When Does Radiation Become Bad For Your Health?

Radiation is everywhere. We catch it from the sun's rays in the sky, and from the rocks beneath our feet. It comes from television sets, radios and mobile phones. We absorb it from certain fruits, vegetables and nuts.

Radioactivity picture from Shutterstock

But not all radiation is equal. Electromagnetic radiation, including radiowaves, microwave, visible and infrared light is known as nonionising radiation, and is largely harmless. On the other hand ionising radiation, from wavelengths shorter than ultraviolet light through the electromagnetic spectrum to X-rays and gamma rays, can cause disease and death.

These effects are from its ability to ionise (that is, separate the positively and negatively charged ions) in bodily tissues. Broadly speaking the risk of damaging health effects is proportional, in a rather complex way, to the extent of the ionisation induced in the body. This is called the dose. How ionising radiation is measured and defined has changed over the decades as we learn more about this relatively young science.

Measuring radiation dose and risk

Dose was originally measured in air by the unit of Roentgens (R, named after the discoverer of X-rays, Wilhelm Roentgens). As ionisation cannot be measured in tissue it was necessary to convert air dose to absorbed tissue dose, originally measured in rads, where 1 R = ~0.8 rad. With the introduction of metric units the basic unit of absorbed dose became the Gray (Gy), which represents an absorbed dose of 1 Joule of energy per kilogramme.

Unfortunately absorbed dose is not very convenient for radiological protection purposes because 1Gy of the different radiations -- gamma and X-rays, beta particles, neutrons and alpha particles -- is not equally damaging to tissue. Consequently a "hybrid" unit, the Sievert (Sv) was introduced. Hybrid, because it is really not a unit of radiation dose but a unit of risk. Thus, we talk of the equivalent dose of 1Sv as carrying the same risk, for example, as 1Gy for X and gamma rays, or 0.05Gy for the more densely ionising, but less penetrating alpha particles.

But there is a further complication, as not all tissues in the body are equally sensitive. Bone marrow and a child's thyroid are much more sensitive than muscle tissue, for example. So the term effective dose which incorporates the correction for equivalent dose and is also measured in Sv, is used. This way, if only part of the body is irradiated the risk can be presented in terms of an effective risk to the person. This allows risks from different exposures to be added together. The unit Sv should not be used for large doses (greater than 1Sv) to the whole body.

Low doses are common

Typically, everybody is exposed to two milliseiverts (mSv) per year throughout our lives from natural background radiation. We might receive a dose of up 10-20mSv from diagnostic radiology -- say 10mSv for a CT chest scan. The firefighters and plant workers at the Chernobyl accident received doses of several Gy and these doses led to deaths from acute radiation sickness within about 60 days. Typically 4-5Gy received over a short period of hours will be lethal, but can be tolerated if delivered over a much longer period.

Recommendations from the International Committee on Radiological Protection limit radiation workers to 20mSv per year or in exceptional cases higher annual doses, limited by 100mSv over five years. Doses to members of the public from discharges from nuclear power plants and laboratories or leakage from, for example, medical radiation sources at hospitals should be limited to 1mSv per year.

Extreme radiation events

Clearly, in the case of accidents such as at Chernobyl and Fukushima the situation is much less well controlled. Doses of around 30mSv were received by the 115,000 people living in settlements close to Chernobyl before the 30km-radius exclusion zone was evacuated days later. In the case of Fukushima evacuation up to 20km from the power plants was much faster. Much higher doses (up to 250mSv) were received by some clean-up workers after Chernobyl, and little is known yet about doses to clean-up workers at Fukushima. If recent reports of doses up to 2.2 Sv/hour from leaking tanks on the site are true and if this dose is from gamma rays, then it may soon become too dangerous to work on the site.

To cause death within hours of exposure to radiation, the dose needs to be very high, 10Gy or higher, while 4-5Gy will kill within 60 days, and less than 1.5-2Gy will not be lethal in the short term. However all doses, no matter how small, carry a finite risk of cancer and other diseases.

A very approximate rule of thumb is that 1Sv carries a risk of a 10 per cent increase in lifetime risk of cancer. This cancer risk may persist for the remainder of life but is unlikely to appear before at least 10-20 years after exposure. So, exposure from accumulated natural background radiation up to the age of 50 years (=100mSv) increases the ~30 per cent lifetime risk of cancer to ~31 per cent and mortality from ~25 per cent to ~26 per cent. On this basis some 30,000 to 60,000 cancer deaths worldwide, but mainly in Europe, will have been caused by the Chernobyl accident, and many still remain to occur.

Much discussion is made of the so-called low dose problem. Effects from doses of less than 50mSv are difficult to assess directly due to the high background of spontaneous (naturally occurring) cancer, so it has been necessary to extrapolate down from measurements of effects at higher doses. The question is whether there is a dose threshold below which there is no effect. From what we know that threshold must be below 10mSv and by the age of ten everybody has received at least 10mSv natural background radiation from natural background sources, so there is no argument for a threshold -- all doses of radiation, no matter how small, entail a finite risk.

Keith Baverstock is Docent in the Department of Environmental Science at University of Eastern Finland. He does not work for, consult to, own shares in or receive funding from any company or organisation that would benefit from this article, and has no relevant affiliations.

