It is commonly admitted that

The dose makes the poison

which means as a person, the more I take a substance, the more risk I take for my health.

There is even an indicator called LD50 (see Median Lethal Dose), which specifies the 'lethal dose' at which 50 percent of subjects will die.

This might seems a stupid question but: are there any substances for which it is more dangerous to take at low dose than at higher dose? For example, you have to ingest at least 1µg of XXX because if you ingest less you will have health problem...

  • 11
    $\begingroup$ Oxygen comes to mind. A low dose is much more harmful than a large one :). $\endgroup$ – terdon Sep 28 '12 at 19:09
  • 3
    $\begingroup$ sugar, water, protein, ... these all have an optimal range of healthy doses- too much OR too little can be harmful. $\endgroup$ – David LeBauer Sep 30 '12 at 5:04
  • 4
    $\begingroup$ I think we need a clarification. Do you mean are there any substances which are themselves toxic or harmful at low doses but which are not toxic or harmful at high doses? You final sentence ("For example you have to ingest ...") makes things a bit ambiguous: obvious examples of such compounds are the vitamins. A lack of of an adequate supply of niacin, for example, causes pellagra, at one time a major disease in Europe and the American south. However, as Rory M pointed out, I don't think this is what you have in mind. $\endgroup$ – user1136 Oct 1 '12 at 10:16
  • 8
    $\begingroup$ Dihydrogen Monoxide is a brutal killer. $\endgroup$ – bobthejoe Oct 3 '12 at 6:56
  • $\begingroup$ All of these substances in lower than required concentrations are not of themselves toxic, but rather prevent a subsequent reaction from occurring and is (indirectly) lethal. As such, not real answers. ;) $\endgroup$ – user560 Oct 27 '12 at 20:34

Yes. Taking antibiotics (and by extension antivirals and antineoplastics) at low doses is dangerous, as this creates an evolutionary pressure to evolve antibiotic resistance (or rather, incorporate through horizontal gene transfer). This is dangerous because an infection resulting from these bacteria will then be harder to treat.

In the paper "Experimental evolution of resistance to an antimicrobial peptide", Perron et al. gave E. coli the chance to evolve resistance against a novel antibiotic. When given a high dose initially, most bacteria died. When exposed to low doses, the bacteria evolved resistance in 600 to 700 generations.

For the same reason it is strongly advised to finish a course of antibiotics even after symptoms fade.

  • 2
    $\begingroup$ +1 although I don’t think this is what OP was referring to because the effect is indirect rather than there being a toxic effect of the antibiotic of the host which decreases with increased dose. $\endgroup$ – Konrad Rudolph Oct 2 '12 at 16:46
  • 1
    $\begingroup$ I agree, but my answer is true to his original question $\endgroup$ – Michael Kuhn Oct 3 '12 at 17:33

This is a bit of a news flash - the answer is probably yes.

Reading through recent Nature articles this one popped up: Toxicology the learning curve.

Where there are some intriguing examples of low concentration only toxic effects.

It might be worth a read. The preface gives a review of the idea, going back to the 16th century physician Paracelcius that the toxic properties of anything increases monotonically with dosage. This is reflected in most of the answers and its one of the few things which toxicologists universally expect to see.

A research group at U Missouri-Columbia is claiming that endocrine disruptors have non-monotonic toxicology curves, meaning that toxic effects actually decrease with greater dosage.

non monotonic curves for endorcine disruptors Nature is kind enough to allow access to this image. The three curves on the right show a decline in drug effect as concentration increases.

Endocrine hormones act at very low dosage - testosterone, estrogen, pituitary and thyroid hormones are some examples. We'll probably be hearing more about this in the future, rather than less.

While this is still controversial, its made news because it is the sole class of compounds which really behave this way. notably the first curve is for BPA (Bisphenol A) - the plastcizer which is common in pool toys and toy inflatable balls and drink bottles but the FDA has flagged for closer study.


No. The influence of dose–response relationship on toxicity is one of the fundamental properties of (bio)chemistry, and follows directly from the fact that concentration influences the reaction rate: a chemical reaction between two or more agents is a stochastic process where agents have a certain chance of interacting based on physical proximity (and other factors). The higher the dose, the higher the chance of two agents meeting, and a reaction taking place.

This has two interesting consequences:

  • Homeopathy doesn’t work.
  • Although not all substances show a clear dose–response curve (simply because they are always reactive no matter their dose), toxicological studies which fail to show a dose–response relationship are met with scepticism. A recent example of this is the GMO toxicity study undertaken by the French scientist Séralini. One of the many flaws in this paper cited by other scientists was precisely this lack of dose–response relationship.
  • $\begingroup$ Example of something always reactive no matter the dose? $\endgroup$ – David LeBauer Sep 30 '12 at 18:09
  • $\begingroup$ @David No, sorry. “no matter the dose” is also stretching it a bit – you always need some minimal amount, it’s just that the steepness of the curve varies. $\endgroup$ – Konrad Rudolph Sep 30 '12 at 19:15
  • $\begingroup$ thanks for the clarification; as written it can be interpreted as "homeopathy" doesn't work, except when it does. On another note (more as a clarification for readers than as a correction to you) the GMO study is a poor example to use. Because of the many flaws, any inference is limited (e.g. my primary conclusion is that there is insufficient information to conclude anything about the relationship dose-response relationship). $\endgroup$ – David LeBauer Sep 30 '12 at 20:45
  • $\begingroup$ No. The influence of dose–response relationship on toxicity is one of the fundamental properties of (bio)chemistry...?!? $\endgroup$ – Lorenz Lo Sauer Oct 1 '12 at 15:54
  • $\begingroup$ @Lo Unclear? It means that things get more toxic with increased dose, never the other way round. $\endgroup$ – Konrad Rudolph Oct 1 '12 at 16:07

I think the best answer is right now would be

Probably no, but we're not certain

The question of non-monotonic dose-response is still being discussed and debated. You will find a good example of this at EFSA


First of all, that is an excellent question! Secondly, please excuse the lack of thorough sourcing for this post. I only like to share some pointers...

