What are the biomechanical principles of resistance training?

If you’re interested in cutting-edge biomechanics, please sign up to our free newsletter using the form below and receive a single free study review every month!

Why study the biomechanics of resistance-training?

This review paper is a great introduction to the biomechanical principles underlying the various different resistance-training methods. In this article Chris Beardsley (@SandCResearch) takes a detailed look at the review.

The study: A Biomechanical Evaluation of Resistance Fundamental Concepts for Training and Sports Performance, by Frost, Cronin and Newton, in Sports Medicine, 2010

For the purposes of their review, Frost et al. categorize resistance training into three different types:

  • Isoinertial resistance (constant resistance)
  • Isokinetic resistance (constant velocity)
  • Variable resistance (e.g. bands and chains)

The researchers suggest that understanding the biomechanical principles that underlie each type of resistance training is key to appreciating their benefits and limitations.  What’s more, it helps us figure out which type of training method to use depending on our goals.  Let’s take a look at their findings:


Isoinertial resistance training

Isoinertial resistance training is basically traditional heavy resistance training in which the weight on the bar stays the same throughout.  “Iso” just means “the same” and obviously “inertia” refers to the loading.  So unless you are a hardcore powerlifter with chains hanging from the bar or have a training partner who delights in unloading the bar in the middle of your set, you are generally performing isoinertial resistance training when you train.

Frost et al. explain that the main benefit of isoinertial resistance is that by adjusting the weight on the bar, athletes can perform sports-specific, multi-joint movements at different speeds and with varying accelerations.  After all, lighter weights allow faster speeds and heavier weights necessitate slower speeds.  Since one of the principles of transfer is that some of the training should be velocity-specific to sports movements, this is a key benefit. However, there are two main limitations to isoinertial resistance training:

  • Rate of force development – during isoinertial resistance training, in order to increase the movement velocity at a given load, the force exerted by the athlete must be increased.  However, this increase in force is not compelled but is discretionary.  Frost et al. explain that this can mean that an athlete’s ability to move at higher speeds when using isoinertial resistance training might be limited by their rate of force development (RFD).
  • Mechanical advantages – during isoinertial resistance training (especially standing lower body exercises), there are certain parts of the range-of-motion (ROM) that are easier and certain parts that are harder, because of different mechanical advantages or disadvantages at specific joint angles.  Frost et al. explain that this means that during isoinertial resistance training the prime mover muscles are not maximally activated at all points in the ROM.  This can lead to reductions in movement speed and power at certain points in the ROM, which may be suboptimal to athletic performance.

While these points apply to all isoinertial resistance training methods, Frost et al. further divide isoinertial resistance training into three sub-categories, being concentric-only, stretch-shortening cycle (SSC) and ballistic.  All of these different sub-categories have more specific benefits and limitations, which we discuss in more detail below.


Isoinertial – concentric-only resistance training

While the most natural isoinertial resistance training method is the eccentric followed by a concentric, also called the stretch-shortening cycle method, the simplest is the concentric-only method.  This involves lifting a weight from a static position to a higher static position by means of a concentric contraction of the prime movers.

An example of this method is the Anderson squat, which was used to great effect by Paul Anderson to develop his squat to inhuman levels.  But what do we know about the biomechanics of this method?


Frost et al. note that since a concentric-only movement starts and finishes with zero velocity, total acceleration must equal total deceleration across the whole movement and therefore mean acceleration must be zero.

In other words, if the athlete works hard at the beginning to accelerate the bar, they ease off to the same degree towards the end of the movement.  This is the criticism noted above, for all isoinertial exercises, which means that the prime movers are not activated throughout the whole ROM and therefore they do not get stimulated to gain strength equally at every point in the ROM.


Frost et al. explain that during concentric-only movements, an increase in mean velocity is associated with a decrease in the size of the load lifted.  In other words, lighter weights can be lifted at faster speeds.  This can be seen in the following chart, which shows the mean velocity at different percentages of 1RM:

While what happens to mean velocity is pretty obvious, what happens to peak velocity is much less clear.  Where movements are longer (such as in ballistic movements, where there is no deceleration phase), peak velocity can be higher but this isn’t necessarily meaningful, as rate of force development is not changed.  Frost et al. explain that this has led many researchers to investigate the time to peak velocity.

Time to peak velocity corresponds with the point in the movement that the acceleration of the external load is zero.  After this point, the barbell begins to decelerate.  Therefore, the time to peak velocity divides the concentric-only movement into its acceleration and deceleration phases.

