The causes of childhood leukaemia

This website is devoted to the issue of whether magnetic fields (or the power lines that are one source of the fields) cause harm to health and in particular whether they cause childhood leukaemia.

Whether or not magnetic fields are a cause of childhood leukaemia, they would not explain more than a small fraction of cases.  So that poses the question, what causes all the other cases? (see more on the statistics)

One known cause (for childhood leukaemia like many other cancers) is ionising radiation - x-rays, gamma, radon etc.  On present understanding that maybe accounts for 25% of cases in the UK.  And there are a couple of rare genetic conditions that are known risk factors too.  But that leaves the majority of cases with no absolutely clear cause.

However, there are two related hypotheses that are becoming more accepted - indeed, they probably don't deserve to be called "hypotheses" any more, they should probably be seen as established mechanisms.  They are usually called after the scientists most responsible for promulgating them: the "Kinlen hypothesis", named after Leo Kinlen, and the "Greaves hypothesis", named after Mel Greaves.

The Kinlen hypothesis applies where you mix two populations with different immunities to infections (it is sometimes called the "population mixing" hypothesis).  They become exposed to new infections that they have not developed an immunity to, and this causes childhood leukaemia rates to rise.  The classic version of how this can happen is when you have a hitherto relatively isolated rural population, and then a new industrial enterprise or some other change leads to a big influx of people into that area from urban areas. Kinlen has demonstrated this many times in a series of elegant epidemiological studies.  The best-known example is probably the way leukaemia rates increased dramatically in the rural Cumbrian village of Seascale with the influx of professionals working at next-door Sellafield.  But Kinlen has also shown the same effect with new towns, wartime evacuations, the oil industry in Scotland, and others.  And, although Kinlen himself limits his hypothesis to major influxes of whole populations, others have shown that leukaemia rates are related to residential mobility in general - how often individual people in an area move in and out and how far.

The Greaves hypothesis says that our immune systems need stimulation in early childhood to develop properly.  If they don't get that stimulation , later exposure to common infections can result in leukaemia.  This is a variant of the "infection hypothesis" which broadly says that early exposure to infections can often be good for us in the long run.  So childhood leukaemia rates should be linked (inversely) to mixing with other children - having older siblings, early attendance at childcare facilities etc, and this too has been demonstrated in epidemiological studies.  This could also explain the observation that breastfeeding is mildly protective against childhood leukaemia.  (Another part of the Greaves hypothesis is about how leukaemia is a "two-hit" process - you need a first hit in utero to cause a genetic abnormality, which lies dormant unless and until it receives a second hit later in childhood.)

Kinlen and Greaves are two different mechanisms. But they are linked in that they both say that leukaemia risk depends on social factors - moving house, mixing with other people - and not just those factors as they apply to yourself, but how they apply to other people in the area.  We often refer to them as socioeconomic or demographic.

Are these theories or fact?

Scientists often have different views as to how strong the evidence is for these hypotheses.  (And sometimes supporters of one hypothesis may not feel as positively about the other...)

In both cases, we don't yet know enough to know exactly how they operate, exactly what the risk factors are, how to apply it to an individual etc.  But in both cases, most people would say the evidence is strong enough that we can regard the basic operation of each mechanism or something very much like it as established fact.

What we don't yet know is whether, if we understood exactly how these mechanisms operate, we'd find that they explain all the cases of childhood leukaemia that don't have an obvious explanation like ionising radiation - or whether there would still be some where we simply have to say that we don't know what causes them.

Could these mechanisms apply to power lines?

Both mechanisms depend on the socioeconomic or demographic makeup of an area - how often people move, what sort of childcare facilities they use, etc.  It seems perfectly plausible that this could be linked to the presence of a power line.  Either the power line could be built close to areas with certain characteristics to start with, or the presence of the power line could subsequently affect the area.

Of course, merely being plausible doesn't mean it is established fact.  We do in fact know a fair bit about socioeconomic factors near power lines, and we have started looking at residential mobility in both the UK and USA.  But we have not yet made very clear connections between all these factors that could lead to an explanation of raised childhood leukaemia rates near power lines.  But nor have we ruled it out.

If these mechanisms do operate with power lines, however, it would fit with the broad scientific evidence as it stands at the moment.  The scientific evidence suggests that leukaemia rates are elevated with either high magnetic fields, proximity to power lines, or both.  But the laboratory evidence, and our theoretical understanding of mechanisms, says that magnetic fields are not the explanation.  Nor have we identified other direct physical mechanisms (corona ions and contact currents are two candidates we've looked at and pretty well ruled out).  So an indirect mechanism, something socioeconomic, that doesn't require any direct physical effect of the power line, would be a way of making sense of all the evidence.

Note: if power lines have an effect on leukaemia rates through one of these indirect mechanisms, that would just apply to some of the leukaemias near power lines, and they are only a small fraction of the total.  In most places, the Kinlen or Greaves hypotheses or some variant of them would apply as normal with no role for power lines at all.

The particular relevance to the CCRG study

The "CCRG study" - an epidemiological study of childhood leukaemia rates and power lines in the UK - found several unexpected features.  The elevated leukaemia rates extended too far from the power lines to be a direct effect of magnetic fields.  The elevated rates seemed to decline from the 1960s to the present day.  And most recently, a new paper showed that the rates didn't actually fall smoothly with distance from the power line, but were highest about 100-200 m away and only started falling after that.

None of these factors is really compatible with it being magnetic fields that caused the increases.  But they all could be compatible with some sort of socioeconomic explanation - a variant of the Kinlen or Greaves hypotheses.  For example, whilst it is not very plausible that the direct biological effect of magnetic fields on cells should have changed over just a few years, it is more plausible that the socioeconomic characteristics of a neighbourhood could change on this timescale.

This is speculative - but some sort of indirect consequence of the presence of a power line is probably the best candidate explanation we have for the CCRG results so far.


Recent publicity in the UK

In 2018 one of the two theories we talk about on this page, the Greaves hypothesis, received some attention in UK media.