Beyond FTP: Capacity, Sustainability and Durability
Why the best endurance profile isn’t one number — and what recent research changes about how I test and train athletes
Most endurance athletes know their FTP, threshold pace or VO₂max.
These are useful numbers. I use them regularly in my own training and with the athletes I coach.
But they answer only one question:
How good are you right now, when relatively fresh?
Recent research is increasingly showing why that is not enough.
Two athletes can possess a similar threshold yet sustain it for very different lengths of time. Two athletes can complete the same hard session yet incur quite different types of fatigue. And during prolonged exercise, performance can deteriorate even while conventional measures such as running economy remain relatively stable.
That has reinforced a framework I’m increasingly using within the Sparks Into Life performance and durability work:
Capacity → Sustainability → Durability
They are related, but they are not the same thing.
And understanding the difference gives us a much better basis for deciding what an athlete actually needs to train.
1. Capacity: what can you produce?
This is the part of performance profiling most athletes already understand.
We might assess:
VO₂max or maximal aerobic power/pace
critical power or critical speed
lactate or ventilatory thresholds
power-duration characteristics
sprint and glycolytic characteristics
These establish the athlete’s current physiological ceiling and important intensity boundaries.
They matter.
But a threshold of 320 W does not tell me how long an athlete can actually sustain 320 W.
That distinction sounds obvious, but training prescription often behaves as though the two things are interchangeable.
They aren't.
2. Sustainability: how long can you actually hold it?
A particularly interesting 2026 study from Pallarés and colleagues looked at time to exhaustion at three commonly used physiological landmarks in trained runners:
VO₂max velocity
respiratory compensation point
maximal lactate steady state
Average time to exhaustion was approximately:
VO₂max: 3:38 RCP: 10:58 MLSS: 56:42
So far, nothing particularly surprising.
The interesting part was the variation between athletes.
Time to exhaustion at MLSS varied by around 23% between individuals. RCP showed even greater between-athlete variation.
Yet within each athlete, these durations were remarkably repeatable.
Perhaps most importantly, the athletes who had the highest threshold speeds were not necessarily those who could sustain their threshold for longest. [Pal26]
That has an obvious coaching implication.
Two runners might both have an MLSS pace of 4:00/km.
One might be capable of sustaining it for 45 minutes.
Another might manage more than an hour.
Those athletes don't have the same performance profile simply because their threshold number matches.
This changes how I think about
threshold
Rather than:
Threshold = X watts / X pace
I increasingly want to know:
Threshold magnitude + sustainable duration
The same could apply to cycling.
If two riders both have a CP around 320 W, knowing that number alone does not tell me whether their sustainable characteristics are equivalent.
This is why I think time-domain capability deserves a bigger role in performance profiling.
SAME THRESHOLD. DIFFERENT ATHLETE.
3. Durability: what remains after the work has accumulated?
This is the next layer.
An athlete can have:
excellent capacity;
excellent fresh-state sustainability;
and still deteriorate substantially after several hours.
That is durability.
This is where endurance profiling becomes particularly interesting for marathoners, ultra runners, long-course triathletes and cyclists preparing for long or multi-day events.
A recent trail-running study provides a good example.
Jaén-Carrillo and colleagues had 23 trained trail runners — including 11 women — complete three hours of controlled running with repeated 12-minute uphill time trials.
Over the session, uphill performance declined by approximately 6.6%.
Stride length shortened.
Ground-contact time increased.
Heart rate and RPE increased.
Carbohydrate oxidation progressively decreased while fat oxidation increased.
Yet something interesting happened:
oxygen cost and overall running economy remained relatively stable. [Jae26b]
That matters.
Because if we only measured economy, we might conclude the athlete was coping remarkably well.
But their ability to actually produce performance had deteriorated.
A stable metric doesn't necessarily mean a durable athlete
This is why I am reluctant to define durability using one measure such as:
HR drift;
efficiency factor;
economy;
or fatigued threshold alone.
