Low Level Radiation on Human Health

Lita Yang
April 20, 2016

Submitted as coursework for PH241, Stanford University, Winter 2016

Background on Low Level Radiation Studies

Fig. 1: Chart illustrating the numerous possible sources of ionizing radiation and the difficulty in accurately quantifying health damage from radiation threat. The numbers listed in this chart are estimates by the author, Randall Munroe, using known physics princples and established medical damage rules. The sources cited in the figure are subject to volatility, demonstrating the challenge in finding reliable measurements). (Source: Wikimeda Commons)

General public perception is that any form of radiation exposure is toxic to the human body, but studies have shown that this is not always the case. It is widely known that nuclear radiation can cause cancer, as Francis outlines in. [1] Research in low level radiation, however, demonstrates that low doses of nuclear radiation can actually have beneficial impacts to the human body. Studies in this area of research is often referred to as radiation hormesis or radiation homeostasis. Results from these studies indicate that certain low levels of radiation can stimulate the activation of biological repair mechanisms and have the potential to repair damage done to cells by ionizing radiation. [2,3] Low level radiation research has been active since even the late 1990s and as of January 2015, there has been renewed congressional interest in funding this type of research program in the United States. [4]

Despite growing interest in low level radiation therapies, one of the major issues with low level radiation health studies is that they have been highly controversial and at times, unreliable and inconclusive. As Baumer explains, the uncertainty in measurements of absorbed doses of radiation below 100 mSv can be very high and contributions from causes other than radiation exposure become more increasingly more significant. [2] This complicates pinpointing the root cause of health benefits seen in experiments done in this field, as Fig. 1 illustrates. Suggestions that "beneficial" hormetic responses should be incorporated into changing current nuclear radiation risk assessment models also disregards years of well-established principles that follow conservative public health protection approaches to regulating radiation exposure. [3]

While there is no debate that, under some circumstances, low doses of radiation can be beneficial while they toxic at higher doses, there are still many challenges associated with making this concept universally adaptive or widespread. In the following sections, this report will briefly cover several major issues with both public and scientific acceptance of radiation hormesis studies.

Dose-Response Relationship Models

Over the years, several models for the relationship between dose of radiation exposure and human health response (such as cancer risk) have been proposed. The problem with using these models to generalize the dose-response relationship, however, is that most of the data comes from atomic bomb survivors which were close to the hypocenter at the time of detonation and thus, received very high dose levels of radiation. Studies estimating received radiation dose based on a person's physical distance from the hypocenter, such as Douple et al., and public acceptance of the conservative ALARA protocol led to the formation of the so- called linear-no-threshold (LNT) model. [2,5] The LNT model shows a linear relationship between received dose and cancer risk for doses above 100 mSv. [5] Unfortunately, while the LNT model is widely used and has successfully limited the amount of radiation exposure to the public, it is overly conservative and inaccurately models cancer risk at low doses. The usage of the LNT model hinders proponents of low level radiation therapies from making strides in research, let alone implementation in practical settings, given the scope of the risks at stake in involving human lives. While other models for quantifying the dose-response relationships have also been proposed, such as the sublinear model or thresholded model, the LNT model is still the most reasonable choice as a conservative safety measure until further research demonstrates the true risk of low dose radiation on human health.

Why Low Level Radiation Experiments are Inconclusive

Continued acceptance of the conservative, yet inaccurate, LNT dose-response relationship model is due to the fact that current research in low level radiation studies is still inconclusive. What prevents acceptance of these studies are the numerous confounding variables, such as environmental or genetic factors, hindering prediction of the effects of radiation on human health. The challenge in low level radiation studies is separating out the "noise" from the actual causative agent. Several variables or "stressors" such as human adaptive responses, bystander effects, low-dose hypersensitivity, and induced genetic instability are amongst the phenomena not currently well understood. [6] Examples of these types of "stressors" include smoking, drinking, age, gender, or concurrent past or future exposures to the same or a different agent. [7]

To illustrate concretely how difficult it is to rigorously prove and define a model for the dose-response relationship, we consider how one would go about choosing subjects for these types of experiments. First, because the human body has its own repair mechanisms, this means separating out and quantifying these effects is even more difficult as this will also vary from person to person (or subject to subject). Second, another prominent reason for this difficulty is that the base irradiation rate in the environment is a finite value that will vary depending on location, and not only from the air, but also from rocks. For instance, the contribution of rock irradiation increases significantly for people who live in places such as mining countries, where radon accumulates over winters. Exposure to sky irradiation also increases for those who fly often. From these examples, it is clear there are several variables that can complicate these kinds of experiments and thus, it is difficult to find any study in which people are exposed to known amounts of radiation in controlled quantities while accounting for genomic instabilities.

