Every runner knows the feeling. You're six weeks into a training block, fitness is building, and you're starting to believe the goal is within reach. Then something goes wrong. A niggle in your shin that won't quiet down. A tightness in your Achilles that appears on the first step of every run. A knee that feels fine at kilometre two and terrible at kilometre twelve.

The instinct is to blame the body. Bad luck. Weak calves. Flat feet. Tight hips. The internet is full of explanations that locate the problem in your anatomy, and full of solutions that involve foam rollers, stretches, and strengthening exercises.

Some of these help. But most of the time, they're treating the symptom and ignoring the cause. The cause, in the vast majority of cases, is training load.

The research is consistent and unambiguous: training errors, too much load, increased too quickly, without adequate recovery, account for the majority of running injuries. Not biomechanics. Not shoe choice. Not the surface you run on. The way you manage your training load is the single biggest factor in whether you stay healthy or break down.

This is the most important concept in endurance training that most runners never learn properly. And it's the foundation that adaptive AI coaching is built on.

What training load actually means

Training load sounds straightforward, but it's more nuanced than most runners realise. It isn't just how far you ran this week. It's the total physiological stress your body experienced from training, which is a product of volume (how far), intensity (how hard), frequency (how often), duration (how long each session lasted), and how well you can recover in the context of your life. That last part is critical, and it's the one most runners overlook. Recovery doesn't happen in a vacuum. It's shaped by sleep, stress, nutrition, and everything else going on outside of training.

A runner who does 50 kilometres in a week at easy pace has a very different training load from a runner who does 50 kilometres with two threshold sessions and a set of hill sprints. Same volume, vastly different stress. This distinction matters because your body doesn't just respond to distance. It responds to the cumulative demand placed on every system involved: musculoskeletal, cardiovascular, metabolic, and neurological.

There are two useful ways to think about training load. External load is what you did: the kilometres, the pace, the elevation, the number of sessions. Internal load is how your body responded: heart rate, perceived effort, HRV trends, recovery time. A 16-kilometre easy run on fresh legs after a good night's sleep produces a very different internal load from the same run on tired legs after four hours of broken sleep. The external load is identical. The internal load is not.

This is why a plan that prescribes sessions based only on distance and pace is, at best, an approximation. It's managing external load while ignoring internal load entirely.

The acute:chronic workload ratio

If you've spent any time reading about training science, you've probably encountered the acute:chronic workload ratio, or ACWR. It's one of the most discussed metrics in sports science, and it's worth understanding even though it has significant limitations.

The concept is simple. Your acute workload is what you've done recently, typically the last seven days. Your chronic workload is what you've been doing over a longer period, typically the last 28 days. The ratio between them tells you whether your recent training is in line with what your body has been prepared for, or whether you've spiked sharply above it.

An ACWR of around 1.0 means your recent training matches your longer-term average. You're doing what your body is accustomed to. An ACWR significantly above 1.0, say 1.3 or higher, means you've done considerably more this week than your body has been prepared for. This is where injury risk climbs.

The logic is intuitive. If you've been running 40 kilometres a week for a month and suddenly jump to 55, your body hasn't had time to adapt to the increased demand. Tendons, bones, and connective tissue adapt more slowly than cardiovascular fitness. Your heart and lungs might handle the extra load. Your Achilles tendon might not.

There's an important nuance here that the ratio alone doesn't capture: training history creates a kind of structural resilience. If you've been there before, you can often go there again with relatively low risk. A seasoned runner who has comfortably handled 100 kilometres a week in previous training blocks doesn't usually need twelve weeks to rebuild to that volume after a break. The body remembers. The adaptations in tendon, bone, and connective tissue don't disappear overnight. But the athlete climbing from 30 kilometres a week to 60 for the first time is in genuinely new territory, and that journey might need six months to a year to navigate safely. The risk profile is fundamentally different, and the ACWR treats both situations the same way.

The ACWR has also come under serious academic scrutiny in recent years. Critics have pointed out that the time windows (7 and 28 days) are somewhat arbitrary, that the ratio conflates different types of load (volume and intensity), and that the statistical properties of ratios make them unreliable at the extremes. A 2024 systematic review and meta-analysis acknowledged the concept's usefulness while noting significant heterogeneity in how studies define and measure workload.

