Lead
"This ingredient is dangerous." "No additives, so it's safe." Parenting media serves up lines like these almost daily.
When the only frame on offer is a binary — dangerous or safe — the information a reader can actually use is impoverished. Before accepting that frame, it is worth pausing on a sentence written nearly 500 years ago.
In 1538, the physician and alchemist Paracelsus (Philippus Aureolus Theophrastus Bombastus von Hohenheim) wrote in his Septem Defensiones:
"Alle Ding sind Gift und nichts ohn Gift; allein die Dosis macht, daß ein Ding kein Gift ist." (All things are poison, and nothing is without poison; only the dose makes a thing not a poison.)
He wrote this as a rebuttal to critics who called him a poisoner for prescribing chemical substances. The medicinal herbs his contemporaries used were themselves lethal in sufficient quantities. What mattered was not the name of the substance but how much of it entered the body [1].
Every concept that modern toxicology has accumulated — LD50, NOAEL, ADI, safety factors, dose-response curves — is a detailed elaboration of that one sentence. Understanding this framework is the foundation for reading any claim about the safety of food additives or household products.
How Toxicity Is Measured: From LD50 to NOAEL
The most basic measure in toxicology is the LD50 (median lethal dose): the amount of a substance that kills 50% of a test-animal population, expressed in mg per kg of body weight, used to compare acute toxicity: the amount of a substance that kills 50% of a test-animal population, expressed in mg per kg of body weight. Table salt has an LD50 of roughly 3,000 mg/kg in rats; caffeine, about 200 mg/kg [1]. LD50 allows comparison between substances but does not directly answer the question "how much is safe to consume?"
The metric at the center of practical safety evaluation is the NOAEL (No Observed Adverse Effect Level): the highest tested dose of a substance producing no detectable harmful effects in animal studies conducted over weeks to years: the highest dose at which no adverse effects are observed when a substance is administered chronically to animals. Toxicity testing covers multiple layers — acute toxicity (LD50), subchronic (90-day studies), chronic (2-year studies), carcinogenicity, and reproductive and developmental toxicity. The NOAEL is set at the highest dose that passed all of these without producing harm.
The Structure of the 100-Fold Safety Factor
In 1954, Lehman and Fitzhugh of the US Food and Drug Administration proposed applying a 100-fold safety factor (margin of safety) when deriving a safe human intake from an animal NOAEL [2]. That number has an explicit internal logic.
- Interspecies factor: 10-fold — The NOAEL in rats or mice does not necessarily translate directly to humans. Differences in body size, metabolic rate, and organ function must be accommodated.
- Intraspecies factor: 10-fold — Within the human population, sensitivity varies considerably among infants, elderly individuals, pregnant people, and those with underlying conditions.
Multiplying these two factors of 10 gives the standard safety factor of 100. Dividing the NOAEL by 100 yields the ADI (Acceptable Daily Intake): the amount of a food additive that can be ingested daily over a lifetime without detectable health risk, expressed in mg per kg of body weight per day [2,3].
Renwick subsequently refined this 10-fold breakdown further. The interspecies factor of 10 is distributed between toxicokinetic: relating to how the body absorbs, distributes, metabolizes, and excretes a substance over time differences (how the body absorbs, metabolizes, and excretes a substance) and toxicodynamic: relating to how sensitively the target organ or receptor responds to a given concentration of substance differences (how sensitive the target organ is), allocating roughly √10 (approximately 3.16-fold) to each component [3]. The same logic applies to the intraspecies factor. Where adequate data exist for a specific substance, these default factors can be replaced with more precise values.
What ADI Is — and What It Is Not
The ADI is defined as the amount that can be ingested daily over a lifetime without detectable harm to health, expressed in mg per kg of body weight per day [4]. The Joint FAO/WHO Expert Committee on Food Additives (JECFA) evaluates and publishes ADI values for individual food additives.
A critical feature of the ADI is its "lifetime" assumption. It represents a value already calculated for continuous daily intake over decades — and then further divided by that 100-fold safety margin. In practice, actual exposure is usually substantially below the ADI.
The TDI (Tolerable Daily Intake) uses the same conceptual structure but applies to environmental contaminants and naturally occurring harmful substances rather than intentionally added food additives.