The ConversationThis article was originally published at The Conversation. Read the original article.


    Wow. This article is so odd that it must have been mangled during editing.

    "Electromagnetic radiation, including radiowaves, microwave, visible and infrared light is known as nonionising radiation, and is largely harmless."

    The paragraph then goes on to say that electromagnetic radiation can cause harm. Confusing, to say the least.

    There is no clear distinction between electromagnetic and particulate radiation or the different effects (particulate is generally much nastier than electromagnetic in terms of absorbed does and biological effects). There is also no explanation that there is an ongoing debate about linear vs. threshold models for radiation exposure, let alone the controversial idea of radiation hormesis.

    Basically, this is too complex a subject to be covered in a few paragraphs, and just causes confusion for readers.

    "...all doses of radiation, no matter how small, entail a finite risk."
    Rubbish, there is plenty of evidence that low doses are good for you.
    Google "Radiation hormesis".

    This article could hardly be more harmful. I came to this article expecting an overly simplistic response. However, I was pleased to read a reasonably good explanation of effective dose. From the second last paragraph, though, it is clear that the author is either severely ill-informed or purposely trying to prolong and support the panic that is usually associated with radiation.

    Having read reports from WHO and IAEA (both UN organisations), I can tell you that it is absolutely invalid to multiply lifetime risks by the population exposed to get excess cancers. This calculation method is specifically advised against. I would be happy to find the references if it would lead to the article being appropriately edited.

    The reason that the calculation method is deemed invalid is due mainly to the extremely uncertain data we have concerning radiation harm at low doses. Some data points to a threshold dose below which radiation is not harmful, some data shows that too little radiation is actually bad for you, and some data shows a linear increase in harm as radiation dose is increased. This last dose response model is known as the linear no threshold model and is used in to form radiation protection guidelines mainly because it makes calculations very simple (but also because it's deemed to be very conservative).

    The main point is that the first section of the article made up of well known fact while the last part it is basically no more than the author's opinion. The deaths that can actually be associated with the Chernobyl disaster are all expected to fall among clean up workers. (There are actually expected to be a few dozen deaths due to thyroid cancer among nearby residents - about 2000 excess cases are expected but almost all thyroid cancer sufferers are cured). No deaths among the general public are expected to occur as a result of radiation exposure due to the Fukushima disaster but it's too early to tell.

    I am a medical physicist - we deal with the risks associated with medical radiation exposure.

    Last edited 10/09/13 1:26 pm

    Knackerbrot is entirely correct. The article is quite misleading, and should emphasize above all else that this is not "settled science". There is a very large uncertainty in risk estimates at low doses, and it is clearly unethical, indeed immoral to use these risk estimates to scare the populace into avoiding necessary, sometimes life saving radiation exposure such as CT scan, etc, and unfairly demonizing nuclear power.

    These low radiation dose ranges may indeed prove to be beneficial to human life, and not detrimental at all, or there may be very very small levels of risk that are genetically mediated, or there may be a simple threshold to zero. We just don't know.

    Intuitively, and given recent significant advances in radiobiology research, the LNT is the LEAST likely scenario. Much of the epidemiological study supporting the LNT assumption is flawed. This is mainly because such studies are designed from the outset to find detrimental effects, and must therefore ignore all evidence to the contrary.

    I personally receive about 3-4 mSv of gamma radiation dose per year occupationally, and have done for 35 years. I have no concern with this, especially since reading "Radiation Hormesis and the Linear-No-Threshold Assumption" by Charles Sanders, which exposes the bad science behind much of this research, and the consequent propagation of radiophobia.

    BTW dolphins are apparently not so smart as we thought, despite decades of research striving to prove just the opposite. Just another case of confirmation bias. Flipper just took a dipper.

    For the same reasons we should accept climate change research consensus very very cautiously. It is simply too new, not balanced, and too far from a mature science (relying as it does, inordinately, on selective mathematical modelling). It also lacks any real empirical basis, that is, we can't design useful experiments to test our climate hypotheses, but must rely on historical observations and intrinsically unreliable predictive models.

    ps I have a masters degree in Health Science with a major in radiation safety, and have studied the subject extensively.

    I also have received occasionally many more times the occupational dose limit for radiation during a 40 year career as a radiation research worker but do not have the slightest concern about this. As a radiation biologist, with emphasis on the biologist, I am well aware that all life has evolved and indeed may be largely due to constant background exposure , in the past at much higher environmental doses than now exist. As a consequence all cells have active cellular DNA and tissue repair mechanisms that in humans can cope with natural background radiation and indeed larger exposures over most of the life-time of the irradiated individual.

    As a cancer biologist I am also aware that the increased risk of getting a clinically detectable cancer is a multistage process requiring a number of "environmental" exposures not only to radiation but also other risk events e.g. infections with virus, exposure to chemicals , hormonal status, trauma, nutritional factors genetic inheritance etc., as well as other epi-genetic factors which as yet are still poorly understood.

    Because of its developmental history, radiation protection guidance and dose-limits have been set, largely by physicists using historical data, adopted by international bodies and followed nationally, mostly on a fail safe basis- and at vast financial cost . It is to be expected, if only due to the politics, that the "No Threshold Theory" is going to be given up to "Radiation Hormesis" that easily.

Join the discussion!

Trending Stories Right Now