Preventing Overtraining Before It Happens: How AI Spots What You Miss
Anton Krupicka has won the Leadville 100 and the Lavaredo Ultra Trail. He knows what hard training feels like. So when he started noticing an “inability to hit times for intervals or uphill tempos that would normally come easily to me,” paired with what he described as a “general lack of motivation,” something was clearly wrong. The confusing part? He couldn’t tell if his motivation was tanking because he always felt terrible running—or if he felt terrible because his motivation had already left. (Uphill Athlete)
“My own well was dry, and when I looked to find a little more in myself, there was nothing.”
Here’s the cruel irony of overtraining: it targets the hardest-working runners. The ones who never miss a session. The ones who push through fatigue because that’s what got them here. In ultrarunning and trail running, where terrain stress compounds in ways flat-road training never does, that drive becomes a liability faster than most athletes realize. A 15-mile run with 4,000 feet of vertical gain puts a fundamentally different load on your body than 15 flat miles at sea level (Footpath)—and most training logs treat them identically.
The question isn’t whether you’ll flirt with overtraining. If you’re serious about this sport, you will. The question is whether you’ll catch it before it catches you.
What Overtraining Actually Is (And Isn’t)
There’s an important distinction that gets lost in most training conversations. Overreaching and Overtraining Syndrome are not the same thing.
Functional overreaching is a normal part of training. You push hard for a block, you accumulate fatigue, you recover, and you come back stronger. That’s how adaptation works. Non-functional overreaching is when the fatigue outlasts the training block—days or weeks of suppressed performance with no fitness gains to show for it, even after rest. And then there’s Overtraining Syndrome (OTS), the deep end of the pool: a systemic breakdown that can take weeks to months to resolve, sometimes longer.
The numbers are sobering. Research published in Current Sports Medicine Reports found that roughly 60% of elite male and female runners experience non-functional overreaching at some point in their careers. A separate multicountry survey found that about 30% of elite adolescent athletes had experienced it at least once, averaging two episodes lasting four weeks each. (Kreher & Schwartz, 2012 — PMC)
The most dangerous pattern isn’t the overtraining itself—it’s the response. When performance drops, the instinct is to train harder. You felt so good last week; it can’t possibly be overtraining. So you add volume. You add intensity. And you spiral deeper. (Marathon Handbook)
“Taking two weeks off completely could save a year of running in your future. OTS is serious and I feel like ultrarunners older than me have completely fallen off their game due to overtraining. It’s so sad, and frankly, so preventable.”
The training axiom worth memorizing: it is far better to be slightly undertrained and well-rested than even a little overtrained. An undertrained runner can dig into willpower reserves on race day. An overtrained runner has no reserves left to find.
The Warning Signs Most Runners Miss
The textbook signs of overtraining are well-documented. Performance declines despite consistent training. Persistent fatigue that doesn’t resolve with a rest day or two. Loss of motivation—the run you used to look forward to starts feeling like a chore. These are the primary indicators, and most serious runners have encountered at least one of them. (Fleet Feet)
But the secondary signs are where it gets tricky: elevated resting heart rate upon waking, sleep disturbances (falling asleep fine but waking repeatedly), increased frequency of minor injuries, mood changes, and getting sick more often than usual. A 2018 study confirmed that sleep disruptions can be a key symptom of overreaching, sometimes caused directly by increased training load. (Fleet Feet)
For trail and ultra runners, there’s an additional layer of complexity that generic overtraining articles completely miss. Terrain multiplies training stress in ways that mileage alone doesn’t capture. Downhill running generates significant eccentric muscle damage—the kind of deep muscular stress that accumulates silently and takes longer to recover from than the cardiovascular fatigue you can feel. Fifteen miles on a mountain with serious vertical is not fifteen road miles, and your body knows it even when your training log doesn’t. As Alyssa Clark, coach and professional trail runner, puts it: “A 10-mile route one place can be significantly different from another with differences in altitude, ascent and descent and terrain.” (Footpath)
Elevation compounds this further. Training at altitude increases physiological stress beyond what pace and distance suggest. Runners who stack races across different distances—a common pattern in the trail community, where you might run a 50K one month and a mountainous half marathon the next—are loading terrain stress on top of incomplete recovery windows.
One of the most important findings in recent overtraining research comes from a 2016 review in Sports Medicine: subjective measures like mood, perceived fatigue, and soreness ratings are often more sensitive and consistent indicators of developing overtraining than objective metrics alone. Heart rate changes, while useful, can be small enough to get lost in daily noise. (Running Explained)
The practical takeaway: your feelings are valid data. One of the earliest clinical signs of overreaching is an increased rating of perceived exertion for a workload that used to feel routine. (Kreher & Schwartz, 2012 — PMC) If your regular Tuesday trail run suddenly feels like a Thursday long run effort, pay attention. That signal is worth more than most runners give it credit for.
