How often should you lift to failure?

Chris Beardsley is a biomechanics researcher. He is also the editor of Strength and Conditioning Research. He is based in Loughborough, Leicestershire.

Lifting to muscular failure when strength training is a tactic used by bodybuilders, pro athletes and the general population for maximising gains in muscular size. But is going all the way to muscular failure really better than leaving a few reps in the tank?

Before we go any further, we need to establish exactly what the term ‘muscular failure’ means. Muscular failure is frequently used in research studies but precise definitions of this term are quite rarely discussed. In an old review, a researcher once defined muscular failure as ‘the point during a resistance exercise set when the muscles can no longer produce sufficient force to control a given load’[1]. This definition has since been tightened by another researcher who suggested muscular failure involves ‘the point during a set when muscles can no longer produce necessary force to concentrically lift a given load’[2].

Another group recently went even further by defining it as ‘the inability to perform any more concentric contractions, without significant change to posture or repetition duration, against a given resistance’[3]. It seems that everyone wants to add their own extra clause on somewhere!

Relative load
Whether such additional clauses are necessary to the original definition is probably a moot point. The really important part of all these definitions is that muscular failure is much defined in relation to a given load. This should be immediately apparent when we consider that bodybuilders often perform repetitions to failure and then immediately drop the weight and using a lighter weight to continue performing several more repetitions, when doing drop sets. They go to muscular failure several times without stopping, simply by changing the weight.

So muscular failure does not really mean that a muscle is incapable of performing further muscle actions. It means that the muscle cannot perform any more muscle actions with one particular load. In reality, we cannot say that muscular failure is equivalent to being maximally fatigued[4]. It is almost always possible to drop the weight somewhat and continue performing repetitions, albeit with a lighter load.
Muscular failure is simply the point at which a level of fatigue has been reached that causes a reduction in force production to the extent that the current weight cannot be lifted any more times. Essentially, it is a marker that a certain level of fatigue has occurred.

Mechanical loading
Mechanical loading is the main stimulus for muscle growth. It is mostly applied actively, when the brain sends a signal to the muscle fibre. When the brain sends this signal, it activates motor units in the muscle. Each of these motor units control a group of dedicated muscle fibres. A muscle fibre will only be stimulated to increase in size after it has been recruited.

Muscle fibres are always recruited in size order. This phenomenon is known as Henneman’s size principle[5-8]. This principle states that smaller and weaker motor units are recruited before larger and stronger motor units. This progressive recruitment of small to large motor units happens in tandem with a neural signal that steadily increases in size. This neural signal will increase in size either because the force required of the muscle is high (as when a heavy weight is lifted) or because the muscle is fatiguing (as when training to muscular failure).

Full recruitment
Some researchers have assumed that muscular failure automatically means full motor unit recruitment has also occurred[9,10]. Obviously, full motor unit recruitment is a good thing, since it means that all the muscle fibres have been stimulated and can grow. This assumption led these researchers to develop the theory that low-load training to muscular failure was just as effective as high relative load training11. Other researchers disagreed and arguments ensued.

For several years now, the issue has been debated back and forth[12-14]. Unfortunately, this debate has occurred largely in the absence of investigations directly exploring whether muscular failure does in fact lead to full motor unit recruitment. We know that heavier loads cause full motor unit recruitment but precisely what happens during fatiguing contractions is much less clear. In any event, the research is currently too limited to say who is right.

Neural drive
Some recent studies have used electromyography (EMG) to measure neural drive and to assess whether muscular failure is synonymous with full motor unit recruitment. This is a difficult task, as EMG is just a measure of the size of the neural drive from the brain to the muscles over a set period of time.
The neural signal recorded by EMG is dependent upon both how many motor units have been recruited (motor unit recruitment) and how many times the signal is sent per second (motor unit firing frequency).

When many motor units are recruited or lots of signals are sent per second, the overall size of the EMG signal is very large. Unfortunately, we don’t know for certain whether a big EMG signal has been caused by greater motor unit recruitment or greater motor unit firing frequency.  Nevertheless, based on the use of EMG activity as a (very rough) proxy for motor unit recruitment, recent findings suggest that training to muscular failure with lighter loads is still somewhat less effective than training with heavy loads.

High versus low loads
For example, one study explored EMG activity during lateral raises performed with light loads (15RM) to muscular failure and with heavy loads (3RM) not to muscular failure. Interestingly, it found a plateau in EMG activity at 10 to 12 reps of the 15RM load and a lower level of EMG activity than the 3RM condition. This may mean that training to complete failure is not necessary to fully recruit the entire motor unit pool and may not be as good as heavy load training, at least in untrained individuals[15].

Additionally, several other researchers have compared the EMG activity with high and low relative loads, using the leg press16, knee extension without blood flow restriction[17], and knee extension with blood flow restriction (BFR)[18]. All of these studies found that high relative loads lead to greater EMG activity than low relative loads, even when going to muscular failure. Whether the addition of BFR has any effect is unclear. One study reported no differences in EMG activity between low relative loads performed to muscular failure with and without BFR[19].

Some researchers have interpreted these findings to mean that muscular failure is not synonymous with full motor unit recruitment. This is very hard to say purely from the EMG activity reports. It could be differences in motor unit firing frequency that are causing the differences. What we can see, however, is that heavy loads can achieve high EMG activity without close to muscular failure, while light loads probably need to go to muscular failure to get anywhere near the same level of neural drive (and even then they might not get there).

Metabolic stress
So if the recruitment of muscle fibres is not sufficient to explain the potentially superior results that come from training to failure, what can? In addition to mechanical loading, another key stimulus for hypertrophy is exercise-induced metabolic stress. The metabolites responsible for this stress are produced during exercise that relies heavily upon anaerobic glycolysis or from exercise performed where venous return is inhibited, such as BFR training. Such metabolites may be produced to a larger extent during periods of greater fatigue, which may in turn stimulate more muscle growth.

What it all means
The theories of how muscular failure might affect hypertrophy are all well and good but we all known that the rubber only meets the road when we look at the long-term studies. After all, over the years, more theories of hypertrophy have been rejected than have survived. So what do we find when we compare groups of trainees who are training to muscular failure with other groups of similar trainees who are not?
Only one long-term trial investigating the effects of proximity to muscular failure has been performed in resistance-trained subjects[20].

In this trial, one group performed repetitions to self-determined repetition maximum and another group performed repetitions to true muscular failure. The group that went all the way to true muscular failure achieved better results, both in terms of strength and size. Similar results have been found in studies investigating the effects of proximity to muscular failure in untrained subjects[21,22].

One of these studies used a very clever comparison where one group trained to muscular failure with a 10-rep set while the other group used the same 10-rep set but took a 30-second rest in the middle, to prevent them going to muscular failure. The other took a similar approach but used more inter-set rests. Both studies reported better results in the groups who trained to muscular failure, by some measures around twice the increase in muscular size!

Learn to fail
So however it happens, training closer to muscular failure seems very likely to be more effective than training further from muscular failure for hypertrophy. The training studies show this for certain.
We think it might happen because the fatigue causes more motor units to be recruited and there might be a beneficial effect of greater fatigue causing more metabolic stress. But either way, if you want to build bigger, stronger muscles, the one place you can afford to fail is the gym.
For all references cited in this article, go to strengthandconditioningresearch.com