Radiation Hormesis: Is It Good for You?

by Brian Dunning
October 4, 2016

Today we're going to gather our courage, set our jaws, and march into the Chernobyl sarcophagus in search of seemingly-improbable health benefits. For our subject is radiation hormesis, the idea that low doses of radiation are not merely harmless, but actually beneficial, in that they stimulate the body's repair mechanisms in response to the radiation and make us better suited to resist normal radiation levels in the future. Is it true? Some say it is, and claim all sorts of science supporting it; some say there's no truth to it and the body's repair mechanisms function just fine without any such kick-start; and some say that the idea is plausible but it's unproven and does not justify self-irradiating in the interest of finding out. Does any one group or another have the preponderance of scientific opinion on its side?

To begin, I'd like to start by identifying some red flags that pop up when you set out to research radiation hormesis. A red flag like these does not prove anything, and I'm absolutely not offering these red flags as evidence of anything. The value of a red flag is to warn that extra caution might be warranted. In this case, these red flags suggest that some of the support for radiation hormesis as a medical therapy might be ideological or commercial, and not necessarily scientific.

Red flag #1 is that surprisingly (or maybe not surprisingly, depending on your point of view), the division between those who embrace radiation hormesis and those who are cautious of it closely matches political party affiliation. Those trumpeting the benefits of radiation hormesis the loudest are often the same ones who deny anthropogenic global warming. This may be the result of people getting their information from political sources rather than from science sources.

Notably, political pundit Ann Coulter has long espoused that radiation — specifically that from events like the atomic bombing of Hiroshima and the explosion at Chernobyl — resulted in medical benefits, in the form of reduced cancer incidences, to all who weren't killed outright. And that's provably false.

Red flag #2 is that sources promoting radiation hormesis are often dedicated to promoting it. (Check websites like HiroshimaSyndrome.com, Radiation-Hormesis.com, and a raft of books on Amazon) They do so through cherrypicked and/or misrepresented science, combined with persuasive anecdotes and charges of conspiracy and suppression. Conversely, sources devoted to presenting valid science conclusions typically present research that isn't cherrypicked, and aren't known for making charges of conspiracy and suppression.

In fact, if we go back to Skeptoid #37, How to Spot Pseudoscience, we find that radiation hormesis ticks virtually every one of our 15 warning signs. So I'd say we have pretty good cause to dive into the research and see what the testing actually shows. Is a little bit of radiation actually good for us?

First, let's lay down some background information. Generally speaking, ionizing radiation is harmful to living creatures. This is in contrast to non-ionizing radiation which is harmless, low-energy things like visible light, radio broadcasts, and thermal radiation from your fireplace. Ionizing radiation is that which carries enough energy to strip electrons from atoms and molecules, or make other chemical changes. Ionizing radiation includes the types we normally think of as dangerous, such as radioactivity from nuclear waste or hydrogen bombs or cosmic rays: alpha particles, beta particles, neutron radiation, gamma rays, or X-rays. All generally bad stuff.

The more of this hits your body, the more of your tissue gets damaged. Governments around the world respond by establishing radiation safety limits. In order to do this, they have to use a dose-response curve. This is a model that establishes how much radiation over how much time results in how much increased cancer risk. (If you get a whole lot of radiation all at once, like a fireman at Chernobyl, a lot of your cells get killed and you get radiation sickness, possibly resulting in organ failure or death. But for long-term, gradual exposure to radiation, the most common health consequence is increased cancer risk from damaged DNA. Generally, most healthy people can withstand a certain amount of constant radiation. Your body can repair damage at a certain rate. But get too much at once, and your body gets overwhelmed.)

The model used by most governments for regulatory purposes is called LNT, or "linear no threshold". It's the simplest possible model. "Linear" means the risk increases equally with dose, starting at zero, and "no threshold" means no threshold dosage has been established below which exposure is considered to be without consequence. There are other possible models. It might use a curved relationship where radiation at a low dosage is more harmful per unit, or the opposite curve where radiation at a high dosage is more harmful per unit.

There is also a model that represents the hypothetical radiation hormesis effect. This curve is U-shaped. It starts at zero along with zero exposure, but then it dips down below the zero-risk line — indicating radiation at that dose actually reduces the risk of cancer, less risk than you'd have with no radiation at all — and then, as the dosage increases, the curve comes back into the increased risk zone, and continues curving upward as the dosage increases. This model does not merely presume that low doses of radiation are harmless, but that they are actually beneficial.

