She did everything the apps told her to. Ate to maintenance — the number the calculator spat out, the one that keeps the scale flat. Trained twelve, fourteen hours a week through a big winter block. Weight steady, steps fine, discipline immaculate. And by February her threshold power had gone backwards, she was catching every cold in the office, her sleep had turned to glass, and her morning resting heart rate had quietly climbed five beats. Nothing in her food diary looked wrong. The plate was sensible. The math underneath it wasn't.

This is the silent mistake in endurance fuelling, and it has a name.

Section 01"Maintenance calories" and "enough fuel" are not the same number

Energy availability (EA) is the energy left for everything your body has to do — repair, hormones, immunity, bone, brain — after the cost of training is taken out, scaled to your fat-free mass. The formula is plain: EA = (energy intake − exercise energy expenditure) ÷ fat-free mass, in kcal per kg of fat-free mass per day (Loucks 2011).

Read that again, because the trap lives in one word: after. A maintenance-calorie target only asks whether total intake matches total expenditure — does the scale stay flat. It never subtracts the ride first. EA does. Those are completely different questions.

Run the numbers on our winter rider. Seventy kilograms, roughly 55 kg of fat-free mass, eating 2,500 kcal a day. On a day she burns 900 kcal on the bike, her EA is (2,500 − 900) ÷ 55 ≈ 29 kcal·kg FFM⁻¹·day⁻¹. Her weight is stable. Her energy availability is under the line. The scale says "fine"; the physiology says "I'm running on reserves."

The thresholds matter. Around 45 kcal·kg FFM⁻¹·day⁻¹ is the value associated with energy balance and optimal function; chronic intake below roughly 30 is where measurable physiological disruption begins (Loucks 2011; Thomas 2016). The band from 30 to 45 is reduced, sub-optimal availability — not catastrophe, but not free either.

Section 02What low availability actually breaks

When fuel runs short for long enough, the body doesn't fail loudly. It triages. It down-regulates the systems it can survive without in the short term to protect the ones it can't, and that conservation state is what the IOC consensus calls Relative Energy Deficiency in Sport — RED-S (Mountjoy 2018).

The list of what gets throttled is long and specific: resting metabolic rate drops, reproductive hormones fall (menstrual disruption in women, suppressed testosterone in men), bone turnover and density suffer, immunity weakens, the capacity to build and repair tissue through protein synthesis is blunted, and mood and cardiovascular regulation take a hit (Mountjoy 2018). Low energy availability is the root cause; RED-S is the syndrome that grows from it.

The performance side reads like a list of every plateau you've cursed: reduced endurance, smaller glycogen stores, a worse response to the same training, slower recovery, and more time lost to injury and illness (Mountjoy 2018). The cruel part is the loop — you train harder to fix the stall, which raises expenditure, which drops EA further, which deepens the hole. The fatigue isn't a discipline problem. It's an accounting problem.

Section 03The fix is older than the problem: fuel for the work required

The corrective isn't "eat more" in the abstract. It's eating in proportion to what each day actually demands — what sports nutrition calls fuelling for the work required (Thomas 2016). Carbohydrate is the lever, because it's the fuel that scales with training load while protein and fat stay relatively fixed.

The position stand sets carbohydrate by the day, in grams per kilogram of body mass (Thomas 2016):

• Light day (recovery, skills, easy spin): 3–5 g·kg⁻¹
• Moderate day (~1 hour of moderate work): 5–7 g·kg⁻¹
• High day (1–3 hours, real intensity or endurance): 6–10 g·kg⁻¹
• Very high day (the 3–4+ hour epics): 8–12 g·kg⁻¹

For a 72 kg rider that's roughly 250–360 g of carbohydrate on an easy day versus 580–860 g on a big one — a swing of hundreds of grams, driven entirely by what the legs did. A flat daily diet can't honour that. It either over-fuels the rest days or, far more often, starves the hard ones.

Protein holds the lean-mass line: 1.2–2.0 g·kg⁻¹ per day, and it works best spread across three or four meals at about 0.3 g·kg⁻¹ each rather than dumped into one (Thomas 2016). Fat fills the rest — but with a floor. Dropping below about 20% of total energy buys no performance benefit and starts costing you fat-soluble vitamins and essential fatty acids (Thomas 2016).

