The literature on rapid force production has been clear for over a decade. Patients can recover peak strength on the involved side and still be neurally a step behind, and that gap is the one that shows up first when they try to cut, jump, or react. The reason RTD has not been routine in clinic isn't disbelief, it's instrumentation: until recently the only way to measure it was on a high-rate signal most clinics couldn't easily capture, and on a metric most tools didn't compute.
ForceIQ computes it on every trial. This is what the number is, what it means, and where the edges are.
The definition
RTD Early is a two-point slope between the force at onset and the force at onset + 75 ms. Both endpoints are linearly interpolated to land exactly on those times rather than snapping to the nearest recorded sample. Units are Nm/s when torque conversion is available, N/s otherwise.
The math is the same secant you would draw on the curve by hand: rise over run, with the run fixed at 75 ms. There is no curve fit, no goodness-of-fit statistic, no R². At low sample rates this matters more than it sounds. A previous version of the algorithm fit a regression line through whatever samples happened to fall inside the window, which meant a 40 Hz signal computed "RTD over 75 ms" using endpoints that could drift by up to a half sample (about 12 ms at 40 Hz). The reported window was anchored only at the labels, not at the data. Interpolating directly to the target times keeps "0 to 75 ms" actually 75 ms, regardless of sample density. The case for treating force during rapid contractions as a time series rather than a few sampled points is made directly in the rate-of-force-development review literature (Del Vecchio 2022).
Why 75 ms
The 75 ms window is not arbitrary. It captures the phase of force production dominated by neural drive: motor unit recruitment speed and the maximal discharge rate of the motor neuron pool, both of which peak before force generation begins (the neuromechanical delay is roughly 42 ms in tibialis anterior). The maximal discharge rate alone has been shown to explain about 71 percent of the variance in explosive-force variables across individuals (Del Vecchio et al. 2019). After the early window, the dependence shifts: late-window force (around 150 ms) is dominated by maximal voluntary force itself, while the 50 ms window is purely neural and the 100 ms window is contractile (D'Emanuele et al. 2023). The two windows are answering different questions about the same contraction, and the relationship between them is more informative than either in isolation.
Concordance, where RTD Early and RTD Late track together, suggests intact neuromuscular coupling. Discordance, say RTD Early recovered but RTD Late persistently low, points at peripheral or muscular limitations. Read the two values as a pair.
Why peak alone misses what RTD catches
Two patients both produce 600 N peak force on the involved side, both look symmetric on the LSI bar. One of them generates that 600 N in 180 ms, the other takes 420 ms. In every functional task that involves rapid force production (change of direction, landing, deceleration) those two patients perform very differently. Peak force does not distinguish them; RTD Early does.
This isn't a critique of peak torque. Peak is still the headline strength measure and the strongest single predictor of re-injury risk in the LSI literature. RTD Early adds the speed axis on the same contraction. You read the two side by side, not one instead of the other.
How sample rate changes what you can trust
Fixing the endpoints at 0 ms and 75 ms makes the window definition rate-independent. What changes with rate is how much of the curve between those two points the platform actually saw. At 40 Hz the interpolated endpoints sit between two recorded samples each, and the only thing constraining the rise between them is the linear interpolation. At 80 Hz there are six samples inside the window, and any inflection in the rise (a hesitant start, a small re-acceleration) is recorded directly rather than smoothed over.
ForceIQ flags the RTD Early reliability on every session:
| Effective sample rate | Samples in 75 ms window | Reliability |
|---|---|---|
| ≥ 80 Hz (most BLE force gauges, isokinetic exports) | ≥ 6 | high |
| ~55 to 79 Hz | 4 to 5 | moderate |
| ≤ 53 Hz (grip devices, low-rate firmware) | ≤ 3 | low |
The threshold is the sample count, not the rate. Six samples is what a 75 ms window holds at exactly 80 Hz and four is what it holds at about 53 Hz, so the moderate band is narrow and most devices fall clearly on one side of it.
A low-confidence badge does not mean the number is wrong. It means the slope is being computed across a window where the platform recorded only a handful of samples, so a single noisy endpoint or an inflection between recorded samples could shift the value by 20 to 30 percent. The number still ships, because clinicians should see it, but the badge is your reminder to weigh it less heavily. The signal property that makes RTD durable is sample density inside the 0 to 75 ms window; what produces that signal property is a clinical and operational decision, not a platform one.
The earliest RFD windows are also where contraction-to-contraction reliability is weakest even at high sample rates: the 0 to 50 ms slope has consistently shown lower intra-session ICC than later windows in both ballistic and explosive-sustained tasks (Kozinc et al. 2022). The reliability badge is the platform's translation of that consensus into a per-session signal.
When to weight RTD Early heavily
- Return to sport for ACLR, ankle reconstruction, Achilles repair. The ACL return-to-sport literature specifically calls out persistent RTD asymmetry as a re-injury predictor independent of peak strength symmetry. A patient at 95% peak LSI and 70% RTD Early LSI is not yet ready.
- Older or deconditioned patients where rapid force matters for falls risk. RTD Early in the lower extremities correlates with the recovery time available to arrest a fall.
- Performance contexts. Cutting, jumping, reactive sports. An athlete at 110% peak LSI and 75% RTD Early LSI tells a specific story: strong, but slow to fire.
When not to over-read it
A two-point slope across 75 ms reports a number whether or not the rise between the two points was clean. A countermovement dip, a hesitant start, or a double peak inside the window will all still produce a slope value, and it will not look unusual on its own. The number is not the diagnostic; the curve is. Look at the trace before interpreting the RTD value, and if the rise is not a single clean climb, exclude the trial and run another one. That is almost always the right call.
Onset placement is the other thing to keep an eye on. Automated threshold methods (detecting onset at 2 to 3.6 percent of MVF, for example) miss the true onset by 24 to 30 ms (Tillin et al. 2013), which is large enough to invalidate the early-window RFD entirely if you trust it without review. The platform's auto-detection is built to do better than that, and the click-to-override workflow exists for the cases where it doesn't. See correcting the auto-detected onset.
References
- Del Vecchio A. Neuromechanics of the rate of force development. Exerc Sport Sci Rev. 2022. doi:10.1249/JES.0000000000000306
- Del Vecchio A, Negro F, Holobar A, et al. You are as fast as your motor neurons: speed of recruitment and maximal discharge of motor neurons determine the maximal rate of force development in humans. J Physiol. 2019;597(9):2445-2456. doi:10.1113/JP277396
- D'Emanuele S, Tarperi C, Rainoldi A, et al. Neural and contractile determinants of burst-like explosive isometric contractions of the knee extensors. Scand J Med Sci Sports. 2023;33(2):127-135. doi:10.1111/sms.14244
- Tillin NA, Pain MTG, Folland JP. Identification of contraction onset during explosive contractions. Response to Thompson et al. J Electromyogr Kinesiol. 2013;23(4):991-994. doi:10.1016/j.jelekin.2013.04.015
- Kozinc Ž, Pleša J, Djurić D, Šarabon N. Comparison of rate of force development between explosive sustained contractions and ballistic pulse-like contractions during isometric ankle and knee extension tasks. Applied Sciences. 2022;12(20):10255. doi:10.3390/app122010255