Time to peak velocity is different depending on the percentage of 1RM used.  In other words, it takes longer to develop the required force with higher loads.  The chart below shows how time to peak velocity increases with increasing load:

What can we learn from these points?  Well, considering both of these charts together, we can see from the first chart that when using isoinertial training at high velocities (in order to train at sports-specific speeds), it is necessary to use lighter loads and therefore generate lower mean forces.  Additionally, from the second chart, we can see that lighter loads correspond to shorter acceleration phase durations and therefore a potentially reduced training stimulus.


Power is the product of the applied force and the movement velocity, an idea most people are familiar with.  However, Frost et al. note that many researchers focus simply on the percentage of 1RM that leads to maximum power output.  This typically results in charts like the following, which show that maximum power is achieved at c. 60% of 1RM, while minimum power is achieved at c. 30% of 1RM.  This graph uses the bench press for an example:

Frost et al. note that this focus on just power in total neglects the individual contributions of force and velocity to the power equation.  They explain that the contributions of force and velocity to power at different percentages of 1RM vary, as can be seen in the following chart:

The researchers explain that the contribution of force increases with increasing load, while the contribution of velocity decreases.  In other words, to create maximum power with high loads, you need to be very strong, while to create maximum power with light loads, you need a very high rate of force development.  Without differentiating between the contributions of force and velocity to power, this would not be clear.


Isoinertial – Stretch-Shortening Cycle resistance training

Movements in which the concentric phase is immediately preceded by an eccentric phase are called stretch-shortening cycle (SSC) movements.  SSC movements are profoundly different from concentric-only movements because of the inertia of the external load and because of certain fundamental characteristics of muscle physiology, including the conservation of elastic energy, reflex activity and higher muscle activity induced by the eccentric phase.

The following analysis sets out the differences between concentric-only and SSC movements.  In other respects, the two methods are similar and SSC movements are subject to the same limitations of isoinertial movements in general.


Interestingly, because of these fundamental differences, the peak forces generated during SSC movements are greater than those during concentric-only movements, as can be seen in the chart below:

This is a strong case for the use of SSC movements in general, rather than concentric-only movements, as they allow more weight to be used.  Of course, since most sports use SSC movements in the main, rather than concentric-only movements, they are more specific to sports.


Comparing SSC and concentric-only movements, Frost et al. explain that researchers have found that the mean velocity of the concentric phase in a SSC movement is significantly faster than in a concentric-only movement, at all percentages of 1RM, as can be seen in the following chart:

This is another reason why the use of SSC movements is to be preferred in general, rather than concentric-only movements, as they allow a faster bar speed.  However, despite the difference in mean velocity, Frost et al. note that there is little difference between peak velocities or in time-to-peak velocities.  This is because there is no specific requirement to increase force, which is required for greater peak velocities.   So while SSC movements allow a greater load and a faster speed, they do not allow a greater RFD to be trained than the concentric-only method.


Frost et al. observe that when comparing SSC movements with concentric-only movements, the mean force is the same while the mean velocity is higher.  And, therefore as expected, mean power is higher in SSC movements than in concentric-only movements.

Interestingly, however, we might expect the same pattern to occur when looking at peak velocity and peak force, since peak force is higher in SSC movements than in concentric-only movements and peak velocity is the same.  However, because peak force and peak velocity are not concurrent events, this does not occur.


Isoinertial – ballistic resistance training

Ballistic resistance training involves releasing the external load into the air without decelerating.  It is an attractive training method because increased acceleration phases are not accompanied by equally increased deceleration phases and therefore there are no sections of the ROM in which there is significantly reduced muscle activity.  However, this type of resistance training is not without its limitations, as we will see below.


Unlike concentric-only and SSC movements, ballistic movements are not bounded by zero velocities at both the beginning and end of the movement.  Therefore, researchers have found that mean force and acceleration are greater for ballistic movements than for concentric-only and SSC movements.

For example, Frost et al. cite that Newton et al. observed that ballistic lifts caused an increase in mean force of 35% over non-ballistic lifts, in addition to increases in mean and peak velocity (caused by the lack of deceleration phase).  Interestingly, however, this study showed that peak acceleration, peak force and the time to peak force were not greater in ballistic movements.

Frost et al. comment that this study therefore showed that RFD was not changed by using the ballistic method but that the mean force was increased by an extension of the length of the acceleration phase.  Again, while the isoinertial method can be manipulated to increase mean force and mean and peak velocity, it seems that rate of force development remains unchanged.


