Aero Until It Hurts: When Your Position Starts Taking Watts Back

Most triathletes have made peace with the basic trade-off in their aero position: get lower, drop drag, and go faster for the same effort, at least for a while. That trade-off is well established, and we've covered how to prioritize aero changes and test drag reliably elsewhere. What's less settled — and rarely discussed — is what happens to that same position over the two, three or four hours it actually has to hold up during a real triathlon bike leg.

This article stays narrowly on that question. Not whether to go aero. Not what to buy. Not how to field-test CdA. Just: does the position that felt fast and sustainable on a 20-minute test ride behave the same way at hour three? The honest answer is that the research here is thin, mostly small-sample, and stops well short of triathlon-relevant durations. That's worth knowing before you commit to a position for a 70.3 or Ironman bike leg.

Why This Matters More for a Triathlete Than a Time Trialist

A time trialist rides an aggressive position for 20 to 60 minutes and then stops. A triathlete rides it for two to six hours and then has to run. If a position costs slightly more energy per watt held over even a 12-minute trial, as the evidence below suggests, that cost compounds across a long-course bike leg in a way no time trial has ever tested. And whatever fatigue does to your position, you still have to swing a leg over the bike and run afterward — a link almost none of this research measures directly.

The steady-state version of this trade-off — that lower positions cost more energy per watt at a given speed — isn't new, and this isn't a re-run of that argument. What's new, and thinner, is the actual question this article asks: does the size of that cost change the longer you hold the position, and does anyone's body adapt to it?

What the Evidence Says

The Steady-State Trade-Off, Briefly (So We Don't Re-Litigate It)

Because the fatigue question only makes sense against a baseline, one line of context: in 19 well-trained male time-trial cyclists tested across torso angles of 0°, 8°, 16° and 24°, every maximal physiological variable measured got worse as torso angle got lower, and the submaximal physiological cost of holding a given speed rose as torso angle dropped (Fintelman, Sterling, Hemida & Li, 2015). Modelling from the same group found the optimal torso angle actually decreases as speed increases — above roughly 46 km/h the aerodynamic gain outweighs the power cost, below about 30 km/h a more upright position wins, and a fully horizontal torso is never the right answer at any speed (Fintelman, Sterling, Hemida & Li, 2014). Most age-group triathletes ride below 40 km/h — exactly the zone where that model says the physiological cost is more likely to outweigh the aerodynamic gain.

A related study of 11 trained cyclists across hip angles of 12°, 16°, 20° and 24° found power at the anaerobic threshold was 16-20 W lower at the most aggressive angle (12°) than in the control position — even though the combined aero-and-physiological "economy" was actually highest at 12°, because at that speed the aerodynamic saving outweighed the power cost (Faulkner & Jobling, 2021). That's the settled steady-state picture. The interesting question — the one this article is actually about — is what fatigue does to it.

The Direct Evidence: Position Cost Grows the Longer You Hold It

The single most relevant finding in this whole area comes from seven trained cyclists tested in a triathlon-style time-trial position versus upright. Critical power was lower in the TT position than upright, while W′ (anaerobic work capacity) was unchanged — and, crucially, the percentage power loss caused by the TT position got larger as trial duration increased. In the longest trial tested, 12 minutes, cadence fell and heart rate rose specifically in the TT position (Kordi, Fullerton, Passfield & Parker Simpson, 2019).

Read that carefully: it's one study, seven riders, and the longest duration tested was 12 minutes — a fraction of even a sprint-triathlon bike leg. That critical power (the aerobically sustainable ceiling) dropped while W′ (the finite sprint reserve) didn't is a genuinely useful clue: it points to the aero penalty being an aerobic-sustainability problem, not an anaerobic-capacity one. But "the penalty grows with duration" is an inference drawn from a trend across trials up to 12 minutes, not a finding demonstrated at bike-leg length. Treat it as the best available signal, not a settled fact.

Gross Efficiency Falls in Every Position — Not Just the Aggressive Ones

A separate study put 12 riders through 20-minute self-paced time trials at torso angles of 0°, 12° and 24°. Gross efficiency was lower at 0° (the most aggressive position) than at 24°, and mean power was highest at 24° — differences you'd expect from the steady-state trade-off. But gross efficiency declined across every single time trial, at every position tested, with no measurable difference in how much it declined between positions (Fennell, O'Grady & Hopker, 2020).

That's an important nuance: fatigue erodes efficiency regardless of how aggressive your position is. Riding upright doesn't protect you from the efficiency loss that comes with time in the saddle — it just starts from a different baseline. Separately, in nine non-elite male cyclists riding just over 40 km, power output in an aero position was measurably lower than in road-bike or fully upright positions, confirming the lab-measured aero penalty shows up in a field-realistic ride rather than only on a stationary rig (Jobson, Nevill, George, Jeukendrup & Passfield, 2008).

Elite Riders Show No Metabolic Penalty at an Aggressive Torso Angle — Trained Non-Elites Do

This is the most provocative finding in this area, and it needs careful handling. World-championship top-10 finishers in time trials cluster around a 4-12° torso angle; national-level elites sit at 8-18°. When 10 national elite TT cyclists were tested at 4°, 12° and 20°, there was no difference between positions in overall energy expenditure, delta efficiency or muscle oxygenation — but perceived effort (RPE) was significantly higher at 4° than at each rider's own habitual position (Cubel, Piil & Nybo, 2022).

Put next to the Fintelman and Faulkner studies above — trained but non-elite riders, both finding clear metabolic penalties at lower torso angles — the pattern is suggestive: elite riders may tolerate an aggressive position with little or no metabolic cost, while trained-but-not-elite riders pay a real one. The honest reading is that position tolerance looks trainable and specific, not that elites are simply built differently. But nobody has run the study that would actually prove that — a longitudinal trial that takes non-elite riders, has them train in an aggressive position over months, and re-tests the metabolic cost. Until that trial exists, "you can train your way into a more aggressive position" is a reasonable hypothesis, not a demonstrated fact.