So should you ignore the ACWR? No. The underlying principle is sound: sudden spikes in training load relative to what you've been doing are dangerous, especially when you're pushing into territory your body hasn't been prepared for. That principle holds up regardless of the specific metric used to measure it. The Garmin-Runsafe Running Health Study, which we discussed in a previous post, confirmed this clearly: when a single run exceeded the longest run of the previous 30 days by more than 10%, injury risk increased by 64%.

The takeaway isn't that the ACWR is the perfect metric. It's that the relationship between what you're doing now and what you've been doing matters enormously, and most runners pay almost no attention to it.

Intensity distribution

Training load isn't just about how much. It's about the distribution of intensity, and this is where frameworks like the 80/20 principle and polarised training come in.

The concept, popularised by researcher Stephen Seiler and based on decades of observation of elite endurance athletes, is this: roughly 80% of your training should be at low intensity (below the aerobic threshold, what most runners would call truly easy pace), and roughly 20% should be at high intensity (above the anaerobic threshold: tempo, threshold, intervals, races). Very little should fall in the middle. This is known as a polarised distribution. A related approach, pyramidal training, allows a slightly larger share of moderate-intensity work and tapers volume as intensity increases. Both have genuine research support, and the best approach depends on the athlete, the event, and the training phase.

The 80/20 framework provides a useful starting point, but it's important not to oversimplify it. Not all intensity is created equal. A VO2max session might only be 12 minutes of hard running, but it can require two or three easy days to recover from. A controlled threshold run, by contrast, could be followed by an easy session the same day. The physiological and psychological cost of intensity varies enormously, and the sustainable dose depends on the individual. It's unlikely you'll be spending your entire 20% at VO2max, for example. The dose needs to be sustainable over months, and it needs to be specific to the event you're training for.

It's also worth considering what a workout takes from you psychologically, especially in the context of your life and work. A session that's physically manageable but mentally draining still costs something, and that cost affects your recovery and your consistency over the long term.

For well-trained athletes, there's another consideration. As aerobic fitness improves and the aerobic threshold climbs, the upper end of zone 2 can approach marathon pace, and the mechanical load at those speeds becomes significant. A runner with a high aerobic threshold may need to spend more time in zone 1 and less in zone 2, because what counts as "easy" for their cardiovascular system is no longer easy for their joints, tendons, and muscles.

The deeper point behind 80/20 isn't that moderate-intensity work is inherently bad. It's that most runners lack intensity discipline. Athletes on prescribed easy days go too hard, turning what should be recovery into additional stress. The result is that they're never truly fresh for the sessions that matter, and the hard days suffer because the easy days weren't easy enough. It's about how the dose is applied. Is 4% of your training at VO2max better or worse than 8% at threshold? That depends on age, background, fibre type, experience, and the specific demands of the event. The higher the intensity, the higher the global stress and the more unpredictable the recovery.

This is also where HRV data becomes genuinely useful. If your easy runs are consistently suppressing your HRV, you're probably running them too hard. The data confirms what the science predicts: easy means easy.

Progressive overload

The third pillar of load management is progressive overload: the principle that training stress needs to increase gradually over time for the body to continue adapting.

This one sounds obvious, and at a high level it is. You can't run the same 30 kilometres a week at the same pace forever and expect to keep improving. The body adapts to the stimulus, the stimulus stops being challenging, and progress plateaus.

But the execution is where runners get into trouble. The conventional wisdom is the "10% rule": don't increase your weekly mileage by more than 10% from one week to the next. In practice, this is too simple a concept to account for intensity (adding a threshold session at the same weekly volume is a significant load increase that the rule doesn't capture), individual variation (a runner with years of training history can absorb larger jumps than a relative beginner), or life context (10% more volume during a stressful work week is very different from 10% more volume during a holiday). It also breaks down mathematically: a runner doing 30 kilometres a week who increases by 10% every week would be running 78 kilometres by week ten. That's not progressive overload. That's a recipe for injury.

Training is better understood as a step-up followed by a stabilisation period. You increase the load, hold it there while your body absorbs and adapts to the new level, and then, when readiness allows, take another step. The principle is one lever at a time: the next logical step. If you're running twice a week, the next step might be three times a week. Once that's stable, perhaps you add strength work. Then you introduce intensity, but gradually: some strides at the end of an easy run, then short accelerations within a session, then short hills, then longer hills, then fartlek, then structured track work. Each step builds on the last. Jumping straight into 20 x 400 metres when you haven't done any speed work in months is not progressive overload. It's a spike.

