The Unescapable Fatigue Effect

timclarkAt the Races, Research

In an earlier article, I discussed the mechanics of how fatigue impacts pace during a marathon, but what happens to other metrics like ground contact time, flight ratio and impact characteristics?  We know these metrics are predictably affected by running speed in an unfatigued state.  As running speed increases, impact Gs, pronation, pronation velocity and flight time usually increase.  Contact time drops because our feet spend less time on the ground.  Using data captured from the 2014 NYC Marathon, we look at how fatigue impacts those dynamics. The study consisted of::

  • Raw RunScribe data from 11 runners split into 3 groups
  • Group 1 – Held a consistent pace throughout race (orange)
  • Group 2 – Consistent through first half and then slowed slightly during 2nd half (pink)
  • Group 3 – Started slowing at around 10k and dropped off significantly toward the end (grey)

While pace did not change for the consistent group, impact Gs slightly increased throughout the race.  Also, the impact Gs of the other two groups steadily increased even though their pace dropped through much the race.  

This is not what you would expect. Fatigue is clearly having an impact (no pun intended!) on footstrike characteristics.  This is likely caused by a gradual loss of neuromuscular control as runners become more and more fatigued.  At submaximal paces, like marathon paces, not all the motor units (bundles of muscle fibers) fire with every contraction.  They actually tend to cycle in and out of use in an attempt to delay fatigue of the muscle group as a whole.  As fatigue becomes more severe, some of these motor units fire less frequently (or not at all) causing firing patterns to change and forcing other, less fatigued units to take up the slack.  As these muscle firing patterns change, runners may become less adept at cushioning the impact as their feet collide with the ground.  As fatigue worsens, impact Gs slowly increase. Not only are runners forced to absorb higher impacts while fatigued, but overall efficiency appears to drop.[Tweet “Fatigue forces runners to absorb higher impacts while their efficiency drops.”]

Running efficiency is a difficult thing to measure. In fact it’s impossible to measure outside of a lab. But a number of descriptive studies have tried to determine some basic biomechanical features of more accomplished runners with more efficient running form. One consistent finding is that more accomplished runners tend to have lower ground contact times (CT) at a range of speeds than less accomplished runners. Their feet are in contact with the ground for a shorter amount of time during each running stride. It’s not clear why this is beneficial but one possibility is that a shorter CT gives the runner a better opportunity to use the elastic energy stored in the muscles and tendons of the feet and lower leg at each footstrike. During the early part of every stance phase, energy is absorbed in the tendons and muscles of the foot and leg as they contract and stretch to stabilize the leg to support the runners weight against gravity. During the latter part of the stance phase those elements recoil like a rubberband under tension, contributing to the force required at toe off.  This is called the stretch shortening cycle (SSC) and the shorter the time duration of this cycle, a larger portion of this elastic energy is available for use at toe off. Stated another way, the longer the CT, more of the absorbed energy in the muscles and tendons is dissipated and the more difficult it is to use this aerobically “free” energy.  Many researchers have studied the SSC and estimate that at faster paces, upwards of 20 to 50% of the total energy required to maintain pace could come from these aerobically “free” sources. (1,2,3,4)

Looking at the CTs for the three groups, the group that slowed the most (grey line) had the greatest change in CT during the race, increasing by over 40%. The even paced and slightly slowing groups increased only 4% and 7% respectively.  It is interesting that the CTs for the group that ran a very consistent pace (orange line) also increased slightly. Their stride rate was basically unchanged (less than 1%) meaning the average time from one foot strike to the next was the same throughout the race.  The 4% increase in CT means the time they spent in the air (flight time) between FSs had to drop.  They were pushing off with less force as the race progressed.  The result of these minor changes was the small shortening of this group’s stride length.  The changes in these metrics for the consistent pace group were slight.  For more evidence of this cascade of events, look at the other two groups.  The same changes in these metrics occur but more dramatically.  CT increases, flight time decreases, SR and SL both decrease and as a result the runner’s pace drops.  In some cases, quite rapidly.

So what is the true significance of this?  At the start, all runners are not fatigued and probably running at high efficiency. As fatigue becomes a factor, pace drops, CT increases and relative running efficiency is reduced, in part because it becomes more difficult to utilize this “free” energy from the SSC. Even the most prepared are not immune to the effect. By tracking metrics like Contact Time, Impact Gs and Flight Ratio during individual runs throughout marathon training, it’s possible to better evaluate your progress and how prepared you are to achieve your race goals.



1) Cavagna GA(1), Legramandi MA, Peyré-Tartaruga LA, Old men running: mechanical work and elastic bounce. Proc Biol Sci. 2008 Feb 22;275(1633):411-8.

2) Asmussen, E. and Bonde-Petersen, F. (1974), Apparent Efficiency and Storage of Elastic Energy in Human Muscles during Exercise. Acta Physiologica Scandinavica, 92: 537–545. doi: 10.1111/j.1748-1716.1974.tb05776.x

3) Cavagna, Giovanni A. “Storage and utilization of elastic energy in skeletal muscle.” Exercise and sport sciences reviews 5.1 (1977): 89-130.

4) T. A. McMahon, G. Valiant, E. C. Frederick, Groucho running

Journal of Applied Physiology Published 1 June 1987 Vol. 62 no. 6, 2326-2337 DOI: