If you’re over 40 and you train seriously, here’s what probably happens every few weeks.
Your Garmin tells you your status is Productive. Your Whoop recovery has been green. Your Training Readiness score has been above 70 for six days running. You feel okay. You do a session that looks textbook on paper. Nothing ambitious, nothing out of pattern.
And you’re wrecked for three days. A session that would have been Tuesday fine at 32 is Tuesday destroyed at 48. The recovery that the models predicted does not arrive. Your legs carry the fatigue into Thursday. You miss Friday’s session because Thursday became the actual recovery day you’d planned for Wednesday.
This is not you getting old and not training hard enough. This is a systematic mismatch between the training load models embedded in every wearable and the physiology of anyone over about 40.
Where the Models Come From
The training load frameworks used by Garmin, Whoop, Training Peaks, and every other major platform trace back to a small set of sports science models developed in the 1970s through the 2000s.
TRIMP, TSS, Banister’s fitness fatigue model, the Acute Chronic Workload Ratio. All of them are built on research populations that were overwhelmingly young athletes. College age endurance athletes. Military cohorts. Professional cyclists in their mid 20s. The datasets used to calibrate the mathematical relationships between training stress and recovery were not meaningfully inclusive of masters athletes.
This is not a criticism of the original researchers. They studied what was available. The issue is that the resulting frameworks were then generalised into consumer products and sold as universal training load models, with the implicit assumption that recovery dynamics in a 55 year old mirror those of a 22 year old with scaled magnitudes.
They don’t.
The Three Physiological Differences That Break the Models
Masters athletes have consistent, measurable differences in how they respond to training stress. The differences are well documented in sports science literature over the last 20 years. They’re just not reflected in wearable algorithms.
Recovery time after muscle damaging efforts extends significantly with age. A high volume strength session or an eccentric loading workout that a 25 year old recovers from in 48 hours might take 72 to 96 hours for a 50 year old. The magnitude of damage is similar. The clearance and adaptation time is not.
This is important because wearable training load models assume a roughly fixed recovery half life. They decay the training stress score over a standard curve that doesn’t account for age-specific recovery kinetics. By the model’s logic, your Wednesday should be fresh because Tuesday’s stress has cleared. By your legs’ logic, it hasn’t.
The autonomic nervous system response to intensity differs. At masters ages, HRV depression following high intensity efforts tends to be deeper and last longer. A 4 by 4 minute interval session might suppress HRV for 24 hours in a younger athlete and 48 to 72 hours in a masters athlete, even when both complete the same session at proportional intensity.
The wearable reads the HRV at 48 hours, sees it approaching baseline, and calls you recovered. Your actual nervous system is still working through the load.
Concurrent training interference is amplified with age. The interaction between strength work and endurance work is not a constant across age groups. Masters athletes appear to experience greater interference between close-together sessions. Strength work taxing endurance adaptation more, endurance work compromising strength recovery more.
Every training load model treats these sessions as additive independent stresses. For a masters athlete, the interaction term is non trivial and the models capture none of it.
What This Looks Like in the Data
Take a specific example. A 50 year old triathlete running a 70 kilometre per week base build with two sessions of strength per week.
On paper, this is moderate. TSS ratio in the 0.9 to 1.1 range. CTL rising on a normal 5 per week ramp. No acute chronic workload flags. Training Readiness scores mostly green. Whoop recovery in the 60s and 70s.
In reality, there’s a recurring pattern. Every third week, the athlete hits a wall. Session quality drops. Sleep quality degrades. HRV sags for 4 to 5 days. They take a forced mini-deload. A week later they’re back on pattern.
The models don’t see this coming because they don’t know that a 50 year old accumulating the same load as a 25 year old needs a more frequent light week, not a once-per-month planned deload.
The athlete’s interpretation is usually wrong. “I’m just not recovering well this week.” Or “Something must have been off.” No. The pattern is structural. The framework being used to plan the training doesn’t fit the physiology doing the training.
The One-Size-Fits-All Recovery Score Problem
Here’s the trap. Wearable recovery scores are normalised to your own baseline. Whoop tells you recovery is 72 percent. That 72 is comparing today’s HRV, RHR, and sleep to your rolling personal baseline, not to some universal standard.
Sounds good in theory. It means the number adjusts to your physiology.
In practice, it doesn’t adjust enough. Here’s why.
The rolling baseline is short. Usually 7 to 30 days. It captures your current state but not your recovery kinetics. If you’re a 55 year old who genuinely needs 72 hours between hard sessions, a recovery score of 70 percent at 48 hours out still tells you you’re ready, because the score is calibrated to how you look relative to your recent average, not to how long your tissues actually need to recover.
You can be at your personal 70 percent and still be in the middle of a recovery curve that isn’t done. The score doesn’t know that.
What Actually Works for Masters Training Load Management
Everything I’m about to suggest is from personal tracking, conversations with masters athletes I know, and the small amount of sports science research that has specifically studied masters populations. Not universal truth. Working heuristics.
Increase your recovery multiplier. Whatever the wearable tells you about recovery time, add 50 percent for muscle damaging sessions, 25 percent for high intensity aerobic sessions. If the device says 24 hours, plan on 36. If it says 48, plan on 72.
Use a longer moving average for acute load. Most platforms use 7 days for acute load. Masters athletes often get better signal from 10 or 14 days. Slower to react, fewer false alarms, less whipsawing in recommendations.
Plan a light week every third week, not every fourth. The 3:1 structure (3 loading weeks, 1 deload) is standard in coaching. For masters, 2:1 or 3:1 with a lighter loading progression often works better. Less aggressive builds, more frequent resets.
Track subjective readiness separately. Morning self rating on a 1 to 10 scale captures fatigue signals that HRV and RHR miss. For masters athletes, the subjective score often leads the objective data by 24 to 48 hours. If you feel off for two days in a row, something is happening that the device will show you later.
Respect muscle soreness as a signal. Young athlete models tend to discount DOMS as an unreliable signal. For masters athletes, deep muscle soreness at 48 hours is a more meaningful indicator of actual muscular recovery status than the nervous system recovery the wearable is measuring.
What Good Would Look Like
The training load models we have today are calibrated to a population you don’t belong to. The data they produce is not wrong. The inferences the platforms draw from it often are.
An intelligent masters athlete framework would take into account age adjusted recovery kinetics, longer baseline windows, explicit modeling of concurrent training interference, and a higher weighting of subjective and muscular signals relative to autonomic ones.
That framework does not exist in any consumer wearable today.
Until it does, the best thing a masters athlete can do is stop trusting the default interpretation of the numbers and build a personal understanding of how your specific physiology responds to specific kinds of load. Track patterns. Note exceptions. Learn your actual recovery kinetics, not the ones the app assumes.
The data is useful. The default interpretation is calibrated for someone else.
Green score. Destroyed legs. There are 6 blind spots in your wearable data. We wrote a free guide covering every one of them.
Green score. Destroyed legs. There are 6 blind spots in your wearable data. We wrote a free guide covering every one of them.
Download the Free Guide