Homeโ€บArticlesโ€บThe nutrient ceiling: where upper limits meet toxicity

The nutrient ceiling: where upper limits meet toxicity

Every pet food standard publishes a floor โ€” minimum crude protein, minimum calcium, minimum

vitamin A โ€” the numbers that say "a diet must contain at least this much to avoid deficiency."

Less visible is the ceiling: the maximum a diet may contain before that nutrient stops nourishing

and starts harming. The space between a published upper limit and a documented poisoning dose is

the margin of safety, and it varies by nutrient. This article compares the ceilings FEDIAF

2025 and AAFCO set for nutrients against the toxicity thresholds poison control centers have

measured for the same substances.

Two flavors of "maximum"

FEDIAF 2025 does not use a single kind of ceiling. In its nutrient tables it distinguishes a

nutritional maximum (N) โ€” "the highest level not expected to cause harmful effects, which

FEDIAF recommends commercial pet foods not exceed" โ€” from a legal maximum (L) set by EU

feed-additive legislation, which is "mandatory for all life stages and applying to the total

amount in the finished product when the nutrient is added as an additive."

Calcium: the growth ceiling, and why it was split

Calcium is the clearest example of a development-protecting ceiling rather than an acute-poison

guard. FEDIAF 2025 places a nutritional maximum on calcium in dogs of **1.60 g (N) for early

growth, 1.80 g (N) for late growth, and 2.50 / 1.60 g (N) for adults** (the 95/110 MER columns);

phosphorus gets no standalone maximum but is bounded by the Ca:P ratio, capped at **maximum 2/1

adult, 1.6/1 early growth & reproduction, 1.8/1a or 1.6/1b late growth, minimum 1/1**.

AAFCO tells the same story with a sharper edge. In the 2013 baseline the calcium maximum was "a

single 2.5% DM for all life stages"; the 2014 revision split it into **1.8% max for Growth &

Reproduction (and All Life Stages) products, 2.5% max for Adult Maintenance products only**, driven

by "expert-subcommittee concerns about large-breed puppy skeletal development from excess

calcium." The "Phosphorus maximum remained 1.6% DM." The ceiling is a chronic-developmental

guardrail, not a toxicity line โ€” over-calciumated large-breed puppies grow skewed skeletons.

Vitamin D3: a nutrient and a rodenticide in one molecule

Vitamin D is the nutrient that most dramatically shows why ceilings exist. FEDIAF 2025 sets a

dog nutritional maximum for Vitamin D of 320 IU (N), with a legal maximum of 227, and a

cat nutritional maximum of 3,000 IU (N). Excess of the parent vitamin is explicitly named as

the hazard: FEDIAF notes for Vitamin A that "excess causes hypervitaminosis A," and the same logic

of a nutritional maximum applies to its fat-soluble cousins.

Now read the toxicology. Cholecalciferol, the pet-toxicology topic states plainly, "also known as

Vitamin D3, is an active ingredient in some of the most potent mouse and rat poisons available

today." The same compound a standard caps at 320 IU per 100 g DM in dog food is, concentrated, a

deliberate poison. The overdose description is exact: "When animals ingest an overdose of

cholecalciferol, they develop gastrointestinal distress (e.g., vomiting, anorexia, and diarrhea),

electrolyte disturbances (e.g. elevation in calcium and phosphorous), and kidney injury," and

"Without prompt treatment the overdosed animal dies secondary to kidney failure."

The margin of safety here is the gap between 320 IU (N) โ€” the level "not expected to cause harmful

effects" โ€” and a rodenticidal overdose large enough to drive calcium and phosphorus up and shut

the kidneys down. The standard does not publish the poisoning number; it only guarantees the

finished food stays clear of it when vitamin D is added as a regulated additive (legal maximum

227). For a fat-soluble vitamin the gap is the whole ballgame, because the harm is accumulation,

not a single meal.

Sodium and salt: the ceiling that is simply absent

Salt is where the margin of safety gets uncomfortable. AAFCO Chapter 6 defines salt as "an

acceptable source of sodium chloride; it must be true to name and contain not less than 95%

sodium chloride," yet "The AAFCO definition sets no inclusion-rate or tolerance limits." AAFCO does

publish minimums โ€” dogs need Na 0.3% DM growth / 0.08% DM adult and Cl 0.45% / 0.12% DM โ€”

and notes that "Excessive salt intake increases thirst and urination, and very high intakes can

cause sodium toxicosis." FEDIAF is blunter: it "sets no maximums for sodium, potassium, magnesium,

or cobalt in these tables."

