# Respiratory exchange ratio: what the breath can and cannot tell us about fuel use

**Working contribution to RQ-FND-101 and RQ-FND-007.** Original source and calculation checks are pending. These constructed examples explain an inference boundary; they do not estimate a person's metabolism or prescribe breathing, exercise or diet.

## Two ratios at two boundaries

Respiratory exchange ratio (RER) compares net carbon dioxide leaving the lungs with net oxygen entering them over a measurement interval. Metabolic respiratory quotient (RQ), as used in the oxidative example below, compares carbon dioxide generated with oxygen consumed by the specified reactions. A changing store separates these boundaries.

Write all rates in moles per minute, or in volume equivalents at the **same stated gas conditions**. Let \(P\) be metabolic CO₂ generation, \(C\) oxygen consumption, \(E\) net pulmonary CO₂ output, \(U\) net pulmonary oxygen uptake, and \(N\) the corresponding body gas store. With no other gas inputs or outputs:

\[
\frac{dN_{CO_2}}{dt}=P-E,
\qquad
\frac{dN_{O_2}}{dt}=U-C.
\]

Therefore,

\[
RER=\frac{E}{U}
=\frac{P-dN_{CO_2}/dt}{C+dN_{O_2}/dt}.
\]

When both stores are stable and the stated boundaries hold, this reduces to \(P/C\). A steady workload alone does not establish either condition. The signs matter: releasing stored CO₂ raises output relative to current production; replenishing an oxygen store raises uptake relative to current consumption. The equality requires a nonzero denominator. These are original balance derivations, not fitted human response models. [Checked-example record](../../research/packets/rer-rq/2026-09-30/calculations.json).

## What the classical substrate equations assume

Frayn's original methods paper uses complete oxidation of glucose and a specified triglyceride, plus an empirical protein/nitrogen contribution. Its glucose reaction has RQ = 1; its chosen triglyceride has RQ = 55/78. This does not make every carbohydrate or fat chemically identical. Its rounded equations are:

\[
c=4.55\dot V_{CO_2}-3.21\dot V_{O_2}-2.87n,
\]
\[
f=1.67\dot V_{O_2}-1.67\dot V_{CO_2}-1.92n.
\]

Here \(c\) and \(f\) are apparent carbohydrate and fat rates in g/min, gas rates are L/min on the paper's common volume basis, and \(n\) is urinary nitrogen in g/min. Table 1 uses 22.4 L/mol. Setting nitrogen to zero is an additional assumption, not a measurement of zero protein oxidation. [Frayn 1983, p628 Eqs 1–6; p629 Table 1](https://doi.org/10.1152/jappl.1983.55.2.628) [RER-S001].

The paper also explains why substrate interconversion changes interpretation. Under its specified lipogenesis calculation, negative apparent fat oxidation can represent net fat synthesis. Other accumulation and excretion pathways complicate the calculation. This interpretation cannot be transferred to every negative result obtained during an unsteady breathing measurement. The uninspected supplement's detailed pathway and ATP calculations are not adopted here. [Frayn, pp629–632, Influence of Other Metabolic Processes and Discussion](https://doi.org/10.1152/jappl.1983.55.2.628).

## Hold fuel use constant and change only storage

Construct a system completely oxidizing **1.0 mmol/min glucose** and **0.1 mmol/min of the stated triglyceride**. It contains no protein oxidation, substrate interconversion or changing oxygen store. Using the source's rounded molecular masses, its fixed fuel rates are 0.1800 g/min glucose and 0.0861 g/min fat.

The reactions consume 13.8 mmol/min oxygen and generate 11.5 mmol/min CO₂. Using a stipulated 22.4 L/mol for both gases gives 0.30912 L/min and 0.25760 L/min, respectively. The metabolic ratio stays **0.8333** in every row. Volumes are rounded ideal STPD equivalents throughout; no BTPS volume is mixed into the ratio.

| Constructed state | CO₂ store change, L-equivalent/min | CO₂ at lungs, L/min | Mouth RER | Apparent carbohydrate, g/min | Apparent fat, g/min |
|---|---:|---:|---:|---:|---:|
| Stable stores | 0 | 0.25760 | 0.8333 | 0.1798 | 0.0860 |
| CO₂ released from a store | −0.100 | 0.35760 | 1.1568 | 0.6348 | −0.0810 |
| CO₂ added to a store | +0.100 | 0.15760 | 0.5098 | −0.2752 | 0.2530 |

The stable row approximately recovers the known inputs; small differences reflect rounded source coefficients. In the other rows, applying the same substrate formulas to lung exchange gives false oxidation rates even though the stipulated fuels have not changed. The negative fat result in the release row does **not** signify fat synthesis: this constructed system explicitly has none.

The storage changes apply only over a finite interval with sufficient initial store or capacity. Their magnitude and duration are not human measurements, and the example supplies no breathing maneuver that would produce them. A low or high measured RER needs interpretation of physiology, measurement conditions and model assumptions; the ratio alone cannot identify the cause.

## Connect the accounting to actual measurements

The [steady and transient mass-balance chapter](07-steady-and-transient-mass-balance.md) separates a synthetic reservoir from selected human experiments. Those experiments contain estimated production and stores, separate measurement times, and unmeasured model inputs. They do not convert these constructed substrate examples into a validated personal calculator.

For a real substrate inference, establish the measurement boundary and time window, gas-volume conditions, treatment of protein/nitrogen, stability of relevant stores, and whether nonoxidative production, accumulation or excretion is material. Independently verify instrument timing and the source-specific method. Apparent stability in one recorded variable does not prove all these assumptions.

This packet addresses a limited part of the registered question. Feeding states, additional human transition experiments, modern measurement standards, citation/search completion, source-version checks and whole-question appraisal remain open. No meta-analysis, formal GRADE rating, clinical validation or external human review is claimed.