ACWR: The Metric That Changes How You Think About Load
The 10% rule—don’t increase weekly mileage by more than 10%—has been the default advice for decades. It’s simple, easy to remember, and better than nothing. But it’s also blunt. It treats every week in isolation and ignores the cumulative context of what your body has actually adapted to over the past month.
The Acute-to-Chronic Workload Ratio (ACWR) offers a more complete picture. The concept is straightforward: divide your training load from the past 7 days (your acute workload) by the average weekly load from the past 28 days (your chronic workload). The resulting ratio tells you whether you’re ramping too fast, maintaining, or detraining. (Runner’s World)
Here’s what the numbers mean in practice. An ACWR between 0.8 and 1.3 is generally considered the sweet spot—the zone where injury risk is lowest. You’re building load at a rate your body can absorb. Push above 1.5 and you’re in the danger zone: a training spike that significantly increases your risk of injury. Drop below 0.8 and you’re undertrained, which creates its own vulnerability—when you eventually ramp back up, your body isn’t prepared for the load. (Gabbett, 2016 — British Journal of Sports Medicine)
For example: if you averaged 40 miles per week over the past four weeks and then ran 52 miles this week, your ACWR is 1.3—the upper edge of the safe zone. If you jumped to 60 miles, your ACWR hits 1.5, and the research says your injury risk climbs substantially.
The International Olympic Committee recognized the value of this approach in their 2016 consensus statement, recommending ACWR as a framework for managing athlete training loads. (Soligard et al., 2016 — BJSM) It’s not predictive in the way that a blood test diagnoses an illness—no single metric can do that for something as complex as injury risk. But as a framework for making smarter load decisions, it’s significantly more useful than watching weekly mileage in isolation.
Here’s where it gets especially relevant for trail runners: standard pace-based load calculations—including Grade-Adjusted Pace (GAP) and Normalized Graded Pace (NGP)—tend to underestimate the stress of steep downhill running. They account for the cardiovascular demand of climbing but underrepresent the eccentric muscular damage, coordination demands, and musculoskeletal impact of sustained descending. A run with 3,000 feet of descent generates a very different recovery timeline than the same distance on flat ground, even if the GAP looks moderate. This means trail runners relying solely on pace-normalized metrics may be accumulating more load than their ACWR suggests. It’s a blind spot that generic training apps aren’t designed to address.
How AI Changes the Overtraining Equation
You can calculate ACWR by hand. Plenty of runners do, in spreadsheets or training logs. The math isn’t complicated. What’s harder—much harder—is consistently cross-referencing that number against weeks of training context, wellness trends, sleep quality, race proximity, and historical patterns to make a sound judgment call about today’s workout.
That’s where AI coaching shifts from novelty to genuine utility.
An AI coach can process four, eight, twelve weeks of your training data in context and surface patterns that are nearly invisible in day-to-day training. Your easy pace is drifting upward by 10 seconds per mile over three weeks—a sign of cardiac drift that suggests accumulated fatigue. Your heart rate at a given effort has been creeping higher. Your current mileage ramp matches the loading pattern that preceded a previous injury.
These are the gradual drifts that humans miss, not because we’re careless, but because we’re inside the training. It’s the same reason coaches exist: an outside perspective sees what the athlete can’t.
Hannah, the AI coach built into our app at Prato Studios, operates on what we call a protective coaching philosophy. The best analogy we’ve found: she’s like a loyal dog who won’t let you run into traffic—even if you want to. When your ACWR climbs to 1.35 and you have a key quality session in three days, Hannah will recommend backing off a tempo run or shortening it rather than risking the more important workout. When your training data shows a pattern that historically preceded a breakdown, she’ll flag it before you feel it.
The advantage isn’t that AI knows more about exercise science than a human coach. It’s that AI can hold weeks of data in active context simultaneously, never forgets a pattern, and doesn’t get caught up in the athlete’s enthusiasm to push harder. It can cross-reference your training load against wellness inputs, sleep quality, and how close you are to your next race—a holistic view that’s difficult to maintain manually across an entire training cycle.
Your Practical Prevention Toolkit
Technology is a safety net. It’s not a substitute for the fundamentals. Here’s what actually keeps trail and ultra runners healthy over the long term.
- Respect the 80/20 principle. The American Trail Running Association notes that most elite athletes run about 80% of their training at easy efforts, with only 20% at harder intensities. (Footpath) This ratio isn’t just tradition—it’s the distribution that allows sustained high-volume training without chronic fatigue accumulation. If you’re running hard more than twice a week, you’re likely eating into your recovery capacity.
