EAR vs PRI: Two paths to vitamin D reference values

In practice, the answer depends on which scientific framework is being used, what outcome the framework is trying to protect, and how confidently researchers can describe the distribution of vitamin D requirements across a population.

That distinction matters because an EAR vs PRI vitamin D reference intakes comparison is not a contest between two interchangeable numbers. The Estimated Average Requirement, or EAR, is designed to meet the needs of approximately half of a healthy population group. The Population Reference Intake, or PRI, is intended to cover nearly all healthy people in that group, conventionally around 97.5%. These values belong to different points in the same methodological pathway, while the Adequate Intake, or AI, used by EFSA for vitamin D reflects a different evidentiary position altogether.

For policymakers deciding whether and how to fortify foods, this is more than technical vocabulary. It determines how deficiency is estimated, how safety margins are interpreted, how labels communicate intake targets, and whether a population intervention is built around a demonstrated requirement or a pragmatic level adopted under uncertainty.

Defining the framework: EAR, PRI, and AI are not substitutes

Dietary reference intakes are often presented in tables as if each value were simply a different daily target. That presentation is convenient, but it can obscure the statistical logic underneath.

The EAR, called the Average Requirement or AR in EFSA terminology, is the intake level estimated to meet the physiological requirement of 50% of healthy individuals in a defined population group. It describes the centre of the requirement distribution, not a universal target for every person.

The PRI, which corresponds broadly to the US Recommended Dietary Allowance, or RDA, is derived from the average requirement plus two standard deviations:

PRI = AR + 2 SD

This construction is intended to cover the requirements of approximately 97.5% of healthy individuals in the population group. It is therefore a population-protective reference value, not an estimate of the amount needed by the average person.

The distinction becomes particularly important when we move from assessment to policy. An EAR can be useful for estimating the prevalence of inadequate intake in a population, provided the assumptions behind the requirement distribution are appropriate. A PRI is more suited to setting an individual-level reference that covers almost everyone, though it should not be interpreted as a guarantee that every person consuming less is deficient.

An AI is different again. It is used when the evidence is not sufficient to establish an EAR or AR and, consequently, a PRI. Instead of being calculated from a quantified requirement distribution, an AI is derived from observed or experimentally supported intake levels associated with a defined physiological or health-related benchmark.

For vitamin D, these differences are not semantic. They are the reason two respected authorities can publish reference values that look similar numerically while meaning something different methodologically.

A vitamin D reference value is never just a number; it is a statement about how much we know, how much uncertainty remains, and whom the policy is designed to protect.

The central distinctions

The table is useful, but it should not flatten the differences. The same unit—micrograms per day—does not make values equivalent. A 15 µg/day AI and a 15 µg/day RDA may lead to similar dietary planning in some settings, yet they carry different implications for population nutrient requirements calculation and for interpreting inadequate intake.

The IOM approach: a formal requirement distribution for vitamin D

The US Institute of Medicine’s 2011 Dietary Reference Intakes provide the clearest example of the EAR-to-RDA pathway for vitamin D.

For adults and children from 1 through 70 years, the IOM established:

• an EAR of 10 µg/day, equivalent to 400 IU/day;
• an RDA of 15 µg/day, equivalent to 600 IU/day.

For adults older than 70 years, the RDA rises to 20 µg/day, or 800 IU/day. The adult tolerable upper intake level is 100 µg/day, or 4,000 IU/day.

Under this framework, the EAR is the value expected to meet the requirement of half of a healthy population group. It is therefore the more appropriate reference for estimating the proportion of a population whose usual intake is inadequate, assuming the relevant statistical conditions are satisfied. The RDA, by contrast, is set higher so that it covers almost all healthy individuals in that group.

That difference has practical consequences. If a national dietary survey shows that many people consume less than the RDA, the result does not automatically mean that the same proportion is deficient. The RDA is deliberately positioned high in the distribution. Nor does consumption above the EAR establish that an individual has achieved an adequate vitamin D status, because intake is only one contributor to circulating 25-hydroxyvitamin D, or 25(OH)D.

The IOM framework also assumes that the dietary reference value can be linked to a defined physiological requirement under stated conditions. For vitamin D, those conditions include assumptions about sunlight exposure and the contribution of cutaneous synthesis. This is one reason vitamin D standards are difficult to compare across regions: the body’s production of vitamin D through sunlight varies by season, latitude, skin pigmentation, clothing, outdoor activity, and other contextual factors.

