Way back in the early 1970s Arthur Jones popularized the notion of training to failure with his series of articles in Iron Man magazine. Training to the point of muscular failure, Jones explained, was the necessary stimulus for maximum muscular growth. Mike Mentzer, a former Mr. Universe and founder of "Heavy Duty" training, was absolutely adament about it, repeatedly stating that if the muscle isn't pushed to the point of momentary concentric failure then no growth will be stimulated. Five-time Mr. Universe Bill Pearl, on the other hand, holds the conviction that one should NOT train to the point of failure. To Pearl's thinking, training to failure is not only unnecessary, it's counterproductive. Top Powerlifters seldom train to failure. Olympic Lifters rarely ever take sets to the point of failure. (Note: By failure here I mean momentary concentric failure, i.e. the inability to complete another full repetition of the concentric phase of the lift - you could, however, continue to do static holds and negatives). Some people have even advocated doing negative-only sets to the point of momentary eccentric failure (the inability to complete another full repetition of the eccentric phase of the lift - you are unable to stop the bar from crashing down on you).
With all of these experts disagreeing with each other, and many of them having very impressive credentials, how can one know who or what to believe? That's what I'm going to attempt to calrify here. If you haven't read the Neuromuscular System series and the article entitled Muscular Fatigue During Weight Training on the 'Physiology Related Articles' page, then I suggest that now would be a good time to have a look at them. A great deal of the knowledge that you need to analyze many weight training approaches, including the practice of training to failure, is there. Let's have a look at those approaches with both neuromuscular physiology and experience in mind.
As already mentioned, some high intensity training advocates have stated, unequivocally, that if you don't train to momentary concentric muscular failure then you will not grow. That's a pretty bold statement. The logic goes like this:
The logic seems bullet-proof. But you really don't have to look very far to dispell it. Top powerlifters, Olympic-style weightlifters and many bodybuilders rarely, if ever, train to momentary concentric muscular failure, yet I probably don't have to tell you that they haven't had a problem with realizing muscular growth and/or strength increases. I recall reading an interview with Powerlifting legend Ed Coan from about fifteen years ago in which he stated that he never went to failure on any of his sets. Bill Pearl says the same thing about his own training. "But they were on steroids", some of you will say. Well, of course they were. But most pre-steroid era bodybuilders didn't train to failure and they never had a problem with muscular growth either. "But they weren't that big," some more of you will say. That's precisely because they weren't on steroids. As most people can appreciate, the drug-created monsters that now call themselves bodybuilders have raised people's definition of 'heavily-muscled' to the point where any man less than 250 pounds with 4% bodyfat is small and fat. If you really think that men such as George Eiferman, Reg Park, John Grimek and Steve Reeves weren't that big, maybe you should see them standing next to 'normal' men, or in more normal circumstances than oiled up on a posing dias. Take a look. If that doesn't convince you, compare your own overhead lifts to what the Olympic lifters were doing years before the introduction of anabolic steroids. 180 pound weightlifters were routinely pressing well over 300 pounds overhead in the early 1950s! The level of strength that these men posessed was developed without steroids and without training to failure. The success of these people in building muscle, power and strength while not training to failure proves that such training is not necessary (at least for some) to realize muscular conditioning and growth.
Incidently, the pre-steroid era bodybuilders, statistically, carried just as much muscle as modern drug-free champions - sometimes even more - though they usually did not compete at such low body fat levels (see Your Muscular Potential: How to Predict your Maximum Muscular Bodyweight and Measurements). So if training to failure is now in vogue amongst drug-free bodybuilders, then the practice does not appear to be succeeding at building significantly bigger muscles than the pre-steroid era bodybuilders who rarely trained to failure.
But what works best for one person might not work best for all. Perhaps if Reg Park had routinely trained to concentric failure he would have been even bigger. Perhaps Jon Harris (the 2006 WNBF World Champion) wouldn't be nearly as big if he didn't train to failure. Perhaps it's the other way around. So the question is clearly not whether training to momentary concentric failure is absolutely necessary, but rather is it the most effective way for you to weight train.
Physiologically, we need to consider what happens when a weight training movement is taken to failure. Muscles fail because they're firing patterns can no longer produce enough force to continue the activity. Taking a segment from the article The Neuromuscular System Part I: What A Weight Trainer Needs To Know About Muscle:
Since productive (not rehabilitative) weight training involves lifting weights that require the firing of the type I, IIA and type IIB fibers, failure will occur when the highest threshold fibers fatigue. When a heavy load is lifted, first the lower threshold fibers are recruited, then the higher threshold fibers are recuited in increasing numbers until the weight is lifted. If all available fibers are recruited yet the resistance not overcome, then force production can be increased by increased rate coding frequency (the fibers twitch faster to develop more force) up until the point of maximum rate coding, at which failure occurs. Since the highest threshold fibers have the poorest endurance characteristics, failure will occur when these fibers fatigue (because the weight could not be lifted without them). Even if the weight isn't initially heavy enough to recruit the highest threshold fibers, as the lower threshold fibers fatigue the higher threshold ones are gradually recruited to compensate for the fatiguing lower threshold fibers. It is the high threshold fibers that have the most potential for growth.
This simple examination of muscle fiber recruitment and fatigue patterns shows that by taking sets to failure you are exhausting more muscle fibers than if you stopped the set short of failure. This is strong support for the practice of training to the point of momentary concentric failure - assuming that fiber fatigue is indeed the stimulus for growth.
