The No Observed Adverse Effect Level
The NOAEL is a five-letter acronym, and every letter is important. In fact, it was originally conceived as a four-letter acronym, but that turned out to be one letter too few. The NOEL was introduced in 1954 by the FDA as part of a procedure that was being used to develop regulations for pesticides based on the results from animal studies (Lehman and Fitzhugh, 1954). It is probably best understood by starting with the last letter and working backwards:- Level. The level is a toxicological dose. Animal studies are typically conducted with three or more dose groups, where each group is exposed to a different amount of the chemical being testing.
- Effect. Toxicological studies are designed to look for one or more possible effects that may be caused by the chemical. These may include effects that are both obvious and subtle effects that can only be detected by careful inspection or highly technical measurements.
- Observed. All studies have limitations about what they can measure. They also all have a finite list of effects that are looked for. In addition, even in a controlled laboratory experiment, it is not always possible to be sure than a particular outcome is caused by the chemical, especially when the potential effect can occur without exposure. Therefore, statistical tests are typically used to exclude the possibility that an apparent effect occurs by random chance.
- No. This modifies Observed. If statistical significance is not attained, then the effect is considered to be unobserved. Because the burden of proof is show that there is an effect, rather than showing that there is not, the NOEL is not precautionary. In fact, it is quite the opposite.
Putting those all together, the NOEL is the highest dose in an animal experiment in which a statistical significant effect is obtained. Since all effects are not necessarily bad, “Adverse” was introduced later as a modifier for “effect”. For example, high doses of a food color can turn a white rat pink. Some people might consider that to be quite fashionable.
The Lowest Observed Adverse Effect Level (LOAEL) is a phrase that is used to describe the dose of the lowest dose-group that exhibits a statistically significant effect. If there is no dose group with no effect, a LOAEL may be used in place of a NOAEL, usually with an extra safety factor.
The ADI and the NOAEL are also used by the World Health Organization for the evaluation of food safety. Their current definition of the NOAEL is this (WHO/FAO, 2009):
Greatest concentration or amount of a substance, found by experiment or observation, that causes no adverse alteration of morphology, functional capacity, growth, development or lifespan of the target organism distinguishable from those observed in normal (control) organisms of the same species and strain under the same defined conditions of exposure.This definition permits a dual interpretation that is probably deliberately vague. The first is that, as described above, the “No” is a modifier of “Observed”. The second interpretation is that “No” is a modifier of “Effect”, so that the NOAEL becomes the Observed No Adverse Effect Level, which allows the NOAEL to be interpreted as an empirical threshold. That’s not what the NOAEL really is, but many toxicologists think that way.
The Benchmark Dose
The BMD is a three letter acronym, and two of them are superfluous. It might just as well be called The Dose. The Benchmark Dose was designed to be a replacement for the NOAEL (Crump, 1984). The basic concept is to use a dose-response model to fit or describe the data, and then pick a point at low end of the curve, typically the dose at some point where the effect can more or less be reliably measured. This process generally results in a dose that is roughly comparable to a NOAEL. The advantages of the BMD over the NOAEL are threefold:- Because a dose-response model is used to interpolate between doses, the BMD is not dependent on whatever doses happened to be used in a study.
- The BMD itself is a central estimate, so therefore it does not err on either the side of certainty or the side safety. Since statistical confidence limits are also generated, a lower-bound estimate (the BMDL) can be used when a precautionary approach is warranted.
- Unlike the NOAEL, the BMD does have a scientific interpretation. That is, it asserts that a particular effect will occur at a particular dose, so unlike the NOAEL it is a risk estimate. Of course, once a safety factor is applied that is no longer true, but that’s another story.
The main disadvantage of the BMD is that the dose estimate can be model dependent. In fact, the BMDL is often highly model dependent. That means that the model uncertainty problem that is supposedly avoided by using the BMD as a cutoff really isn’t really avoided at all. The BMD also has one underlying policy choice that the NOAEL does not have: In addition to picking the adverse effect and the statistical test, the response level that corresponds to the BMD, usually referred to as the benchmark response (BMR), must be selected as well.
Benchmark doses can be estimated for both quantal and continuous endpoints. The BMR for quantal endpoints is typically the BMD10, which corresponds to the dose at which 10% of the population exhibits the effect. Benchmark responses for continuous endpoints often arbitrarily define any deviation of more than two standard deviations as being adverse.
Benchmark Dose Software
Once the upfront policy choices have been made, the process for fitting curves to laboratory animal studies is pretty much standardized. There are two software programs that have developed for it:- EPA Benchmark Dose Software. This program has a better user interface and generates output to Microsoft Excel, and uses Excel for producing graphs. It also has a Visual Basic Interface that allows for fitting multiple alternative models.
- RIVM PROAST. The open code program is written in R, which means it can be adapted to other uses.
Both of these programs can also be used for dose-response modeling without the Benchmark Dose, which means no upfront policy choices are needed. How very cool.
References
Crump, K (1984). A new method for determining allowable daily intakes. Fundam. Appl. Toxicol. 4:854-871.Lehman AJ and Fitzhugh OG (1954). 100-Fold Margin of Safety. Quarterly Bulletin of the Association of Food and Drug Officials 18:33-35.
World Health Organization (2009). Principles and methods for the risk assessment of chemicals in food. International Programme on Chemical Safety. Environmental Health Criteria 240.
No comments:
Post a Comment