(3). These values are dramatically lower than those cited by Hall et al (1). Moreover, the authors did not discuss important points that could alter the credibility of their caculated doses. The bias related to retrospective dosimetry, which I suppose (considering both the administration method and the ancillary equipment used) was high, was not taken in account. Also, the differences between administered doses and absorbed doses were not discussed.

Answer 2

UPDATE: As noted by @Oddthinking, I missed the report referenced by @redgrittybrick.

I could not find a link between CT scans and intellectual development in children. I searched PubMed using MeSH terms such as "tomography, x-ray computed" (technical term for "CT"), risk, intelligence and various terms for youth (child, infant or fetus).

Of course, not finding something doesn't disprove it but it does suggest that it isn't well-studied or generally accepted in the medical community ... if you accept that I made an adequate search of the scientific literature.

In contrast, there's plenty of information on CT scans and an increased risk of cancer for infants/children (eg, recent article in Lancet). An important note here is that the risks need to be balanced against the dangers of not running CT scans (see Catch study and a recent review in J Pediatrics). In other words, NOT running a CT scan (for high-risk cases) is more likely to lead to lowered "IQ", due to improperly diagnosed/treated head trauma.

Answer 3

This is a (too long) comment about the Swedish study, in addition to particularly @igelkott's link about balancing the risks of death by head trauma and death by cancer due to the unnecessary CT.

---

I'm wondering how much can be skeptically concluded from the The Swedish study. They list as participants

Participants 3094 men who had received radiation for cutaneous haemangioma before age 18 months during 1930-59.

In other words, the studies population are boys who got radiotherapy because of tumours on the skin of the head or somewhere else. I assume (though I did not find that clearly stated in the text!) that the controls are boys that got radiation therapy for this type of tumour somewhere else than the head and neck region.

IMHO the more interesting control group would have been boys who had the hemangiomas on head and neck but did not get radiation therapy (stratified for severity of the tumour / the radiation dose they "should" have gotten). Or at least, one would need to establish that high school attendance of people with untreated infantile hemangiomas was the same compared to boys without hemangiomas (probably not, if "only" because of the psychosocial difficulties in school).

Now according to this webpage these hemangioma are benign tumours, 60% of them appear on head and neck. But a particular form of these hemangioma appears within the PHACE syndrome, which also includes brain malformations and other problems. I have no information about the frequency of this, but it would have been good if the Swedish study had explained if any/how many such patients were in their patient population.

The other point that needs to be kept in mind is that according to Benner 2001 the cancer risk decreases with the age of the child when the CT takes place. The Swedish boys were less than 1 1/2 years old when they got radiotherapy. Now cancer isn't the same as not attending high school, but this tells something about how sensitive the brain reacts to radiation.