So the standards put a floor under sodium but no published ceiling; the ceiling lives in

formulation tolerance. The toxicology is unambiguous. The ASPCA warns that "Excessive salt intake

can produce increased thirst and urination, and sometimes abnormal electrolytes in the blood

depending on the amount ingested," and "Signs of salt toxicity can include vomiting, diarrhea,

depression, tremors, seizures and even death." The same substance AAFCO declines to cap is, in

excess, a toxin with a syndrome ending in death. The margin of safety for sodium is therefore not

a printed number but the distance between a palatability-driven inclusion level and the point at

which "very high intakes" tip into "sodium toxicosis"; reduced-sodium formulas substitute "part of

the salt with potassium chloride to maintain electrolyte balance and palatability."

Onion and garlic: poisons with measured thresholds

Onion and garlic are the mirror image of the nutrients above. They have no "upper limit" in any

feed profile because they are not nutrients at all โ€” yet their poisoning dose is the best

quantified of everything here. The onion/garlic topic reports that "Onion (Allium cepa) and garlic

(Allium sativum) contain compounds that damage red blood cells and cause haemolytic anaemia with

Heinz bodies," and gives hard numbers: "Consumption of as little as 5 g/kg of onions in cats or 15

to 30 g/kg in dogs has resulted in clinically important hematologic changes," while "A relatively

high single/short-term dose (600โ€“800 g) can damage red blood cells." (Garlic is "toxic to horses

at >0.2 g/kg.")

Contrast this with calcium or vitamin D. For onion and garlic there is no nutritional maximum to

consult โ€” the relevant figure is the toxic dose, because the substance has no required dietary

role. The margin of safety is effectively zero by design: the only numbers on record are the ones

that hurt.

Why the margins are different

Put the ceilings side by side and a pattern emerges:

  • Calcium is capped to protect growth (1.8% AAFCO large-breed; 1.60โ€“1.80 g FEDIAF growth),

backed by a second standard โ€” the Ca:P ratio, max 2/1 adult.

  • Vitamin D is capped tightly (320 IU N dog, 227 legal; 3,000 IU N cat) because the same

molecule is a recognized rodenticide whose toxic endpoint โ€” kidney failure โ€” is severe.

  • Trace elements carry the tightest legal maxima: FEDIAF's EU legal maxima are **Zinc 22.70 mg

(L) = 227 mg/kg, Copper 2.80 mg (L) = 28 mg/kg, Iron 68.18 mg (L), Manganese 17.00 mg (L), Iodine

1.10 mg (L), Selenium 56.80 ยตg (L) = 0.568 mg/kg.** Selenium at 0.568 mg/kg shows how narrow a

trace-element margin can be.

  • Sodium has no published ceiling from either FEDIAF or AAFCO's definition, yet ASPCA

documents a full salt-toxicity syndrome โ€” the margin is managed by formulation, not a number.

  • Onion/garlic have only a toxic threshold and no nutrient ceiling, because they are poisons,

not nutrients.

A "maximum" on a pet food label is not a uniform promise. Some maxima defend against acute

poisoning (vitamin D, selenium), some against chronic deformity (calcium), some are absent and

delegated to the formulator (sodium), and some substances sit outside the nutrient system entirely

and are governed only by their measured toxic dose (onion, garlic). The gap between the printed

ceiling and the documented poison is the margin of safety โ€” and knowing which kind of gap you are

looking at is the difference between a diet that is merely complete and one that is safe.

Sources

All figures and quoted wording are drawn verbatim from these repository topics (C1 verified

against their source files):

  1. FEDIAF_2025_micronutrient_upper_limits โ€” nutritional (N) and legal (L) maxima for vitamins

A, D and trace elements; the N-vs-L distinction.

  1. FEDIAF_2025_mineral_limits โ€” calcium maxima by life stage, Ca:P ratio caps, and the note

that FEDIAF sets no maximums for sodium, potassium, magnesium, or cobalt.

  1. AAFCO_2013_baseline_dog โ€” the single 2.5% DM calcium maximum across all life stages.
  2. AAFCO_2014_Ca_P_split โ€” the split to 1.8% (G&R / All Life Stages) vs 2.5% (Adult

Maintenance), driven by large-breed puppy skeletal-development concerns; phosphorus max 1.6% DM.

  1. AAFCO_Ch6_sodium_chloride โ€” salt defined as โ‰ฅ95% sodium chloride with no inclusion-rate or

tolerance limit; AAFCO sodium/chloride minimums; sodium toxicosis at very high intakes.

  1. pet_tox_cholecalciferol โ€” cholecalciferol = Vitamin D3, a potent rodenticide; overdose signs

and kidney-failure outcome.

  1. pet_tox_aspca_citrus_coconut_salt โ€” signs of salt toxicity (vomiting, diarrhea, depression,

tremors, seizures, death).

  1. pet_tox_onion_garlic_thresholds โ€” Allium red-blood-cell damage and numeric toxic thresholds

(5 g/kg cats, 15โ€“30 g/kg dogs, 600โ€“800 g single dose).