Your health comes first. This educational article provides general information. Discuss your individual circumstances, pregnancy, illness or symptoms with your clinician; the article cannot replace an individual assessment.

‘Less positive work at the knee’ sounds like a promising result in a running-shoe study. But precisely what decreased? The whole body’s energy demand, thigh-muscle work or a net joint quantity calculated in a movement model? These are not the same. Interpreting a research graph starts with understanding how the curve is produced, what its sign means and which component processes its summary hides.

This article works through a common report containing joint moment, angular velocity, power and work. Original numerical examples help make the formulas understandable, while specific shoe and tendon-modelling studies show how these quantities are used. You do not need to be a biomechanist to recognise when an interesting knee curve is being made to say too much about the runner’s whole function.

First column: net joint moment

Joint moment describes a turning effect, measured in newton-metres. In movement analysis it often comes from inverse dynamics: motion, external forces and a body model are used to calculate the required net turning effect. ‘Net’ matters: it represents the combined result of several contributions. It does not separately reveal every muscle’s force and function.

The official OpenSim documentation specifically emphasises that net model quantities must not be confused with forces between joint surfaces. Net moment likewise differs from a single muscle’s pulling force. If a graph is labelled only ‘knee loading’, request the precise definition and unit. Its colour does not reveal the physical quantity represented.

An original simple example: two opposing turning effects are 60 and −40 newton-metres. Their sum is 20 newton-metres. Another case with 35 and −15 also gives 20. The net result is identical, but contributions differ. These are deliberately stripped-down number pairs, not actual muscle tests. They remind us that an aggregate cannot reliably reveal its components without further information or modelling assumptions.

Second column: angular velocity

Joint position alone is insufficient for power. You need to know how quickly the angle changes. Calculations often express angular velocity in radians per second. A complete circle is 2π radians, and 180 degrees is π radians. If an angle changes by 0.20 radians over 0.10 seconds, average angular velocity is 2 radians per second. This is approximately 114.6 degrees per second, but the two numerical values must not be carelessly substituted in multiplication.

A common misunderstanding is that a large angle must mean high power. Not necessarily. At maximum flexion, movement may be reversing direction and angular velocity temporarily small. The same moment gives different power during fast and slow rotation. The peak of an angle curve is therefore not necessarily the peak of the power curve.

Moment and angular-velocity signs must refer to the same coordinate system. If two graphs use different conventions for positive flexion and extension, visual comparison may mislead. A transparent report describes directions, units and whether the calculated moment follows an internal or external naming convention. Without these, positive and negative values can easily tell the wrong story.

A runner undergoes a treadmill measurement wearing a respiratory-gas analysis mask under professional observation in a sports-physiology laboratory.Enlarge image
AI-generated editorial illustration of laboratory assessment of running economy; not a documentary photograph.

Power is moment multiplied by angular velocity

The simple rotational-mechanics relationship is P = M × ω. If moment is 100 newton-metres and angular velocity is 2 radians per second with the same sign, power is 200 watts. With the opposite angular velocity of −2 radians per second, the result is −200 watts. Our illustrative example disregards all other circumstances to examine the product alone.

Positive power represents energy generation and negative power energy absorption in the chosen net joint description. This does not automatically mean one muscle performed the same amount of positive or negative work. A net joint curve summarises at model level. The muscles, tendons and neighbouring segments involved require further description.

Timing matters too. Imagine two situations both with a moment peak of 100 newton-metres and an angular-velocity peak of 2 radians per second. If they occur together, their product may be 200 watts. If they occur at different instants, actual peak power is different. Multiplying two separate maxima is therefore not a general way to calculate maximum power. The simultaneous data must be connected.

Work is the area under the power–time curve

Summing power over time gives work. An interval imagined at a constant 200 watts for 0.10 seconds gives 20 joules. Power varies in real running, so areas under its positive and negative portions may be calculated separately. The window may cover the whole stance or a defined portion. Stance duration and the selection rule are therefore part of the result.

An original example: in one fictional knee model, positive work is +20 joules and negative work −15 joules. Net work is +5 joules. Another model with +10 and −5 joules also gives +5 joules net. Positive and negative portions differ despite the identical net result. Showing only the final report column loses that distinction.