Q: Can substances do more harm in a mammal at a lower dose than at higher dose?

A: Yes.

Some may know the example of the blood thinner, and widespread NSAID Acetyl-salicylic acid (Aspirin). I will use this example because Aspirin is one of the most known and used drugs in the world, with a long and interesting history. As being part of the class of nonsteroidal anti-inflammatory drugs (NSAIDs) it has recently gotten a bad wrap for blocking an enzyme which is crucial for prostaglandin synthesis and processes that maintain a healthy stomach wall in the presence of an potentially harmful bacterial and chemical milieu. On a sidenote, as an acid, Aspirin actually hardly dissolves directly in the acidic stomach.

To quote wikipedia (DOR: 28/09/2012):

Salicylates are excreted mainly by the kidneys as salicyluric acid (75%), free salicylic acid (10%), salicylic phenol (10%), and acyl glucuronides (5%), gentisic acid (< 1%), and 2,3-dihydroxybenzoic acid.[148] When small doses (less than 250 mg in an adult) are ingested, all pathways proceed by first-order kinetics, with an elimination half-life of about 2.0 to 4.5 hours.[149][150] When higher doses of salicylate are ingested (more than 4 g), the half-life becomes much longer (15–30 hours),[151] because the biotransformation pathways concerned with the formation of salicyluric acid and salicyl phenolic glucuronide become saturated

Aspirin can interfere with a lot of drugs, including Ibuprofen. The known list of drugs is so high, in part due to the widespread application of aspirin, and in part due to the chemical composition of this very small drug with two functional groups. A functional group is a sort of chemical feature that determines how and with what probability a chemical interacts in a given milieu.

As a certain metabolic degradation pathway may get saturated, other pathways can take over and lead to different potentially harmful metabolites resulting in damage to the excretion/clearance organs such as the liver (hepatotoxic) and kidney (nephrotoxic). A metabolite is the product of metabolism ( a weak back-referencing definition). metabolic pathways are a sequence of chemical reactions within a cell, and metabolism is the entire set of metabolites and pathways.

Liver/Kidney Toxicity is believed to result from an interplay of a) high concentrations of metabolites or direct toxins (=parent chemicals) within specialized high-concentration-forming cells, and b) the non/covalent binding of cellular macromolecules to interfere with their normal function and/or production of reactive oxygen species (ROS).

Another example (from memory) is the interaction pharmacokinetics of theobromine (from cacao) and other xanthines (unverified).

There are many more examples. I hope this example and mechanism serves you well as a starting point...

From a biochemical standpoint:

  • In fact drugs may act differently at different concentration thresholds at different times at different places within an organism and its cells that make it up. Evolution exploited this very early on and used it for things like body segmentation along defined axis during development.

  • Drugs may interact with each other differently at different levels of concentrations, say as binary, ternary, and higher order complexes. The rate of formation of such complexes is denoted through the dissociation constant - which is a type of equilibrium constant such as pH is.

    The dissociation constant describes the propensity of a chemical species to separate reversibly and, - this is crucial! - is concentration dependent. This is accounted for by the use of so called activities or the apparent concentrations, as a result of incomplete hydration of the chemical and other effects of a "crowded environment". Now don't just think of chemicals, but macromolecules as well. In immunology for instance new molecular features can arise on the surface of a macromolecule, due to the association (complexation) with a small molecular drug. This in turn can lead to an immunological reaction, even anaphylaxis ( a so called Type I hypersensitivity reaction) and ultimately shock. Such drugs are classified as haptens.

Important in an organisms sensitivity and susceptibility to drugs however, is the current allometric growth rate and types of organs and cells available. As such a fetus is more susceptible than an adult.

I hope a simple yes accompanied by this admittedly to-long to really read post gave you some insight into what scientist across many, many fields are dealing with. Naturally, the issue is more complex.

Personally I believe the issue of dosing and drug reactions will improve significantly over the coming decade...

  • 6
    $\begingroup$ Maybe I'm misreading your post, but I don't see any example of a substance that is more harmful at lower doses in your post. $\endgroup$ – Mad Scientist Sep 28 '12 at 19:06
  • 2
    $\begingroup$ I may be missing something here, but I can find no mention of how lower doses of aspirin are more harmful than higher ones in your answer. I assume that one of the metabolites it gets broken into activates a different pathway and can lead to toxicity. Is that what you mean? $\endgroup$ – terdon Sep 28 '12 at 19:08
  • $\begingroup$ Correct, I didn't include a specific example only an example mechanism. Substances acting differently at different levels is well known within the concerned fields of science. Substances/Drugs which are toxic at low levels cannot be approved nor are they generally appealing for studies. Substance's "dualism" can offset the LD50. The post also doesn't discuss how substances can be toxic at low dose or discerns entantiomeric purity (e.g. thalidomide) and/or entantiomer-specific pathways. Specific examples may best be extrapolated from systems-biological and/or chemotherapeutics literature... $\endgroup$ – Lorenz Lo Sauer Sep 29 '12 at 6:02
  • 2
    $\begingroup$ The question is not about substances acting different at lower concentrations but specifically about being more harmful/dangerous at lower concentrations. Your examples don't fit that, strictly speaking you're not answering the question that has been asked. $\endgroup$ – Mad Scientist Sep 29 '12 at 7:51

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.