As noted above, the mean velocity of ballistic movements is greater than non-ballistic movements because there is no deceleration phase.  For example, using a load of 45% 1RM, Newton et al. found that ballistic movements had a 27% greater mean velocity than non-ballistic movements.  This is simply because there is a longer time available to build up the speed, like a plane on a longer runway.


Since the mean velocity and mean force are both increased in ballistic movements over non-ballistic movements, Frost et al. observe that we should expect significantly higher power measures.  For example, again using a 45% 1RM load, Newton et al. found that mean power as 70% higher using a ballistic method than a non-ballistic equivalent.  Frost et al. note that increases in peak power would be expected for the same reasons and studies confirm that this is the case.  Again, however, this is a function of having a longer run-up before launching, which leads to greater power production.

Isokinetic resistance training

Frost et al. explain that isokinetic machines were developed to provide a third leg to isoinertial (constant resistance) and isometric (zero speed) training methods.  Isokinetic training involves constant velocity, irrespective of the force applied.  The velocity can be changed, to allow sports-specific training at different velocities.


Frost et al. explain that isokinetic training enables force to be maintained at a maximal level by increasing the mass, instead of maintaining a high level of acceleration, which is not possible in non-ballistic isoinertial movements.  Additionally, since there is no significant deceleration phase in which there is a reduction in muscle activity, Frost et al. note that, for comparable velocities, isokinetics require greater mean forces but lower peak forces than isoinertial movements.


Aside from the initial acceleration and deceleration to begin and end the movement, in isokinetic training, there is no change in velocity involved.  However, clearly, as the machines are set at higher velocities for training at higher speeds, this will result in greater periods of initial acceleration and deceleration.


Variable resistance training

Frost et al. explain that variable resistance devices are similar to isokinetic machines in that they alter the force during the movement intentionally, to match changes in the mechanical advantage at different points in the joint ROM.  They note that the original Nautilus machines were developed with this principle in mind and pneumatic machines have since also been created in light of the same idea.  However, they note that most common currently used variable resistance techniques are bands and chains, made popular by powerlifters.

Frost et al. explain that the use of bands allows for greater accelerations at the bottom of the lift, where the bands are loose and therefore the resistance on the bar is at a low percentage of 1RM.  As the lift moves through its ROM, the bands stretch and add resistance proportional to their stretch coefficient.  This requires the production of even greater accelerations in the first part of the lift, since the deceleration phase is augmented by the increased inertia of the external loading.

Frost et al. also note that the use of bands leads to greater eccentric accelerations and therefore a greater involvement of the SSC for subsequent repetitions.  Research has not noted changes in speed as a result of such use of bands and chains but studies into adding weight on eccentric repetitions have found increased strength on the subsequent concentric phase.

Frost et al. explain that chains also allow greater accelerations at the bottom of the lift, where the chains are in contact with the floor and therefore do not add any resistance.  Again, as the bar moves through its ROM, the chains cause an increase in mass that means that greater forces are needed to finish the lift.  These greater forces again require greater accelerations in the first part of the lift, as the deceleration phase is augmented by the increased weight of the chains.

Frost et al. observe that both bands and chains add additional resistance where lifts have a mechanical advantage at the end ROM of a squat or bench press.  This is thought to increase RFD, although research in this area is still young.


What did the researchers conclude?

Frost et al. drew a number of key conclusions from their review, as follows:

Isoinertial (traditional) resistance training is limited by mechanical disadvantages, which make certain parts of the movement harder than others.  This means that the muscle activity of the prime movers is much greater at certain points in the ROM than in others and therefore that training adaptations occur unevenly.

Isoinertial (traditional) resistance training is limited in respect of its ability to increase velocity-specific power because higher power is generated during isoinertial resistance training at lighter loads and shorter acceleration phases, which reduces training adaptations.

There are significant differences in force, velocity, power and muscle activity between concentric-only and SSC movements because of the inertial properties of the load and the physiological behavior of muscle fibers.  However, exactly why muscles behave in this way is not entirely clear.

Ballistic movements can be used in isoinertial resistance training to increase mean force and mean and peak velocities but do not alter peak force or RFD.

Compared with equivalent isoinertial (traditional) resistance training, bands and chains allow for greater accelerations and velocities at the start of the concentric phase and prevent a reduction in force, velocity, power and muscle activity at the end of the concentric phase.

Bands and chains  also increase the eccentric phase loading and increase the effect of the SSC.