Here is a message I get some version of every September. "My long run pace was identical start to finish, but my heart rate went from 142 to 161. Am I losing fitness, or am I just dehydrated?" It is a fair question and the honest answer is neither, probably.
The rise has a name. Cardiac drift, or cardiovascular drift, is the slow upward creep in heart rate paired with a slow decline in stroke volume during prolonged steady exercise. It shows up in trained runners and untrained ones. It shows up when you are perfectly hydrated. It is not a symptom of anything going wrong.
That does not make it useless. It just means the number is only readable once you know what is producing it.
What is actually happening in your chest
Your heart has two ways to move more blood: beat harder or beat more often. Stroke volume is the first, heart rate the second, and cardiac output is the product of the two.
Somewhere between ten and twenty minutes into a steady effort, stroke volume begins to fall. Your heart rate climbs to cover the shortfall, and cardiac output stays roughly level. Pace does not change. Effort quietly does.
For decades the explanation was straightforward: as you heat up, more blood is sent to the skin to shed heat, less returns to the heart, and stroke volume falls as a consequence. In 2001, Edward Coyle and Jose Gonzalez-Alonso published a review in Exercise and Sport Sciences Reviews arguing the causality runs the other way around. The rise in heart rate is not simply compensating for a falling stroke volume. The faster rate itself shortens the time the ventricle has to fill between beats, and the lower fill is what drops stroke volume.
The distinction matters more than it sounds. If drift were mostly about blood pooling in the skin, drinking more and cooling off would largely erase it. It does not. Drift persists in cool conditions and in well-hydrated runners, because part of it is simply what a faster-beating heart does.
Why the same pace gets genuinely harder
This is the part most runners have never been told, and it reframes the whole thing.
Kevin Wingo and colleagues put nine trained cyclists through 45 minutes at 60 percent of VO2max in 35 degree Celsius heat and then measured VO2max immediately afterward. Between minute 15 and minute 45, heart rate rose 12 percent and stroke volume fell 16 percent. That much was expected. What was not: VO2max measured at the end was 19 percent lower than it had been, even though the riders hit a higher maximum heart rate.
Read that again in running terms. Your ceiling came down. The workload did not change, but the fraction of your available capacity it consumes went up sharply. Your 7:40 pace at mile four and your 7:40 pace at mile sixteen are not the same physiological event, and your heart rate is telling you so accurately.
One caveat worth stating plainly, because it is the kind of thing that gets stripped out when a study becomes a graphic: this was cyclists, in a lab, at 35 degrees. The size of that VO2max drop is a hot-condition number. Drift in 55 degree October air is real but considerably smaller.
Where hydration actually fits
Dehydration is not the cause of drift, but it absolutely makes it worse, and the interaction is steeper than most people assume.
Gonzalez-Alonso and colleagues, working with Coyle in 1997, had fifteen endurance-trained cyclists exercise in the heat for up to two hours, either drinking to stay even or finishing down 4 percent of body mass. Hyperthermia alone and dehydration alone each cost about 7 to 8 percent of stroke volume, and in each case heart rate rose enough to protect cardiac output. Stacked together, stroke volume fell roughly 30 percent, and at that point heart rate could no longer cover it. Cardiac output and blood pressure both dropped.
So the practical version is not "drink more and your heart rate will stay flat." It is that hydration determines whether drift stays a mild, manageable slope or turns into the thing that ends your long run at mile eighteen. The two mechanisms are multiplicative, not additive.
The 5 percent rule, and what it is worth
The standard way to quantify this is aerobic decoupling, written Pa:HR for runners. Take your pace-to-heart-rate ratio for the first half of a run, compare it to the second half, and express the change as a percentage. Under 5 percent is widely treated as the marker of a well-built aerobic base.
That threshold comes from Joe Friel, by way of his training bible and later TrainingPeaks. It is a coaching heuristic built from experience with a lot of athletes. It is not a peer-reviewed cut point, and I have never seen a study establish 5 percent as the line between built and unbuilt.
I still use it, with one change: I do not care where you sit against 5 percent. I care where you sit against you, six weeks ago, on a comparable run. Decoupling is a decent personal trend line and a poor universal standard. Some runners drift 3 percent in peak shape, some drift 8 percent and run 2:45.
And the number is astonishingly easy to corrupt. Temperature and humidity, whether you fueled at mile four or mile fourteen, sleep, a hilly back half, caffeine, starting too fast for the first two miles, and the ordinary noise of an optical wrist sensor will all move it further than a month of good training will.
Reading your own long runs honestly
Four things make the comparison worth making.
- Hold the conditions still. Same route, similar temperature, similar time of day. A drift comparison across a 25 degree swing in weather is not a fitness measurement, it is a thermometer.
- Adjust for terrain. Raw pace on rolling ground makes drift look worse than it is. Grade-adjusted pace is the only fair way to compare a first half against a second half unless the route is genuinely flat.
- Fuel and drink on the same schedule. Given the 1997 findings, a long run where you started drinking at mile ten is not comparable to one where you started at mile three.
- Separate easy runs from marathon-pace runs. Drift on a truly easy effort says something about aerobic durability. Drift during a workout at goal pace mostly says the workout was hard, which you already knew.
Why your September numbers will look better than July's
There is a seasonal trap here, and it catches people right about now.
Julien Periard, Sebastien Racinais, and Michael Sawka have described the cardiovascular side of heat acclimation in detail. Plasma volume expands within the first few days of heat exposure, and that expansion improves ventricular filling and stroke volume, which lets heart rate come down at any given workload.
Run a July summer of long runs and you build that adaptation. Then October arrives, the air cools, and your drift numbers improve dramatically. Some of that is the fitness you genuinely built. Some of it is that you are no longer fighting a 30 degree thermal load, and some of it is a plasma volume expansion that starts decaying within a couple of weeks once the heat stimulus stops.
None of that is a reason to be less pleased with your autumn long runs. It is a reason not to build a goal marathon pace on the strength of one cool, flattering Sunday in late September.
What I would actually do with it
Stop treating drift as a score and start treating it as a pacing input.
If your heart rate is climbing steeply in the back half of an easy long run while pace holds, the run is no longer easy, whatever the watch says. Let the pace go. The aerobic benefit of the last hour does not come from the pace number, and holding it turns a run meant to build durability into one that costs recovery.
On marathon-pace long runs, the drift pattern is closer to a rehearsal. If effort and pace separate at mile fourteen of a twenty, they will separate earlier on race day, when it is warmer and you started faster. That is useful information about your goal pace with three weeks left to act on it.
And build the trend over months, not weeks. A single flat-heart-rate long run is weather. Four of them across a build is fitness.
Where NXT RUN comes in
Most of the work here is comparison, and comparison is where watch software tends to give up. You get one average heart rate for a two-hour run, which is precisely the number that hides drift.
That is what Advanced Run Analytics is for. Highlight any segment of a completed run and you get that segment on its own, including grade-adjusted pace, so you can put the first forty minutes of a long run next to the last forty and see what actually happened between them. The hill at mile fifteen stops distorting the comparison, and the question becomes answerable instead of just interesting.
From there it is a normal coaching conversation. If the split is widening across a build, the answer is usually more aerobic volume and more patience on easy days, not more threshold work.
This article is general training information, not medical advice. A heart rate that climbs sharply and without an obvious cause, or that comes with chest discomfort, dizziness, or shortness of breath out of proportion to the effort, is worth a conversation with a healthcare professional rather than a training app.