We should therefore read the IOM values as products of a particular evidence base and set of assumptions, rather than as globally transferable constants. They are formal reference standards, but they still operate within a model of vitamin D biology that must be interpreted in context.

Why the EAR matters in population assessment

In public health work, the EAR is often the more informative starting point because it is designed around the median requirement. If usual intake falls below the EAR for a substantial share of a population, that can signal a meaningful risk of inadequate intake, although the inference depends on the quality of the dietary data and on the underlying assumptions of the EAR cut-point approach.

The RDA is less suitable for estimating population inadequacy because it is constructed to cover almost all individuals. Using it as a simple threshold can exaggerate the apparent scale of inadequacy. This is a familiar problem in nutrition surveillance: a value intended to protect nearly everyone becomes treated as the minimum requirement for everyone.

For fortification policy, the temptation is understandable. A higher reference value appears safer and more protective. Yet a policy cannot be designed responsibly by selecting the largest number in the table. Fortification must also consider baseline intake, the number and type of foods being fortified, total exposure across the diet, the UL, vulnerable groups, and whether the intervention reaches communities facing the greatest nutritional inequity.

The EFSA stance: why vitamin D lacks formal AR and PRI values

EFSA took a more cautious position in its vitamin D assessment. The EFSA Panel on Dietetic Products, Nutrition and Allergies concluded that an Average Requirement and a Population Reference Intake could not be derived because of variability and uncertainty in the available evidence.

Instead, EFSA established an Adequate Intake of 15 µg/day, or 600 IU/day, for individuals aged 1 year and older. For infants aged 7–11 months, the AI is 10 µg/day, or 400 IU/day.

The adult AI was based on achieving a target serum 25(OH)D concentration of 50 nmol/L, under assumed conditions of minimal cutaneous vitamin D synthesis. That assumption is a deliberate policy choice in the reference-setting process: it allows the dietary contribution to be considered without relying heavily on sunlight to meet the target.

But an AI should not be described as an EFSA PRI. EFSA did not claim that 15 µg/day covers 97.5% of the population in the same way that a PRI or RDA is intended to do. The value is a practical reference established in a context where the evidence did not support a formal estimate of the requirement distribution.

This is the point most likely to be lost in a simplified EFSA PRI vs IOM EAR vitamin D comparison. The apparent numerical similarity between EFSA’s AI and the IOM’s RDA does not erase the methodological divide:

• the IOM established an EAR and then an RDA;
• EFSA concluded that an AR and PRI could not be established;
• EFSA used an AI linked to a serum 25(OH)D target under minimal sunlight assumptions.

These values may be used in overlapping policy discussions, but they answer different scientific questions.

An AI is not a failed PRI

It is easy to frame the absence of a PRI as a weakness in the evidence system, but that would be too simplistic. The AI is not an incomplete RDA waiting for a final calculation. It is a transparent response to uncertainty.

Vitamin D requirements are affected by factors that are difficult to standardize internationally. Sun exposure changes the relationship between dietary intake and serum concentration. The evidence linking serum 25(OH)D to health outcomes is not uniform across all groups or endpoints. Laboratory methods may differ, and the observed relationship between intake and status can vary according to baseline vitamin D stores and population characteristics.

By setting an AI, EFSA provides a usable reference while acknowledging that the statistical requirements for an AR and PRI have not been met. From a policy perspective, that honesty is valuable. A number that openly carries uncertainty is more useful than a number presented with unjustified precision.

The challenge is communication. Once an AI appears in a dietary guideline, a food label, or a fortification model, it can easily be treated as though it had the same status as a PRI. That is where regulators, health professionals, and researchers need to be precise, particularly when comparing national dietary surveys or evaluating the likely effect of mandatory vitamin D fortification.

Divergent thresholds: the role of serum 25(OH)D

The methodological divide between authorities is also visible in the serum thresholds used to derive or support vitamin D reference values.

EFSA based its AI on achieving a serum 25(OH)D concentration of 50 nmol/L, assuming minimal sunlight exposure. The UK Scientific Advisory Committee on Nutrition, or SACN, derived a Reference Nutrient Intake of 10 µg/day, or 400 IU/day, for individuals aged 4 years and older, using a population-protective serum 25(OH)D threshold of 25 nmol/L.

These thresholds are not identical, and the difference is not a minor technical footnote. A reference intake tied to 50 nmol/L is built around a higher serum target than one tied to 25 nmol/L. Consequently, the resulting intake recommendations should not be compared as though they were generated from the same definition of adequacy.

That does not mean one authority is automatically correct and the other incorrect. It means the interpretation of vitamin D sufficiency depends partly on the chosen health or physiological benchmark. When governments assess deficiency prevalence, they need to state which serum threshold is being used, which assay or measurement approach underpins the estimate, and whether the threshold is intended to represent risk of deficiency, population protection, or another outcome.

Why serum targets complicate dietary policy

Dietary fortification operates upstream, while serum 25(OH)D is a biological outcome downstream. The path between them is shaped by the food environment and by social conditions.

A fortified food may be available but unaffordable. A national label may list vitamin D content clearly but fail to reach people who rely on informal food markets. A public programme may achieve good average intake while leaving homebound older adults, people with darker skin living at higher latitudes, or communities facing food insecurity with limited practical access to fortified products. These are not marginal considerations. They are systemic barriers that determine whether a policy produces nutritional equity or merely improves the national mean.

The relationship between dietary intake and serum status also changes across seasons and population groups. A fortification model based on annual average intake may not protect people during periods of minimal sunlight. Conversely, a policy that stacks multiple fortified foods without considering supplements and the UL may increase exposure in groups already receiving substantial vitamin D.

This is why reference values should guide policy, not replace it. The number establishes a planning point; it does not describe the whole intervention.

The strongest fortification policy is not the one with the most confident number, but the one that makes its assumptions visible and tests whether the people at greatest risk are actually reached.

Comparing the pathways without collapsing them

A practical dietary reference intakes comparison needs to show both the values and the reasoning behind them.

The table shows why a single ranking of values would be misleading. The IOM EAR, IOM RDA, EFSA AI, and UK RNI are not interchangeable markers on a single scale. They represent distinct decisions about evidence, population coverage, serum status, and uncertainty.

For researchers comparing national dietary surveys, the first task is therefore not to ask which number is higher. It is to map the reference value to its purpose:

1. For estimating inadequate intake, determine whether an EAR or AR exists and whether the survey data support the required method.

2. For communicating an individual-level target, identify whether the value is an RDA, PRI, RNI, or AI, and explain its intended coverage.

3. For modelling fortification, combine the reference value with baseline intake distributions, food consumption patterns, and the UL.

4. For interpreting serum data, state the 25(OH)D threshold and recognize that it may not match the threshold used to derive the dietary recommendation.

5. For comparing countries, document sunlight assumptions, age categories, analytical methods, and the role of supplements.

The last point is especially important in international work. Two countries can report similar dietary intakes while having different vitamin D status distributions because their populations differ in seasonal exposure, food patterns, supplement use, and skin pigmentation. Conversely, two populations with comparable serum values may have reached them through very different combinations of diet and sunlight.

What this means for vitamin D fortification policy

Mandatory vitamin D fortification is often discussed as though the policy question were simply whether to move the population toward 10, 15, or 20 µg per day. The actual decision is more layered.

A government first needs to identify the public health problem: widespread low dietary intake, seasonal deficiency, a high-risk subgroup, or an inequitable distribution of vitamin D status. It then needs to decide whether fortification is the appropriate instrument, whether voluntary measures are sufficient, and which food vehicles can deliver vitamin D consistently without making the intervention dependent on purchasing patterns that exclude the most vulnerable households.

The chosen reference framework affects each stage.

Setting the intervention target

If a policy uses the IOM framework, the EAR can inform population-level estimates while the RDA may serve as a communication or planning benchmark. If it uses EFSA’s framework, the AI can provide a practical intake reference, but policymakers should not describe it as an intake demonstrated to cover 97.5% of the population.

The intervention target may also be lower than the reference intake if fortified foods are intended to supplement, rather than replace, vitamin D obtained from ordinary diet and sunlight. Alternatively, a programme may need to provide a larger contribution during seasons or life stages when endogenous synthesis is limited. Those decisions require modelling rather than simple adoption of a single headline number.

Protecting against excessive exposure

Fortification policy has two sides: closing an inadequacy gap and avoiding unnecessary excess. The UL of 100 µg/day, or 4,000 IU/day, for adults in both the EFSA and IOM frameworks provides a safety boundary for chronic intake under the relevant assumptions. It is not a target, and it should not be treated as evidence that intake near the UL is desirable.