Instead I want to understand the pattern of deterioration.
For example:
Power/pace ↓RPE ↑HR ↑Economy ↔Carbohydrate oxidation ↓
That is far more informative than reducing everything to a single score.
PERFORMANCE FELL. ECONOMY DIDN'T.
Illustrative summary based on Jaén-Carrillo et al. 2026.
4. Capacity → Sustainability → Durability
Put the three together and the performance profile becomes much more useful.
CAPACITY
What can you produce?
VO₂maxThresholdCP/CSPower/pace
↓
SUSTAINABILITY
How long can you sustain it?
TTEFractional utilizationLong intervalsEvent-specific duration
↓
DURABILITY
How much remains after accumulated work?
Performance retentionHR/RPE driftMechanical changeMetabolic change
That is increasingly how I see the Sparks Into Life performance profile developing.
Not another attempt to find one magic metric.
A framework that tells us where the athlete loses performance.
5. The session matters — not just the intensity zone
Another useful recent paper examined what happens after different forms of high-intensity running.
Dutra and colleagues compared three sessions covering almost exactly the same total distance:
5 × 8 min at RCP
10 × 3 min at 130% RCP
20 × 80 sec at 150% RCP
All produced significant fatigue.
Maximal voluntary force was still reduced six hours afterwards and perceived fatigue remained elevated.
But the type of fatigue differed.
The longer RCP intervals generated greater reductions in neural drive.
The short, higher-intensity efforts produced greater impairment in muscular contractile function, alongside greater leg pain and tiredness. [Dut26]
This is practically important.
We often describe sessions simply as:
“HIIT”
But 5 × 8 minutes and 20 × 80 seconds are not physiologically interchangeable just because both are hard.
And their position in the week should reflect that.
If tomorrow contains strength training, sprint work, technical running or an important long session, the recovery cost and nature of the fatigue matter.
That reinforces one of the principles I use constantly:
The best session isn't necessarily the hardest session. It is the session whose stimulus fits what the athlete needs — and what they can absorb.
SAME DISTANCE. DIFFERENT FATIGUE.
Based on Dutra et al. 2026.
6. Polarised versus pyramidal: I think we're asking the wrong question
Training-intensity distribution remains one of endurance sport's favourite debates.
Polarised?
Pyramidal?
Threshold?
Recent Bayesian network meta-analysis again failed to identify a training-intensity distribution that was definitively superior to polarised training for either VO₂max or time-trial performance.
Threshold-oriented training ranked highly for VO₂max and HIT ranked highly for TT performance, but the credible intervals did not support claiming either as universally superior. [Li26]
For me, that is useful rather than disappointing.
It reinforces the idea that training distribution should be an output of good programming, not an ideology we force every athlete into.
An athlete with six hours per week and an athlete with eighteen hours per week should not necessarily have the same distribution.
Nor should:
an ultra cyclist;
a 10-mile time triallist;
a marathon runner;
and a criterium racer.
The questions I would rather ask are:
What capacity are we trying to develop?
How sustainable does it need to become?
How durable does it need to remain?
Once we know that, the appropriate intensity distribution usually becomes much clearer.
7. Fueling: more is not automatically better either
There is a similar trend in endurance nutrition.
Carbohydrate intake has progressively moved upward.
60 g/h became 90.
90 became 120.
Some elite athletes now report considerably more.
There is good reason for the progression: carbohydrate availability clearly matters to endurance performance and trained athletes can often oxidize more carbohydrate than older guidelines implied.
But a recent review by Plews and colleagues makes an important point:
practice has moved faster than the evidence. [Ple26]
There is not yet strong evidence that routinely moving from roughly 90 g/h to extremely high intakes such as 120–150+ g/h automatically improves endurance performance.
Higher intakes can increase exogenous carbohydrate oxidation.
But oxidation efficiency declines as intake rises, and inter-individual differences are substantial.
