Open any supplement bottle and you’ll see a number, usually followed by a percentage. Open a supplement company’s blog and you’ll see an argument that the percentage is meaningless — that RDAs were set decades ago “just to prevent scurvy” and that the real requirement is several times higher. Both the label and the argument are doing something slightly dishonest, and untangling them is genuinely useful, because the difference between these frameworks determines whether a dose is sensible or wasteful.
The four numbers, briefly
Nutrient reference values come as a set, and the RDA is only one member.
EAR (Estimated Average Requirement) is the intake that meets the needs of half of healthy people in a group. It’s the statistically honest midpoint and the value nutrition scientists actually use when assessing whether a population is getting enough.
RDA (Recommended Dietary Allowance) is the EAR plus a safety margin — typically two standard deviations — set to cover roughly 97-98% of healthy people. Because it’s deliberately padded, most individuals need less than the RDA. It was never meant to describe any one person’s requirement.
AI (Adequate Intake) is used when the data are too thin to calculate an EAR. It’s an educated observation of what healthy people typically consume. Calcium for infants, potassium, and choline have historically sat in this category. An AI carries more uncertainty than an RDA.
UL (Tolerable Upper Intake Level) is the maximum daily intake unlikely to cause harm in the general population over the long term. This is the number that matters most once you start supplementing, and it’s the one labels rarely mention.
The critique that RDAs are “only about preventing deficiency” is partly fair — they are anchored to defined deficiency endpoints and adequacy markers rather than to optimizing athletic performance or mood. But the follow-on claim, that the real requirement is therefore much higher, doesn’t follow automatically. It needs its own evidence.
Where “optimal intake” claims come from
“Optimal” almost always means one of three things, and they aren’t equally credible.
Surrogate markers. A dose that maximizes some measurable thing — a blood level, an enzyme’s saturation, a biomarker — gets labeled optimal. This is reasonable reasoning but incomplete: maximizing a marker only matters if the marker tracks an outcome you care about. Plenty of markers move without health following.
Observational associations. People with higher intakes or higher blood levels of some nutrient have lower rates of some condition. This is a real signal worth investigating, but it’s badly confounded — people with high nutrient status tend to differ in a dozen other ways. Vitamin D is the cautionary tale here: strong observational associations across many outcomes, considerably more muted results when large randomized trials tested supplementation. Our vitamin D guide works through that gap in more detail.
Expert-panel opinion. Some clinical organizations publish intake targets above the RDA for specific populations. These are sometimes well-reasoned and sometimes just a different committee’s judgment call. Either way, they’re recommendations, not measurements.
None of this makes higher intakes automatically wrong. It means the burden of proof sits with the higher number, and a label claiming “optimal” has usually not met it.
When higher-than-RDA genuinely makes sense
There are real cases, and they’re specific rather than universal:
- Documented deficiency or insufficiency. Repletion doses for low ferritin, low B12, or low vitamin D are therapeutic and legitimately exceed the RDA — under clinical supervision, with follow-up testing.
- Malabsorption. Celiac disease, inflammatory bowel disease, and bariatric surgery all change the arithmetic substantially.
- Medication effects. Long-term proton pump inhibitors affect B12 absorption; metformin does too. Some diuretics increase mineral losses.
- Life stage. Pregnancy raises folate and iron needs meaningfully; older adults absorb B12 from food less efficiently.
- Restricted diets. Vegans need a reliable B12 source, full stop — that’s not an “optimal intake” argument but an adequacy one.
- Heavy sweat losses. Sodium and other electrolytes can genuinely run higher during sustained heat and endurance work.
Notice what these have in common: each is a specific, identifiable reason, usually verifiable. “I want to feel my best” is not on the list, because it doesn’t tell you which nutrient or how much.
The gap between RDA and UL is the practical variable
The most useful mental model isn’t RDA-versus-optimal. It’s how much room is there between adequacy and the ceiling?
Wide-window nutrients give you room. Vitamin C has an adult RDA around 75-90 mg and a UL of 2,000 mg/day — a large gap, and the main consequence of overshooting is GI upset. Magnesium’s UL of 350 mg/day applies only to supplemental magnesium, not food, and the limiting factor is loose stools well before anything serious. These are nutrients where modest experimentation is low-stakes.
Narrow-window nutrients don’t. Preformed vitamin A (retinol) has an adult UL of 3,000 mcg RAE, and chronic excess carries real liver and bone concerns — plus specific teratogenic risk in pregnancy. Selenium’s UL is 400 mcg/day, and the gap from the 55 mcg RDA is smaller than it looks once a multivitamin and a Brazil nut habit stack. Iron supplementation without documented need is a genuinely bad idea, and iron is a leading cause of pediatric poisoning at home. Zinc’s UL is 40 mg/day, and chronic intakes above it induce copper deficiency.
This framing does more work than any argument about whether the RDA is “too low.” A dose that sits comfortably between RDA and UL is defensible even without strong optimality evidence. A dose that crosses the UL needs a clinical reason.
What to actually do
Start with intake, not with pills. Most people can identify their genuine gaps by looking honestly at a week of eating. Our nutrient deficiency signs guide covers what’s worth noticing, with the caveat that most early deficiency symptoms are non-specific.
Test the few things worth testing. Vitamin D (25-hydroxyvitamin D), ferritin, and B12 are the ones where a blood value changes decisions. Routine testing of most other micronutrients isn’t clinically useful. See blood tests before supplementing.
Read labels for the total, not the single product. A multivitamin plus a fortified cereal plus a standalone mineral is where ULs get crossed accidentally — supplement upper limits has the numbers in one place.
Be skeptical of “%DV” as either endorsement or insult. 5,000% of a B vitamin isn’t proof of potency; it usually means the RDA is tiny and the excess is excreted. Equally, 100% isn’t proof a product is adequate for you specifically.
Bottom line
The RDA is a padded floor that covers nearly everyone, not a personal requirement and not a performance target — the critique that it wasn’t designed for optimization is fair. But “optimal intake” is usually built on surrogate markers and observational data rather than outcome trials, so it deserves to be treated as a hypothesis rather than a number. The practical question is where a dose sits between the RDA and the tolerable upper limit, and whether you have a specific, identifiable reason for going higher.
This guide is educational and not medical advice. Consult a healthcare provider before starting a supplement or changing a dose, especially if you are pregnant, nursing, taking medication, or managing a health condition.