Hormesis: When Low Doses Reverse the Effect
Toxicology also contains phenomena that do not fit neatly into a simple linear "more dose, more toxicity" model. Hormesis: a biphasic dose-response pattern in which a substance causes stimulation at low doses and inhibition or harm at high doses, the opposite of a simple linear toxicity curve describes a biphasic dose-response relationship in which low doses produce a stimulatory or beneficial effect, while high doses produce inhibition or toxicity.
Calabrese and Baldwin systematically organized this phenomenon and analyzed large toxicology databases, demonstrating that hormesis is not an exceptional outlier but is widely observed [5,6]. The magnitude of the low-dose stimulatory response is typically modest — around 1.5 times the control value — and the stimulatory zone is limited in range.
However, the discovery of hormesis does not translate directly into "a small amount is fine." As Grandjean points out, substances such as endocrine disruptors may cause problems even at low doses; developmental exposures carry different risks from adult exposures; and the effects of combined exposures remain difficult to characterize [1]. These complications are an honest part of contemporary toxicology.
Stepping Back from "Dangerous or Safe"
In the actual practice of risk assessment, when asked whether a substance is "safe" or "dangerous," a toxicologist uses language about the magnitude of health risk at a specific dose and exposure context — not a label attached to the substance itself [4].
As Vermeire and colleagues note, the safety-factor framework is deliberately designed to incorporate uncertainty and interindividual variation [3]. The question "what percentage of the ADI does this ingredient represent?" is more informative than the question "is it dangerous?"
Parents making everyday choices do not need to master this framework completely. What helps is recognizing that any claim framed purely as "dangerous" or "safe" probably omits the relevant context: the substance's name, the dose, the actual exposure level, and what is being evaluated (acute versus chronic, infant versus adult).
Recording matters here too. When a household's food and product use is logged over time, it becomes possible to estimate what a specific child is actually exposed to and at what quantities. Estimating exposure is the first step in any risk assessment.
Summary
For 500 years, since Paracelsus, the core of toxicology has not changed. The question is not the name of the substance but the dose and the exposure context. NOAEL, the 100-fold safety factor, and the ADI represent humanity's accumulated effort to answer that question quantitatively.
Most claims of the form "it's safe because it contains no additives" or "it's dangerous because this ingredient is present" bypass this framework entirely. Developing the habit of asking, whenever such claims appear, "what is the dose?" and "what is the safety factor?" is what it means to be an honest consumer of risk information.
References
- Grandjean P. Paracelsus Revisited: The Dose Concept in a Complex World. Basic Clin Pharmacol Toxicol. 2016;119(2):126–32. doi:10.1111/bcpt.12622. PMID: 27214290.
- Lehman AJ, Fitzhugh OG. 100-Fold margin of safety. Assoc Food Drug Off US Q Bull. 1954;18:33–35. [No PMID; 1954 publication predates PMID indexing. Entry confirmed via Semantic Scholar citation records.]
- Vermeire T, Stevenson H, Pieters MN, et al. Assessment factors for human health risk assessment: a discussion paper. Crit Rev Toxicol. 1999;29(5):439–490. doi:10.1080/10408449991349256. PMID: 10521133.
- JECFA (Joint FAO/WHO Expert Committee on Food Additives). General guidance for the preparation of safety evaluation by the Joint FAO/WHO Expert Committee on Food Additives. WHO Food Additives Series. https://www.inchem.org/documents/jecfa/jecmono/v08je01.htm
- Calabrese EJ, Baldwin LA. Hormesis: U-shaped dose responses and their centrality in toxicology. Trends Pharmacol Sci. 2001;22(6):285–291. doi:10.1016/S0165-6147(00)01719-3. PMID: 11395156.
- Calabrese EJ, Baldwin LA. Defining hormesis. Hum Exp Toxicol. 2002;21(2):91–97. doi:10.1191/0960327102ht217oa. PMID: 12102503.
- Renwick AG. Data-derived safety factors for the evaluation of food additives and environmental contaminants. Food Addit Contam. 1993;10(3):275–305. doi:10.1080/02652039309374152. [No PMID; pre-MEDLINE indexing period for this journal volume. Journal and author confirmed via citation records.]