The authors also discuss intriguing possible mechanisms involving:
substrate selection;
lactate shuttling;
exercise economy;
and central/perceptual effects.
But these remain interesting hypotheses rather than established reasons to prescribe enormous carbohydrate intakes to everybody.
My practical approach remains individualisation
For very long events:
60–90 g/h remains a strong evidence-based range for many athletes.
For some trained athletes, 90–120 g/h may be useful and worth deliberately developing.
Beyond that, I want evidence from the individual athlete:
Can they tolerate it?
Does performance improve?
What happens to GI comfort?
What happens late in the event?
Does the demand of the event actually justify it?
Don't fuel according to Instagram.
Fuel according to the athlete and the event.
MORE CARBOHYDRATE ≠ AUTOMATICALLY MORE PERFORMANCE
Interpretation informed by Plews et al. 2026. Note: I would explicitly avoid making the recommendation that 120 g/h is a target everyone should reach.
What this means for Sparks Into Life testing
Taken together, these papers reinforce the direction I've been moving toward.
I don't want an athlete profile that simply reports:
FTP = 320 W
or:
Threshold pace = 4:00/km
I want to build a more complete picture.
1. Capacity
What can this athlete produce when fresh?
2. Sustainability
How long can they actually maintain the relevant physiological intensity?
3. Durability
What happens to that performance after meaningful accumulated work?
4. Execution
What happened with:
carbohydrate;
hydration;
pacing;
environmental stress;
GI tolerance?
Only then can we start answering the useful coaching question:
Why did performance deteriorate?
What an athlete can implement immediately
You don't need a laboratory to start thinking this way.
Instead of simply recording your next long ride or run as:
4 hours completed. Felt tired at the end.
Record:
Early performance Power / pace / HR / RPE
Late performance Power / pace / HR / RPE
Fuelling Actual carbohydrate intake in g/h
Hydration Actual fluid intake in ml/h
Work accumulated Duration, elevation and — for cyclists — kJ and work above CP where possible
Then ask:
What changed?
And importantly:
Why?
That one question can make your long endurance training considerably more useful.
The bigger message
Endurance performance cannot be summarised by one number.
FTP matters.
VO₂max matters.
Threshold matters.
But so do:
how long the athlete can sustain those qualities
and
how much remains after hours of work.
That is why I'm increasingly thinking about endurance performance in three layers:
CAPACITY → SUSTAINABILITY → DURABILITY
Build the capacity.
Extend how long it is usable.
Then teach the athlete to preserve it when fatigue accumulates.
For long-distance and multi-day endurance athletes, that final part may ultimately be the difference that matters most.
Suggested final CTA
Want to know what happens to your performance once freshness disappears?
The Sparks Into Life profiling approach combines fresh performance testing with durability, fueling and hydration assessment to identify what is actually limiting your long-event performance.
Get in touch or book a consultation to discuss your target event.
Research referenced
Li H, Yang Q, Wang B. Effects of Different Training-Intensity Distribution Models on Maximal Oxygen Uptake and Time-Trial Performance in Endurance Athletes: A Bayesian Network Meta-Analysis. [Li26]
Pallarés JG et al. Time to Exhaustion at Traditional Physiological Indicators in Runners: Between-Subject and Between-Day Variability. [Pal26]
Dutra YM et al. Neuromuscular Fatigue and Perceived Fatigability in the Hours Following a High-Intensity Endurance Running Depend on the Exercise Protocol. [Dut26]
Jaén-Carrillo D et al. Durability in Trail Running: Coupled Physiological and Biomechanical Responses to Prolonged Submaximal and Repeated Uphill Time Trials in Trained Trail Runners. [Jae26b]
Plews DJ et al. Fuelled or Fooled? Examining the Evidence and Mechanisms Behind Ultra-High Carbohydrate Intake in Endurance Athletes. [Ple26]
Comments