The ability to train with power has been available to cyclists for more than twenty years now, and researchers have been using power as a metric in the lab even longer. Training and racing with power in cycling has grown from a nerdy niche for early adopters (like me), to nearly ubiquitous in athlete groups ranging from junior cyclists to pro triathletes to UCI pro teams. Consequently, metrics related to training and racing with power have grown up with (and for) cyclists.
In contrast, power training has been available to runners for only about three years. Some power metrics don’t carry the same utility when crossing over from cycling to running, and some of the assumptions that apply to cycling do not apply equally to running.
Needless to say, there are differences in the use of power as a metric for cycling and running. Let’s take a look at ten of these differences.
Normalized Power (NP) has diminished utility in running.
NP has a great utility and wide range of applications in cycling. Cycling power output is highly variable, as cycling involves periods of coasting, especially in the peloton or on downhills. Running power output is typically not nearly as variable as in cycling, and there is no coasting in running.
It’s quite atypical to see a run produce NP that is more than a watt or two different from the entire run’s average power (AP). Often, the NP and AP of a run are the same — it’s possible to see NP a watt or two lower than AP in running, which is virtually never seen in cycling.
Perhaps the main application of NP in running would be in the assessment of an overall interval session or in mountain/trail running, where power can be more variable due to power hiking some uphills and gingerly “running” technical downhills. However, for the majority of running, NP has no profound utility over AP.
Variability Index (VI) is typically near 1.00 in running.
The equation for VI is normalized power divided by average power (NP/AP). In cycling, the VI of a workout is typically greater than 1.00, and often over 1.10 in criterium racing. As mentioned already, in running, NP and AP are typically the same, or within one or two watts of each other. This means that VI in running is often 1.00 to 1.01. VI becomes a rarely-used, dusty metric shoved in the back drawer of running analytics.
Intensity Factor (IF) is typically no different from AP as a percent of set functional threshold power (FTP).
IF is calculated as NP divided by FTP. Because NP is rarely any different from AP, AP/FTP (or %FTP) typically produces the same, or nearly the same, value as IF. In other words, in running, IF and %FTP can generally be used interchangeably.
Running FTP does not equal cycling FTP.
Any resemblance of your run FTP and your cycling FTP is purely coincidental. Different muscle activity, different mechanics, and different power-measuring technology result in different FTP values.
Running involves eccentric muscle contractions.
Cycling involves little or no eccentric muscle contractions. The differences in eccentric contractions between running and cycling may account for differences in training load tolerances. For example, a cyclist may sustain CTL ramp rates of up to +7 CTL/week quite well, while a runner may max out at +5 CTL/week (or less, depending on hills and high-intensity training in the CTL composition).
Reported Pmax and FRC may be a bit lower than reality in runners.
Cycling power meters tend to be strain-gauge-based direct-force power meters (DFPM) that can instantly capture short-duration, high-power bursts. Metrics like Pmax and FRC, which reflect the shorter-duration portion of an individual’s power-duration curve, likely reflect reality in cycling (barring technical glitches).
Running power meters (such as Stryd) may use accelerometers and gyroscopes to collect the data from which power is calculated, and the power algorithm may smooth the data to some extent. Because of these factors, the short-duration, high-power data (think sprinting) generated from a running power meter may be lower than reality. In turn, modeled metrics like Pmax and FRC may be a bit lower than reality for runners.
Distance runners may tend to have lower W’, AWC, and/or FRC than cyclists.
Putting aside the differences between very short-duration power calculations and handling described above, it’s quite possible that a distance runner’s W’, anaerobic work capacity (AWC), and functional reserve capacity (FRC) are generally lower than those of cyclists.
Just as there are differences in W’, AWC, and FRC among climbers, road sprinters, and track sprinters within the cycling world, there are likely differences in W’, AWC, and FRC between cyclists in general and runners in general. Whether it’s due to morphologic differences, selection, or training, distance runners may actually have lower W’, AWC, and/or FRC values than cyclists.
Runners typically demonstrate a much more feeble VO2max slow component than cyclists.
In cyclists, there is a demonstrable VO2 slow component, as there is a range of power-duration relationships that can access VO2max. For example, it might be that a cyclist can access VO2max while riding at 380 watts for four minutes. The same cyclist might also access VO2max by riding at 370 watts for eight minutes, wherein VO2 rises early, starts to plateau, and continues to rise slowly until VO2max is reached. This slow-rise phase until VO2max is reached is the slow component of VO2. “Power at VO2max” can be variable, depending on power-duration relationship and protocol.
In runners, the slow-rise component of VO2 is comparatively more feeble, as the range of power-duration relationships that access VO2max is far more limited. In a runner, 380 watts for four minutes might access VO2max, but 370 watts for eight minutes might not because of the more feeble VO2 slow component.
FTP lies closer to power at VO2max in runners than in cyclists.
LT2 (second lactate threshold) has been demonstrated to lie closer to VO2max in runners than in cyclists. Since FTP and power at LT2 are relatively close, one can infer that FTP likely also lies closer to power at VO2max in runners than in cyclists. In trained runners, one might expect FTP to lie at about 85 to 87 percent of their power at VO2max, and closer to 90 percent (or higher) of power at VO2max in elite runners. In contrast, FTP in trained cyclists might lie at 80 to 82 percent of their power at VO2max.
In a runner whose FTP is 87 percent of power at VO2max, VO2max occurs at about 115 percent of FTP (nearly 1600m race power for many runners). In the same runner, 90 percent of VO2max (a common demarcation of higher-intensity interval training) could be achieved with appropriate-duration work intervals done at 103.5 percent of FTP (a bit higher intensity than 10K power for many runners), and 95 percent of VO2max could be achieved with appropriate-duration work intervals done at just over 109 percent of FTP (higher intensity than 5K power and lower intensity than 3K power for many runners).
In contrast, in a cyclist whose FTP is 81 percent of power at VO2max, VO2max occurs at about 123.5 percent of FTP. This cyclist would have to perform an appropriate-duration work interval at about 111 percent of FTP to access 90 percent of VO2max, and nearly 118 percent of FTP to access 95 percent of VO2max.
Metabolically, runners demonstrate more variability in efficiency than in cyclists.
Metabolic efficiency is more variable in runners as a group than cyclists as a group. Further, as individuals, runners tend to have more improvements in metabolic efficiency with training than cyclists (intra-athlete variability of metabolic efficiency is lower in cyclists than in runners).
While there are similarities in training to run and training to cycle, there are also distinct differences. It’s wise to understand these differences so that faulty assumptions are not mistakenly carried over from training for cycling with power to training for running with power.