One Sodium, Seven Bodies, Eight Numbers: Why the Gap Between 1.5 g and 2.4 g Is Not About the Number

In one line: Line up the official sodium figures and you get 1,500 · 2,000 · 2,300 · 2,400 mg/day. They look like versions of the same metric. They are not the same kind of thing. Some answer "how much is enough" (Adequate Intake); some answer "how much to avoid" (chronic-disease risk reduction); one is explicitly "achievable today, not optimal." The United States withdrew sodium's Tolerable Upper Intake Level in 2019. Before comparing, ask what role the number plays in its own system.

1. Sodium and salt are not the same quantity

This is the precondition for everything below.

So "less than 5 grams of salt" and "less than 2 grams of sodium" are the same recommendation written two ways. National documents mix the two, and units vary (g / mg / mmol). The first error usually happens here.

2. Eight official figures side by side

Lowest 1,500, highest 2,400 — a spread of nearly 60%. Read only the numbers and the conclusion looks like "experts disagree." Read the Type column and it becomes clear that four different categories of thing have been placed side by side.

3. First fault line: "enough" and "don't exceed" are two different numbers

An AI and a CDRR cannot be subtracted, and neither is "stricter." A 1,500 mg AI is not a tighter limit than a 2,300 mg CDRR — they answer different questions.

⚠️ The most common misreading: treating "AI 1,500 mg" as "recommended intake 1,500 mg," then using it to argue that WHO's 2,000 mg is too permissive. That is using one category as if it were the other.

4. Second fault line: the United States withdrew sodium's ceiling

This is the single most instructive change in the set.

• Before 2019: sodium, like most nutrients in the US system, had a Tolerable Upper Intake Level (UL) — a level above which risk rises.
• In 2019: the National Academies replaced sodium's UL with a CDRR. The stated reason: evidence was insufficient to establish a Tolerable Upper Intake Level, while evidence was sufficient that reducing intake lowers cardiovascular and hypertension risk.

The implication is easy to miss: officially, there is no determinable line at which sodium becomes "too much" — only a direction in which lower intake means lower risk.

When national documents appear to disagree about sodium's "limit," part of the reason is not that standards differ in stringency, but that the concept itself no longer exists in the US framework.

5. Third fault line: a target is not an optimum

The Food Safety Authority of Ireland wrote, verbatim:

The target for the adult Irish population is a mean intake of 2.4 g sodium (6 g salt) per day. Whilst this is considered to be an achievable goal for the population at this time, it should not be regarded as an optimal or ideal level of consumption.

In the same document, the individual advice cites an RDA of 1.6 g sodium/day.

One body, one document: population target 2.4 g, individual RDA 1.6 g. The first is what society can reach now; the second is what an individual needs. Reading 2.4 g as "Ireland's sodium standard" misreads the document.

6. Fourth fault line: the populations differ

Different age floors (15 / 16 / 14), different inclusion of pregnancy and lactation, different treatment of kidney impairment — any one of these makes a cross-country comparison invalid before it starts.

7. Why the figures differ: the evidential basis differs

EFSA's statement deserves separate attention: it means that within the EU dietary reference framework, sodium has a "safe and adequate" tier but no "requirement" tier — structurally different from nutrients such as protein that carry an AR and PRI.

Boundaries and scope

1. This piece compares institutional definitions and categories only. It gives no individual intake advice. How a person's sodium intake should change, or interact with hypertension or renal medication, is a clinical matter outside this scope.
2. WHO's 2025 lower-sodium salt substitute advice is conditional, not universal: it explicitly excludes pregnant women, children and people with kidney impairment. It must not be quoted as a general recommendation.
3. Some institutional documents were not read in full; values and definitions come from official publication pages and official summary materials, confidence B (marked per row in the source table).
4. Values of different types must not be subtracted or compared for stringency (AI vs CDRR, population mean target vs individual RDA).
5. Guideline revision cycles differ; the year and version must always accompany the figure. Each row is dated.
6. No commercial data-vendor sources were used. All figures come from public material published by governments and international organisations.

Sources

Sourcing floor: all data here comes from public international organisations, governments and national academic bodies. No commercial data vendors.

Grade key: L1 = government / international organisation / intergovernmental body / customs; L2 = international standards bodies / regulated exchanges / national think tanks / national academic societies; L3 = industry associations / company statements / official-media relays (grey, original source must be named); L4 = commercial data vendors / self-media (not used by this project).

Rows marked "—" are published here for the first time and are not yet in the site's record index.

Data as of 2026-10-11. Individual records can be retrieved by Record ID via the site's data index or the MCP endpoint.