There's also an honest question that most training advice never asks: if you increase your training volume by 10%, do you actually have the time and capacity to recover from that extra training? An additional hour of running each week means an additional hour of recovery demand, and that recovery has to come from somewhere in your life. If the answer is that something else gives, whether it's sleep, family time, or stress management, the net effect on your training might be negative even though the volume went up.

What progressive overload actually requires is a more sophisticated conversation between ambition and capacity. You need to increase training stress enough to drive adaptation, but not so much that you exceed your body's ability to recover and rebuild. The margin between these two points is narrower than most runners think, and it shifts constantly based on sleep, stress, nutrition, and accumulated fatigue.

This is also where the most common training error lives. Runners tend to increase load in bursts rather than gradual progressions. A good week of training leads to enthusiasm, which leads to a bigger week, which leads to another bigger week, which leads to a breakdown. The overreach doesn't feel like overreach at the time. It feels like momentum. The injury announcement arrives two or three weeks later, by which time the connection between cause and effect is invisible.

Why load management is hard to do yourself

If training load management sounds like something any disciplined runner could do with a spreadsheet and some self-awareness, you're right in theory and wrong in practice.

The theory is straightforward: increase load gradually, keep most of your running easy, avoid sudden spikes, and back off when fatigue accumulates. The practice is complicated by three things.

First, the data is fragmented. Your weekly volume is in one app. Your intensity distribution is in another. Your HRV trend is on your watch. Your sleep data is somewhere else. Your subjective feel is in your head, unrecorded. Synthesising all of this into a coherent picture every day requires effort that most runners, understandably, don't invest.

Second, the decisions are subtle. Is an HRV dip of 8 points enough to modify tomorrow's session? Does the fact that your easy pace has drifted 10 seconds per kilometre slower than last month mean you're fatigued, or just that it's gotten warmer? Should you extend your recovery week by two days or push through? These aren't questions with obvious answers, and getting them wrong in either direction has consequences.

Third, runners are biased. As we discussed in a previous post, every runner alive is biased toward training more, not less. They focus too much on when the event is and where they want to be, rather than honestly assessing where they are right now. Rest feels like regression. Easy days feel wasted. The temptation to turn an easy run into a moderate run, or to add an extra session in a recovery week, is constant and powerful. Self-coaching requires zooming out and taking the emotion or compulsion out of the decision, which is a skill most people don't have, or they are new to the sport and simply do not know what to do next.

Where AI coaching fits

Load management is the foundation of adaptive coaching because it's the thing that matters most and the thing that's hardest to do consistently without help.

An AI coaching system doesn't just track your load after the fact. It manages it prospectively. It knows your acute and chronic workload. It knows your intensity distribution. It knows whether your easy runs are actually easy. It knows your HRV trends, sleep data, and subjective feel. And it uses all of this to make daily decisions about what your training should look like today, given everything that's happened this week, this month, and across your entire training history.

When it detects a load spike building, it intervenes before the spike becomes an injury. When it sees your intensity distribution drifting toward too much moderate work, it adjusts your pacing guidance. When your progressive overload is progressing too aggressively for the recovery signals it's reading, it dials things back, not because a rule says so, but because the data says your body isn't absorbing the current load.

Research on over 11,000 athletes found that this kind of daily load management, powered by incoming physiological data, reduced injury rates by up to 23% and improved workout consistency by roughly 16% compared to static plans. The gains came not from a different training philosophy, but from better daily execution of the principles outlined in this article.

The irony of training load management is that the concepts are simple and the execution is complex. Increase gradually. Keep most of your training easy. Avoid spikes. Recover properly. Everyone knows this. Almost nobody does it consistently, because the daily decisions required to implement it are subtle, continuous, and resistant to the biases that every motivated runner carries.

The best thing you can do for your running isn't a new session type or a faster workout. It's managing the load you're already carrying, intelligently, every single day. That's the foundation everything else is built on.

Zepho is adaptive coaching for serious runners. It's live and available now on the App Store.