Proponents of radiation hormesis often compare it to vaccines. Challenge a defense system, and you will strengthen it. But although this sounds logical, this particular comparison is invalid. When the immune system responds to a vaccine, new killer T cells are created that populate the body and bear protein markers specifically designed for the targeted pathogen. The vaccine causes an actual physical change in your body. Conversely, when the body responds to radiation damage, the only tools available are the innate metabolic systems for replacing damaged tissue. Nothing new is created, nothing is multiplied, nothing is changed.

Homeopathy is a type of hormesis. An assumption made by homeopathy — now known to be pseudoscientific — is that some dose of a toxin that's so low it's actually represented only by a "spiritual essence", can confer protection against real doses of that toxin, or even reverse damage done by it. We know for a fact that this is false, but might hormesis actually be real when the amount of toxin is non-zero? Like, very very small, but non-zero?

It's really hard to say, and that's one reason there is controversy about radiation hormesis. We all live our lives bathed in low doses of harmful radiation, mainly from the sun. We get occasional X-rays at the doctor; we breathe radon gas when we're around natural bedrock; we go through airport security; we eat bananas, the potassium in each of which gives us a tiny shot of ionizing beta particles. This is all just background radiation. It's noise in the data. Patterns in the data don't become clear enough for us to draw reliable dose-response curves until we get to much higher dosages, the dosages at which people do get sick, allowing us to learn what the risks truly are. If any hormesis effect does exist, it's also lost in that noisiest of data at the very bottom of the dose scale.

Since we all live somewhere in this noise, it seems reasonable to infer that very low doses of even ionizing radiation are harmless, at least for all practical purposes. It might therefore seem like the LNT model is flawed in that it would be more appropriate to have a threshhold; to draw the curve in such a way as zero increased risk of cancer lies at some non-zero dosage. And, accordingly, there are some arguments that this is exactly what should be done. The basic argument is that the extreme conservatism prompted by the LNT model, and the resulting regulatory burden with its associated expenses and limitations, produces no discernible benefits for the overwhelming majority of people.

Let's have a look at some examples. The Health Physics Society is the professional organization of radiation health scientists; basically the world's best informed experts on this subject. Their 2016 position statement says:

...Because of statistical uncertainties in biological response at or near background levels, the LNT hypothesis cannot provide reliable projections of future cancer incidence from low-level radiation exposures... Substantial scientific data indicate that the LNT model of radiation effects oversimplifies the relationship between dose and response.

Yet regulators like the United States Nuclear Regulatory Commission still embrace it:

This dose-response model suggests that any increase in dose, no matter how small, results in an incremental increase in risk. The U.S. Nuclear Regulatory Commission (NRC) accepts the LNT hypothesis as a conservative model for estimating radiation risk.

And the International Commission on Radiological Protection says, in essence, "Stop! You're both right":

...While existence of a low-dose threshold does not seem to be unlikely for radiation-related cancers of certain tissues, the evidence does not favour the existence of a universal threshold. The LNT hypothesis... remains a prudent basis for radiation protection at low doses and low dose rates.

You can dig and dig through the literature, and this is the basic theme you'll find time and time again. LNT probably doesn't represent the real risk at low doses, but it's probably still the safest model to use for regulatory purposes.

But nowhere will you find recommendations that radiation hormesis be made part of the model — at least, nowhere outside of the fringe sources. A dedicated literature search will indeed find claims that hormesis is real, but these effects are always found within the noise; and when you find references to these studies in mainstream sources, they always point out the flaws. To sum up radiation hormesis in one sentence, we can say that it is one claim of a pattern that some say can be found in the data, but that most dismiss because the data is simply far too noisy at that low level to support the drawing of any conclusions at all.

Most radiation researchers will probably agree that radiation hormesis is potentially plausible, but unproven and probably unprovable. Even if it exists, the data that can be interpreted to support it would show that the hypothetical protective effect is also too small to be detectable. So don't go running to your local nuclear waste dump for a restorative bath in a barrel of anything glowing green.

By Brian Dunning
Follow @BrianDunning