Our model Positive work Negative work Net work
A 20 J −15 J 5 J
B 10 J −5 J 5 J

Distinguish joules from work values normalised to body weight or body mass. J/kg means work divided by mass; watts measure power. A larger work area can come from longer duration, not only a higher peak. A step graph normalised to percentages may also hide total duration, so interpreting area physically requires the actual time base.

Why does a net curve not show every muscle separately?

The OpenSim description of static optimisation introduces another calculation criterion precisely to distribute a net task among individual muscle forces. Alongside the model and external forces, the method uses a selected optimisation rule. This requires more information than a moment curve alone. Other simulation approaches may use different assumptions, so method name and settings belong with the result.

A muscle spanning two joints is particularly thought-provoking. When joint positions at both ends change, their total length change may contribute to that same unit’s function. One knee graph does not explain what happens at the other joint. Energy transfer between several segments is therefore a more detailed question than reading the knee’s positive area.

You do not need a full muscle simulation to interpret every result. What matters is the level of the claim. ‘Net knee work decreased’ may follow directly from the displayed analysis. ‘The thigh muscle consumed this much less energy’ is another claim requiring further evidence about muscle function and energy demand. Measuring the first number precisely does not make the second statement true.

Tendon movement and muscle-fibre movement can differ

The total length change of a muscle–tendon unit connected in series is not identical to each component’s length change. For an original mechanical illustration, imagine two serial elements: one lengthens by 5 millimetres while the other shortens by 5 millimetres. Total length is unchanged, but the components have moved. Watching only the whole unit hides this opposing internal change.

In Uchida and colleagues’ whole-body simulation, changing tendon compliance affected individual muscles’ estimated energy demands differently. The model accounted for fibre operating length and shortening too. The result illustrates how elasticity and active muscle function jointly determine the system; elastic return does not make running passive.

A joint-power curve is therefore not a direct calorie meter. Adding positive and negative mechanical portions does not replace metabolic data derived from respiratory gases or a separately evaluated muscle-energetics model. Place mechanical and physiological results beside each other and ask which question each answers. This avoids immediately promising an equal proportional energy saving from a reduced work value.

A man resting at a table after running calmly checks the sports watch on his wrist; water and a folded towel are beside him.Enlarge image
AI-generated editorial illustration; not a documentary photograph.

What does a shoe-stiffness experiment teach?

Cigoja and colleagues, in thirteen recreational male runners, examined the same shoe with and without a stiffness-increasing insert. The stiffer condition produced more positive and less negative work at the metatarsophalangeal joints while positive knee work decreased. The study thus provided a specific example of changed net joint-work distribution.

A related randomised crossover study modelled the foot arch and muscle–tendon units with a similar shoe modification. Arch deformation, shortening of some units and shortening velocity also changed. These quantities came from models and movement data, not direct measurements of every muscle fibre. The shoe–body relationship can alter several component processes simultaneously.

The practical lesson is to read the whole figure. Do not isolate only the apparently most favourable knee column: examine foot, ankle and hip data, positive and negative work, and running speed. A short laboratory situation may offer a mechanical explanation but does not alone establish long-term injury protection or identical race benefits for every runner.

A reading sequence you can check in one minute

Begin with the quantity’s name. Moment, power or work? Then check its unit and sign, followed by the period: the entire stance or a selected part? Only then examine percentage differences between groups or shoes. The change is then attached to a precisely defined quantity rather than the runner’s ‘loading’ in general.

  • Moment: a net turning effect, for example Nm.
  • Angular velocity: the rate of angle change, in the appropriate unit for the product.
  • Power: simultaneous moment multiplied by angular velocity, W.
  • Work: the area under power over time, J; positive and negative parts can be read separately.
  • Energy demand: a separate measured or modelled physiological question.

If an article’s title mentions ‘saved knee work’, be able to state exactly which metric decreased and under which conditions by the end. That is already a meaningful finding. You do not need to imagine reduced muscle force, smaller contact force and lower calorie use alongside it. A biomechanics graph becomes useful when read using its own concepts and when you seek the appropriate additional data for a decision based on it.

Sources

Helpful? Pass it on.