The Upper-Body Cost: Where "Aero Fatigue" Actually Lives

Neck, shoulder and lower-back fatigue in an aero position isn't just discomfort — it has a measurable muscular signature. In nine male recreational cyclists tested in a wind tunnel with EMG on six upper-body muscles across ten forearm angles, drag area and total upper-body muscle activation both tracked forearm angle almost perfectly (R² > 0.9 for both): lower forearm angles cut drag and simultaneously raised muscle activity, with the triceps working hardest and the biceps, anterior deltoid and upper trapezius all rising significantly in the more crouched postures. A forearm flexor muscle also showed a classic fatigue signature — its EMG frequency dropped in the most aggressive position (Li, Zhou & Zhang, 2026).

This was a road-bike drops-and-hoods study, not a TT-bar study, involved recreational rather than trained riders, and only measured acute effort — not what happens after two hours. But it's the clearest evidence available that the aerodynamic gain from a lower position is bought, muscle by muscle, with exactly the upper-body load that riders describe as neck and shoulder fatigue. Every degree of position that saves you watts of drag appears to cost you watts of postural muscle work to hold it there.

Position Changes the Run, Not Just the Bike

One further study is worth including because it's the only one in this area that measures a triathlon-specific outcome. Eight male triathletes rode 40 km at roughly 70% of VO2peak in two different bike setups, then ran a 10 km time trial. A steeper 81° seat-tube angle produced a run time of 42:55 versus 46:15 with a shallower 73° angle — a meaningful difference, concentrated entirely in the first 5 km (Garside & Doran, 2000).

This is a single study from 2000, in eight athletes, with no replication found in the newer literature. Seat-tube angle and torso angle aren't the same variable, and this doesn't directly answer the fatigue-over-time question above — but it's a reminder that position choices on the bike don't stay on the bike.

What to Actually Do With This

Given how much of this is unsettled, the practical guidance has to be proportionate to the evidence, not ahead of it:

• Match your position to your actual race speed, not your fastest possible speed. The Fintelman modelling suggests the aero gain only clearly outweighs the physiological cost above roughly 46 km/h; if your realistic race pace is closer to 30-35 km/h, a slightly less aggressive position may cost you less than it saves.
• Treat "my neck and shoulders always hurt by hour two" as data, not just discomfort to tolerate. The EMG evidence says that pain has a real, escalating muscular cost attached to it — it isn't only about comfort.
• If you want to hold a more aggressive position for longer, build toward it gradually in training rather than assuming race-day adrenaline will carry you. The elite-vs-trained contrast is suggestive that tolerance can be built, but no study has proven a specific way to build it — treat any specific protocol, including this guidance, as reasoned practice rather than validated training.
• Don't assume the position that felt sustainable on a 20-minute field test will feel the same at hour three. Nothing in the literature has tested that duration, so the honest plan is to rehearse your race position on your longest rides, not just your fastest ones.

Common Mistakes

• Choosing a position based on a single fast field test or a fitting-session feel, then never testing it on a genuinely long ride.
• Assuming an upright position is fatigue-proof. Gross efficiency fell across every time trial in the Fennell data regardless of torso angle — riding upright doesn't stop fatigue from eroding efficiency, it just starts from a less aggressive baseline.
• Copying a pro's torso angle without accounting for the difference in adaptation. The Cubel data suggest elite tolerance may be earned over time, not purely anatomical.
• Pushing through numbness, sharp back pain or a stiff neck because "aero is supposed to feel uncomfortable." Discomfort that changes your cadence or forces you to keep sitting up is exactly the signal the Kordi and Li data would predict, and it's a cue to get a professional opinion, not grit through it.

How to Apply This Week

You don't need a wind tunnel to act on any of this.

• On your next long ride (90+ minutes), hold your race position for extended blocks and note honestly when cadence starts to drop or you find yourself sitting up. That's the closest field proxy to the cadence and heart-rate signal Kordi's lab data captured.
• Add one session this week where you hold your intended race position continuously for at least 30-40 minutes, rather than only ever testing it in short bursts.
• If neck, shoulder, lower-back or hand numbness shows up consistently, book a bike fit rather than adjusting stack and reach yourself by feel.
• If you're within 4-6 weeks of a race, don't make aggressive position changes now. Any adaptation this literature suggests is plausible takes months, not weeks, and a new position close to race day trades a small aero uncertainty for a much larger comfort and pacing risk.

References

• Fintelman, Sterling, Hemida & Li, 2015 — Journal of Sports Sciences: torso angle and the aero-physiological trade-off
• Fintelman, Sterling, Hemida & Li, 2014 — Journal of Biomechanics: modelling the optimal torso angle by speed
• Faulkner & Jobling, 2021 — International Journal of Sports Physiology and Performance: hip angle, power and heat production
• Kordi, Fullerton, Passfield & Parker Simpson, 2019 — European Journal of Sport Science: critical power in TT vs upright position
• Fennell, O'Grady & Hopker, 2020 — Journal of Science and Cycling: gross efficiency across torso angles in self-paced TTs
• Jobson, Nevill, George, Jeukendrup & Passfield, 2008 — Journal of Sports Sciences: field validation of aero-position power suppression
• Cubel, Piil & Nybo, 2022 — Journal of Science and Cycling: elite TT cyclists across torso angles
• Li, Zhou & Zhang, 2026 — International Journal of Sports Physiology and Performance: EMG and drag across forearm angles
• Garside & Doran, 2000 — Journal of Sports Sciences: seat-tube angle and the subsequent 10 km run