Put together, the protocol is unglamorous and effective: take the day's training cost, set carbohydrate to the band that matches the work, anchor protein across the day, keep fat above its floor, and let the total land where energy availability stays north of 30 — ideally near 45 on your hardest blocks.

Section 04Run it on yourself for one week

You don't need a lab. You need an honest week.

1. Estimate fat-free mass. Weight × (1 − body-fat fraction). No body-fat number? A waist/neck/hip circumference estimate gets you close enough to start.
2. Log intake honestly for five to seven days. Honestly is the hard word — under-reporting is the default failure mode.
3. Estimate each ride's cost. Kilojoules from a power file are a near 1:1 stand-in for kcal; otherwise use your device's session estimate.
4. Compute EA per day: (intake − ride) ÷ fat-free mass.
5. Read the week, not the day. If you're averaging under ~35 across a hard block — or dipping under 30 on several days — that's your signal. Add carbohydrate to the offending days first.

One low day means nothing. A pattern of them, stacked under heavy training, is the thing to catch.

Section 05Where the number lies to you

Here's the part most calculators won't tell you: the 30 figure is a research cut-point, not a diagnosis (Mountjoy 2018). It came from short-term, tightly controlled laboratory studies in women, focused on reproductive and metabolic markers — not a season of male road racers. Individuals vary enormously in the EA at which problems appear; some systems wobble well above 30, others tolerate dips. For male athletes the consensus is blunt: the critical threshold isn't established, and it may sit lower than 30 (Mountjoy 2018).

A measurement problem sits on top. Every term in the equation is shaky in the field. Self-reported intake is chronically under-reported. Exercise expenditure is an estimate, not a fact. Even fat-free mass carries error. Three noisy inputs make a noisy output — which is exactly why the IOC tells clinicians to read RED-S as a continuum through signs and history, not to diagnose anyone off a single number on a screen (Mountjoy 2018).

So treat your EA number the way you treat a single power figure off a cold trainer: directionally useful, not gospel. One low day is noise. A month of low days, plus the symptoms, is the signal.

Section 06Reading the body, not just the diary

This is where a number-on-a-form gives way to something more honest: what your body does night after night. A one-shot EA calculation is a snapshot of a guess. Chronic under-fuelling, by contrast, leaves fingerprints that accumulate — and those are trackable.

The HELIOS ring isn't measuring energy availability, and it shouldn't pretend to. What it sees is the consequence pattern: resting heart rate drifting upward across weeks, heart-rate variability trending down off its own baseline, sleep fragmenting — all while training load stays high and performance flattens. None of those is meaningful alone. Together, sustained, against your personal baseline, they trace the same energy-conservation state the research describes — and they show up in your data before they show up in a blood test, and earlier than a food diary you've quietly stopped filling in honestly.

That's the job THRIVE's coaching does: not hand you a calorie verdict, but watch the trend and say the quiet thing out loud — your readiness signals have slid for two weeks under this load; this looks like under-fuelling, not under-training. The fix it points to is the one the science supports: add fuel to the days that earned it, hold protein, and let the trend recover before you add more training.

Section 07Bottom line

Hitting maintenance calories tells you your weight is stable. It tells you nothing about whether you're fuelling the work — because the ride is never subtracted. Fuel by the day's demand, keep energy availability off the floor (and near 45 on your big blocks), and judge it by the trend in how your body recovers, not by a single number on a screen.

A note on limits. The 30 kcal·kg FFM⁻¹·day⁻¹ threshold is a research reference point derived largely from short-term studies in women; the male threshold is not well defined and may be lower (Mountjoy 2018). Field measurement of energy availability is imprecise — intake is under-reported, exercise expenditure is estimated, and fat-free mass carries error — so a single calculated value should never be read as a diagnosis. Energy availability and RED-S are best assessed as a continuum through trends, symptoms and history. This is general training information, not medical advice; persistent fatigue, menstrual changes, or repeated illness or injury warrant a qualified sports physician or dietitian.