Confounding Variables/Stressors to Low Level Radiation Studies

The complex matter of identifying problems associated with low level radiation studies is addressed by Mothersill and Seymour, who attempt to highlight the major uncertainties with this line of work. [7] They challenge currently accepted approaches to studying carcinogenesis and propose a balanced, system-like approach to examine all effects at both the cellular and organismal levels. [6] A quick summary of the "stressors" is outlined below:

Confusion in the low dose exposure research field comes from the lack of consideration that biological organisms are organized hierarchically. Most of the studies regarding whether radiation has a net positive or negative effect is due to the lack of consideration of the level at which the effects occur, and because many of the arguments have only used data from cancer incidents or deaths. [7] From an evolutionary perspective, Mothersill and Seymour argue that death of radiosensitive individuals in non-human populations which cannot adapt to the radioactively polluted environment can be seen as "good" in the context of hierarchical organizations. While this argument sounds "bad" from an individual perspective (especially in human populations), one can consider that as the plants and fruits humans consume become evolutionarily stronger from radiation exposure, humans also similarly adapt to these environmental effects over time.

The second major contributor to complications with low level radiation studies include concepts related to the age of the organism (i.e. state of maturity and complexity of the system) at the time of irradiation and the deposition pattern of the ionizing energy. Examples of what Mothersill and Seymour call "concepts related to time and space" include the considerations of age or maturity of the individual entity or system, the density of the radiation, and the lifetime of the entity or system. [7] Because most measurements are made at the individual level, models must extrapolate data from a variety of subjects to extend to larger populations. This is especially difficult since younger organisms are typically less stable and more vulnerable than older organisms due to higher metabolic rate and higher rate of growth and cell division. At the same time, the younger individuals have more capacity to adapt to change and have better reproductive rates. The last bystander effect is that radiation sometimes has a delayed effect on humans, further complicating measurements of its effect in time or space.

The last major "stressor" is consideration of multiple exposures to pollutants in the environment. Chemical, as well as radioactive aspects, need to be included in environmental radioactive contributors since these pollutants rarely occur in isolation. The recognition that there are complex interactions in the environment contributing to radiation exposure lead to a multitude of questions regarding regulation of proper dose unit per exposure unit, including "How to deal with multiple stressors especially if the interactions are not known? How to deal with mixed chronic and acute exposures? How to factor in possible adaptive, hermetic or antagonistic effects?" [7]

Concluding Remarks

In summary, the inaccuracy of models currently being used to define the relationship between radiation dose and the resulting human health response, and the challenges with accurately predicting the effects of radiation on the human body due to bystander hindrances and uncertainties in experimental procedures are preventing widespread acceptance of the benefits of low level radiation on the human body. We are still in the early stages of even accepting that radiation effects at low doses are non-linear, that multiple stressors impacts these types of studies, and defining what exactly is "good" or "bad" radiation effects. Clearly, there are still many difficult and unanswered questions before we can change the public perception of radiation exposures with respect to human and environmental health. Despite this, it is still imperative that the community receives a more accurate depiction of the risk of radiation on human health. There is still a large gap in understanding between nuclear radiation experts and the general public, as is evident from a study done from interviews with more than 200 employees of a nuclear installation compared to opinions of more than 100 members of the general public. [6] Once general public perception is improved, there may be more hope for further progress in pushing research in low level radiation forward.

© Lita Yang. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.


[1] J. Francis, "Inducing Cancer From Nuclear Decay Products," Physics 241, Stanford University, Winter 2014.

[2] M. Baumer, "ALARA: The History and Science of Radiation Safety," Physics 241, Stanford University, Winter 2015.

[3] D. Axelrod et al., "'Hormesis' - An Inappropriate Extrapolation From the Specific to the Universal," Int. J. Occup. Environ. Health 10, 335 (2004).

[4] "Low-Dose Radiation Research Act of 2014," Congressional Record 160, No. 140, H8011, 17 Nov 14.

[5] E. B. Douple et al., "Long-Term Radiation-Related Health Effects in a Unique Human Population: Lessons Learned from the Atomic Bomb Survivors of Hiroshima and Nagasaki," Disaster Med. Public Health Prep. 5, Suppl. S1, S122 (2011).

[6] A. N. Jha, W. H. Blake, adn G. E. Millward, "Preface: Environmental Radioactivity: Implications For Human and Environmental Health," J. Environ. Radioact. 133, 1 (2014).

[7] C. Mothersill and C. Seymour, "Implications For Human and Environmental Health of Low Doses of Ionizing Radiation," J. Environ. Radioact. 133, 5 (2014).