You don’t need to spend too long on youtube to find a clip of a bodybuilder (or anyone really) performing an exercise with what the official internet exercise police deem illegal technique. The comments degrade to the usual back and forth, those talking of potential injury, how he’d be so much bigger if he just used proper technique, and then the discussion eventually reverts to mudslinging and name calling. Sometimes I wonder if Arnold would have had the success he did if the internet existed back then. I doubt they’d let him get away with curls like those shown in the picture below.

But if many of these bodybuilders are using such sub-par form, how is it that they are still able to create significant amounts of muscle mass? We can chalk it up to genetics, or maybe supplements and drugs, all contributing factors when it comes to building ridiculous muscle mass, but these arguments are easy, tired and old, at least in my opinion.

Fortunately for us a recent article in the European Journal of Applied Physiology has taken a new look at whether or not cheating pays, at least in the weight-room. The answer may shed some light on the paradox where everyone complains about how many large bodybuilders have such success despite form that is often considered less than ideal.

How exactly do you model cheating?

This particular study wasn’t a training intervention comparing those who use some body english (cheating) in their lifting with those who don’t, but rather a mathematical model used to determine how applying external momentum alters the net muscular torque produced by the muscle that would normally perform the lift (the agonist). The authors tested a mathematical model of shoulder abduction (lateral raise exercise) under conditions of varying load and momentum. In this case, they mimicked the application of cheating by applying an initial angular velocity to the lift, similar to what would happen if you put your back into your biceps curls to get the bar moving before completing the curl with the activity of your elbow flexors.

When the initial angular velocity in the lift was set to 57.5 degrees per second to simulate cheating with a fixed 10 kg load, peak torque was higher on all reps except the first in the set, indicating that cheating improved torque during the lift. Torque was also maintained at a higher level as the set progressed and fatigue set in, however the authors didn’t explicitly state whether the increased external momentum would allow the completion of additional reps with the training load. Under these conditions, the time to complete each rep also decreased, such that the total time of the set was reduced. The simulation didn’t include an increase in the number of repetitions, which given a modelled 10RM load, I think it may be reasonable to assume that additional reps could be performed when external momentum is added to the equation.

They ended the paper with a situation closest to what would happen in the gym, the use of momentum (added initial velocity) and a greater training load (12.5kg, a 25% increase) was compared with lifting the original 10kg load without momentum. The load increase resulted in fewer repetitions completed per set, from 10 to 8, however the set could be extended provided that the cheating was progressively increased (increased initial added velocity). This is an interesting point, as all the simulations in the study involved setting a specific initial angular velocity, whereas in the gym it’s likely that lifters attempt to use progressively greater momentum as fatigue sets in.

All in all, these numbers demonstrate that the use of muscles additional to the prime movers to generate momentum can still result in increased torque of the target muscles, contrary to popular belief. The authors do note that there is a point of diminishing returns, so there’s a fine balance between providing just enough ‘kick’ to aid in completion of the reps to going all out and compromising the lift.

How does this relate to hypertrophy?

The final portion of the paper attempts to take the mechanical variables and convert them to the relevant physiological adaptation, in this case muscle hypertrophy. Unfortunately, in order to transform the data into a physiological adaptation, it was assumed the the potential for hypertrophy follows a sigmoidal relationship (think ‘S’ shaped), where lower %RM training results in negligible hypertrophy, moderate %RM training rapidly promotes hypertrophy, and above a certain threshold, no additional benefit is obtained from increased load, at least from a hypertrophic standpoint.

I don’t want to continually hype these papers, but given recent data suggesting that muscle hypertrophy is similar between 30%RM and 80%RM training (3), and that protein synthesis is at least equally stimulated between the two conditions (2), it’s apparent that this sigmoidal relationship doesn’t hold true and it’s time to re-evaluate.

Nevertheless, the mechanical data is still interesting and relevant, although it would have been nice to see how loading and ‘cheating’ interplay to manipulate the number of reps performed in a set and the potential that increased repetitions at a given weight when cheating compensates for the reduction in time-under-tension on a per-repetition basis.

So does it really pay to cheat?

This was an interesting study but it’s important to keep in mind that it was a computer simulation, not an actual training study. It is an interesting start though, and it fits with what we’ve seen anecdotally: many deviate form and have muscle-building success despite straying from what is considered optimal form.

This is not a call to arms to throw caution to the wind and to get the bar or dumbbell up by any means necessary. Proper exercise technique is paramount in the weight-room, but for experienced lifters there’s potential for momentum derived from muscles other than the prime movers to facilitate additional torque production from the target muscles.