Before I begin, let me state that these graphs and information are all drawn from actual race performances: I can quote the date, the time, the actual event and the runner involved. I have the permission of the runner(s) involved to post these graphs. The information is true, genuine and accurate. I used a Polar Sport-Tester HRM to record them, with HR recordings being taken every 15 secs (this is programmable: on this Polar model recordings can be made every 5, 15, or 60 secs, these can be downloaded later to a computer for analysis).
Although in a 2hr40 marathon I will get ~640 data points (one every 15 secs) for clarity only representative data will appear in the graphs on this thread, and it is enough to show only one single HR recording per 1-km. This does not reduce the accuracy of the graphs, and no information has been removed that would alter the interpretation of any of the graphs. As shown, these graphs offer true and accurate representations of actual performances. Some day I may post the unedited originals in an eBook or online in a future Hadd website.
Although these graphs are being posted in a public forum, I retain all and full rights to them and their usage. It is possible they will promote discussion in other fora and I foresee no problem with others linking to these graphs for discussion purposes, so long as they represent them exactly as they appear here; unaltered in any way.
Finally, I would like to thank Weldon for generously hosting the graphs and thus allowing the following discussion(s) to take place.
Let’s begin with a discussion of the attached graph:
This is the HR profile of a marathon. As a performance standard it would generate ~1,000 points on the IAAF Scoring Tables (available for download from the IAAF website). Such a point score represents a sub-2:40 performance for a woman and a sub-2:20 performance for a man. I can also produce graphs of higher-quality performances, but these will suffice for demonstration.
I have mentioned earlier that an HR graph for a good marathon performance is a “flat” graph. This can readily be seen from the attached graphic. "Flat" is to be taken as meaning "not rising over the course of the event", although as can be seen, HR does fluctuate within a narrow HR-band. This is to be expected.
From the 5k point onwards in the race, it can be seen that the HR remains secure within a HR-zone between 175-180 bpm.
Only for a total of less than 2 minutes in ~2hrs 40 does the HR rise above 180. It hits 182 at 3 points in the race, but drops instantly back under 180. The combined total time over 180 bpm is less than 2 mins in the whole race.
Once the race passes the 5k point, the HR does not drop under 175 for the entire remainder of the race; not for one single second.
This particular athlete has an HRmax of 196.
Therefore the lower limit of the Marathon HRzone of 175 bpm = 89.28% of HRmax and the upper limit of 180 bpm = 91.83% of HRmax
Therefore the athlete runs the entire way (after initial 5k) with the HR fluctuating within the narrow band between 89.2—91.8% HRmax.
The average HR (mean) is 176.84 bpm. This = 90.22% of HRmax. (for interests sake, the average HR (mode) is 177 bpm).
Therefore the HR fluctuates only ~1% either side of the mean throughout the race: never higher, never lower.
Let me make the point here that this athlete did not run "according to HR". This athlete most definitely did not run constantly looking at the HRM and adjusting pace accordingly. From memory, the runner cannot recall looking at the HRM more than twice in the whole race... if that often.
The runner ran this marathon purely and simply as a race. The HRM was worn simply for my benefit and in no way to guide or influence performance. It was worn so that I could learn what actually happens during a high-quality marathon performance. Antonio has already remarked that too many studies are drawn from sedentary individuals and extrapolated (with questionable conclusions) to elite runners. This was an opportunity for me to "observe" a genuine high quality performance under actual conditions.
The Green Line on the graph is set at 177 bpm and represents this runner's Lactate Threshold (LT) which corresponds to 2 mmol/l blood lactate ([La]b). It can thus be noted that 2mM is the effort level this athlete can maintain for a full marathon. This is in line with the published literature on world class performances.
Readers should not confuse what I term LT as being that intensity which correspondes to "maximal effort for one hour". The sports science terminology can often be confusing (with LT, OBLA, MLSS, LTP, OPLA, AT, IAT, etc).
Those whom I have advised online for some years will know that the training model I use mandates two "thresholds". Up until very recently I referred to these as Aerobic Threshold (AeT) and Anaerobic Threshold (AnT). Given that this terminology is now recognised to be both inaccurate and misleading, in line with the literature I now use the (respective) terms Lactate Threshold (LT) and Lactate Turnpoint (LTP). The former (LT) corresponds extremely well (and is useful) with training for marathon performance, while the Lactate Turnpoint correlates better with MaxLaSS (Maximum Lactate Steady State) and thus HM and 10k performances.
So, just to repeat; the Green Line on the Marathon Graph corresponds to this runner's LT. It should be noted that it also corresponds almost exactly to this runner's meanHRavMarathon (177 HR vs 176.84 bpm)
To answer some FAQs:
Why is a flat graph optimal in a marathon?
We can look at this in some depth, but a flat graph generally signifies a steady power output over the whole race: an even distribution of fuel. Obviously since this runner averaged ~90% HRmax, she could have run the whole race at any effort level below that (e.g.: 85% HRmax) and still have had a "flat" graph... so a flat graph on its own is not indicative of the best possible marathon performance on any given day. As much as possible the athlete wants to maintain a flat graph at the highest possible percentage of HRmax.
Here we must again note the quality of performance (sub-2:40 female or sub-2:20 male). Those runners unable to achieve this level of performance may need to revise (down) the estimated percent of HRmax they can sustain for the marathon, and work the numbers accordingly. The evidence of this graph is not to be taken as proof that 90% HRmax performance is achievable by everyone.
Some data has been posted in the past on LetsRun representing athletes I have advised online and demonstrating average marathon HRs of 88-89% HRmax. That these are lower percentages of HRmax than the example given in this graph may be partly explained by the fact that they are lower in terms of quality of performance (i.e.: slower than sub-2:40 female and sub-2:20 male).
Conversely I have every reason to believe that Paula Radcliffe maintained ~92% HRmax when she ran the far superior marathon performance of 2:15 (AM Jones, 2006). Were the HR data available, I would also expect Ryan Hall to have done close to the same in London ‘08 (>91% HRmax).
So the percent of HRmax that can be sustained for a marathon may ultimately be linked to the time duration of the performance. Whatever the level, if well-trained, the runner should always allow a 5-bpm HRmarathon zone to run in (e.g.: 171-176 - and in this case the runner might hear me tell them pre-race, "you never want to see 170 or 177 at any time after 5k.")
Why is a flat graph optimal in a marathon?
An HR graph that falls over the course of a marathon should mean only one of two things:
a) the runner is winning comfortably and can afford to slow down (thus the running pace also slows along with HR). I have a graph which can demonstrate exactly this.
b) the runner is beginning to experience severe drop in muscle glycogen and the amount of muscle mass that can be recruited is dropping (thru fuel exhaustion). Lower recruited muscle mass will require reduced blood flow, so cardiac output drops. Obviously the running pace will also drop and the runner will slow in the later stages of the race. Once again I have graphics that will support this.
More graphs and discussions to come...