- Build in recovery weeks. Every three to four weeks, reduce volume and intensity deliberately. This is especially critical during ultra training blocks, where cumulative terrain stress stacks in ways that flat-road training doesn’t. A recovery week isn’t lost training. It’s when adaptation actually happens. (Marathon Handbook)
- Monitor both sides of the equation. Track objective markers: resting heart rate trends, heart rate variability, pace at a given effort level. But equally, track subjective markers: mood, energy, motivation, perceived effort. Research consistently shows that the subjective side often detects trouble earlier. (Running Explained) A simple daily 1–10 rating of how you feel before a run can be remarkably informative over time.
- Listen to the athletes who’ve been there. Darcy Piceu, one of the most consistent long-distance trail runners in the sport, puts it plainly: “I never log 100-mile weeks, even when training for a 100.” (Uphill Athlete) More isn’t always more. The runners with decade-long careers in ultra aren’t the ones who trained the hardest in any single block—they’re the ones who trained the most consistently across years.
- Use technology as a safety net, not a taskmaster. The best role for data and AI in your training isn’t to drive you harder. It’s to catch the moments when your drive is outpacing your body’s ability to absorb the work.
Let Someone Watch Your Back
Overtraining is preventable. That’s what makes it so frustrating when it happens. The warning signs are usually there—in the data, in how you feel, in the gap between what you planned and what your body can actually deliver. The problem is that most runners aren’t equipped to synthesize all of those signals in real time, especially when they’re deep in a training cycle and the next race is circled on the calendar.
Hannah was designed to be the coach who watches your training load even when you won’t. She tracks your ACWR with color-coded zones—green when you’re in the sweet spot, yellow when you’re approaching a spike, red when it’s time to back off. She activates a protective coaching mode when your data suggests you’re accumulating risk. And she does it by looking at the full picture: recent workouts, historical patterns, wellness inputs, and race timelines.
If you’re a self-coached trail or ultra runner who wants a second set of eyes on your training—one that won’t get swept up in your excitement about the next big effort—try Hannah free for 7 days and see how the conversation changes your training.
Not ready for coaching? Download Altitude, our free training calendar and race planner, and start tracking your training load basics today.
Sources
- Uphill Athlete — “Overtraining: The Elephant in the Ultrarunning Room” — Quotes from Krupicka, Moehl, Gallagher, and Piceu on overtraining experiences and prevention in ultrarunning.
uphillathlete.com/trail-running/overtraining/ - Kreher, J.B. & Schwartz, J.M. (2012) — “Overtraining Syndrome: A Practical Guide” — Current Sports Medicine Reports. Prevalence data (~60% lifetime NFO in elite runners, ~30% in adolescent athletes), clinical signs, and pathophysiology.
ncbi.nlm.nih.gov/pmc/articles/PMC3435910/ - Gabbett, T.J. (2016) — “The Training–Injury Prevention Paradox” — British Journal of Sports Medicine. Introduced the ACWR framework and the 0.8–1.3 sweet spot for injury risk management.
bjsm.bmj.com/content/50/5/273 - Soligard, T. et al. (2016) — “How Much Is Too Much? (Part 1)” — International Olympic Committee consensus statement on load in sport and risk of injury. British Journal of Sports Medicine.
bjsm.bmj.com/content/50/17/1030 - Running Explained — “Overtraining in Runners: Signs, Metrics, Prevention” — 2016 Sports Medicine review on subjective vs. objective monitoring, HRV findings, and the importance of perceived exertion.
runningexplained.com/post/overtraining-in-runners-signs-metrics-prevention - Runner’s World — “Acute-to-Chronic Workload Ratio: What It Is and How to Use It” — ACWR calculation guide with practical examples for recreational runners.
runnersworld.com/training/a63657755/acute-to-chronic-workload-ratio-calculator/ - Footpath — “Overtraining in Running: Signs, Causes, and How to Recover” — Trail-specific overtraining risks, terrain stress analysis, and the 80/20 principle from the American Trail Running Association.
footpathapp.com/blog/overtraining-in-running/ - Fleet Feet — “6 Signs of Overtraining in Runners” — Secondary symptoms (sleep disturbances, elevated RHR, HRV), recovery strategies, and practical monitoring advice.
fleetfeet.com/blog/6-signs-of-overtraining-in-runners-how-to-know-when-you-need-a-break - Marathon Handbook — “Overtraining Syndrome: Warning Signs and Recovery” — The death spiral of training harder when underperforming, 10% rule, recovery protocols.
marathonhandbook.com/overtraining/
Related reading: AI Coach vs. Human Coach: The Real Cost-Benefit Analysis · Your Running Data Deserves Better: Why Privacy-First Architecture Matters