Total exposure matters. A person may receive vitamin D from fortified milk or plant-based alternatives, breakfast cereals, spreads, supplements, and medical products. If each product is regulated separately without an integrated view of the food supply, the combined exposure may be underestimated.

This is where food safety regulation and nutrition labelling laws need to operate together. Fortification standards should be accompanied by clear declarations of vitamin D content, consistent units, and monitoring systems capable of identifying changes in product formulation. The policy should also specify how compliance will be measured and how the programme will respond if intake or serum status shifts unexpectedly.

Reaching people who are usually missed

Average population intake is a poor proxy for equitable impact. Fortified foods reach people through existing diets, and existing diets are structured by income, geography, culture, disability, housing, retail access, and trust in public institutions.

Grassroots implementation can expose gaps that national modelling will miss. Community health workers, clinicians, local food organisations, and public health teams can identify whether the selected food vehicle is actually consumed by the groups the intervention is meant to serve. They can also clarify whether a label is understandable, whether the product is available in smaller shops, and whether the fortified option costs more than its non-fortified equivalent.

A national fortification strategy that improves intake among already well-served households while leaving food-insecure communities behind may still improve the average, but it will not fully address nutritional inequity. The policy should therefore include distributional monitoring, not only national mean intake or overall serum concentration.

Keeping the science visible in public communication

Public communication should avoid presenting the EAR, PRI, RDA, AI, and UL as five competing daily prescriptions. That language encourages confusion and can lead people to interpret a population reference as a personal diagnosis.

A more accurate explanation is that:

• the EAR describes the estimated requirement of the average person in a defined group;
• the PRI or RDA is set high enough to cover nearly all healthy people in that group;
• the AI is used when evidence does not allow a formal requirement distribution to be established;
• the UL is a safety boundary, not an amount to aim for.

This framing is not merely educational. It protects trust. When communities learn that scientific bodies use different thresholds and methods, they deserve an explanation of why, rather than a succession of apparently contradictory numbers.

From national standards to a shared policy language

The international discussion around vitamin D reference values would benefit from more explicit reporting of assumptions. A standard should state not only the final intake value but also the serum target, sunlight conditions, age group, evidence limitations, and intended use.

This would make it easier to compare dietary guidelines, health claims, food labels, and fortification rules across jurisdictions. It would also help researchers avoid a recurring error: treating a value designed for individual guidance as though it were a direct measure of population inadequacy.

The unresolved questions are substantive. It remains uncertain whether future individual participant-level data meta-analyses will allow international bodies to move from an AI to formal AR and PRI values for vitamin D. The distribution of dietary requirements may also differ across populations with different seasonal sunlight exposure and skin pigmentation, and those differences cannot be resolved by changing the unit from micrograms to international units.

For now, caution does not require inaction. It requires policies that are explicit about what is known and flexible enough to be updated. Governments can monitor dietary intake, serum 25(OH)D, supplement use, product compliance, and subgroup outcomes together. They can review whether the selected food vehicles reach low-income and geographically isolated communities. They can assess whether the intervention reduces seasonal vulnerability without pushing total intake toward the UL.

The practical route is therefore neither to choose the highest reference value by default nor to treat methodological uncertainty as a reason to delay every intervention. It is to connect reference standards with surveillance and community implementation, so that the policy can learn from the population it is intended to protect.

A more honest way to compare vitamin D standards

The question of EAR vs PRI for vitamin D is ultimately a question about purpose.

The IOM framework offers a formal sequence from EAR to RDA, with the EAR aimed at the median requirement and the RDA intended to cover nearly all healthy individuals. EFSA, facing variability in the evidence, established an AI of 15 µg/day for people aged 1 year and older rather than assigning formal AR and PRI values. The UK SACN reference of 10 µg/day rests on a different serum 25(OH)D threshold from EFSA’s, further demonstrating that reference values cannot be separated from the biological targets used to derive them.

For nutrition policy, the next step is clear: stop comparing the numbers without comparing the methods. Every fortification decision should identify the framework behind its target, distinguish adequacy from safety, account for total dietary exposure, and test whether the intervention reaches communities affected by systemic barriers.

That is how reference values become useful public health tools rather than isolated figures in a guideline table. We should preserve the uncertainty where the evidence is uncertain, strengthen surveillance where the population burden is real, and adjust implementation until the benefits of vitamin D policy reach beyond the national average to the people who have been least well served by the food system.