So, what exactly is muscle fiber fatigue? The contributing factors to fiber fatigue were covered extensively in the Muscular Fatigue During Weight Training article and somewhat in the Neuromuscular System series on the 'Physiology Related Articles' page. Reviewing some information from those sources we have:
From the phosphagen system:
And during the anaerobic glycolysis mechanism:
As explained in the Muscle Growth series, the stimulus for muscle growth is complex and multi-faceted. A major contributing factor, however, is believed to be myofibrillar damage done as a result of insufficient cycling of actin-myosin cross-bridges. When cross-bridges cannot be released and formed in a sufficiently timely manner and sequence, trauma is experienced at some of the cross-bridge sites which cannot release properly and are 'torn' under the tension being produced by the fiber. This is thought to be a major stimulus for the growth process. All of the above factors of fatigue result in cross-bridge cycling impairment and, therefore, result in growth stimulus via the same mechanism.
Training to failure results in more fiber fatigue and, therefore, more micro-trauma within the fiber than stopping sets short of failure. Logically, this could be extended to conclude that training to failure also results in a greater growth stimulus.
However, consider some research involving heavy eccentric contractions that has shown that negatives (eccentrics) produce more microtrauma to muscle fibers than concentrics or isometrics. This occurs because fewer total fibers are recruited during the eccentric portion of a lift than during the concentric phase. Fewer fibers doing the job mean more tension is developed in each fiber and, therefore, more damage is sustained by each individual fiber. Research has indicated that this does not necessarily translate into accelerated growth, however, as programs consisting of negative repetitions have failed to produce more muscle growth than other types of training programs. As was covered in the Muscle Growth series, muscle damage and muscle recovery and supercompensation, though intertwined, are different processes. High levels of microtrauma (as caused by strong eccentric contractions) are known to interfere with glycogen replenishment and other metabolic processes in muscle after training - this may blunt the growth process. Clearly, greater microtrauma does not necessarily equate to greater growth.
Before you decide to try to minimize the negative portions of your lifts, however, bear in mind that many other studies have indicated that the negative phase is, in fact, the most important phase of the lift for stimulating hypertrophy (growth). The lesson to be learned is that negative-accentuated training will stimulate growth (perhaps moreso than any other type of training) but the degradative processes may outweigh the growth processes and because of the level of damage they do, negative-emphasis reps may, very likely, impose a longer recovery period.
Finally, within the fibers themselves there is point at which maximum tension and work occurs. This has been found to be not at the failure point, where it can be assumed that maximum rate coding of unexhausted motor units momentarily occurs, but rather at the point of about 90% rate coding. Since weights above 80% of 1-rep maximum (1RM) recruit practically all available muscle fibers in most muscle groups, and strength decreases by roughly 3% per rep with weights above 80% of 1RM (for most individuals), then we can estimate that this point of maximum tension and work (on a per fiber level) occurs for a majority of the recruited fibers at roughly the third rep away from failure - perhaps an oversimplification but good enough for a working argument. In other words, if you did a set of 8 reps to failure, maximum tension and work would occur for most fibers somewhere around the fifth rep. The question then begs to be asked, are the extra three reps needed to reach failure really necessary - particularly if maximum strength increase is the goal? Or is it sufficient to stop sets with one to three possible reps able to be performed? Alternatively, if one could perform a certain number of reps (to failure) with a certain weight, is it sufficient to use only 90% of that weight for the same number of reps? Because strength gains are largely a result of increased rate coding "efficiency", neuromuscular physiology hints that this may, in fact, be true. It is also in accordance with the majority of Olympic Weightlifting programs which traditionally avoid haphazard training to failure, and also recent research into strength training and periodization. Very few, if any, of the strongest drug-free Powerlifters and Weightlifters in history have advocated regular training to failure.
As was covered in the Muscle Growth series, the anabolic process relies on several major hormones and prostaglandins. Resistance training has been shown to have a profound effect on these substances. Several studies have indicated, rather conclusively, that one-set-to-failure training programs do not result in as great a training induced increase in anabolic hormones (testosterone, growth hormone and IGF-1) as do programs consisting of several sets of each exercise. Regular training to failure with multiple sets per exercise does appear, however, to increase resting cortisol levels, whereas training shy of failure appears to decrease it.
Although some would argue that this hormone release is inconsequential, assuming the body does it for some physiological reason (which is my belief) it can be concluded that the optimal approach is certainly to perform several sets of each exercise. The remaining question is whether any, or all, of those sets should be taken to the point of failure.
In any case, many studies have shown that the growth process inside trained muscle is completed within 36-48 hours of even very intense, high volume bodybuilding style training. This means that if longer rest periods are necessary between body part training sessions, then it is most likely not the muscles that require the additional recovery time...
It was covered in the Neuromuscular System series that contracting a muscle involves more than just what occurs in the muscle itself. The nervous system is intimately involved in the process. Taking another few lines from that series:
So, extending this, as the muscle fibers fatigue, and you reach the point of failure, the nervous system will recruit all available motor units and fire them as frequently as is possible at that moment (maximum rate coding). It is a well-established fact, though, that as a maximum muscular contraction continues, the frequency of motor units firing decreases. In fact, one study showed that by the end of a 30 second maximum voluntary contraction the firing frequency decreased by 80%. Eventually the frequency of twitching of the high threshold fibers becomes insufficient to sustain the effort.
We know that each neuron must release the neurotransmitter acetylcholine (ACh) every time that it fires (or 'twitches') a motor unit. We also know that the neurons transmit impulses down the length of their axons by way of Sodium/Potassium transport and the Sodium/Potassium ATPase Pump. The signal is carried across the membrane of the muscle cell in the same manner. Inside the cell, calcium is released via the T system to facilitate contraction. (For my purposes I am including the T system as part of the peripheral nervous system.) The process relies heavily on optimum calcium levels and transport, and enzymes that are involved in the synthesis and breakdown of acetylcholine and numerous other substances. The frequency of motor unit firing decreases, therefore, as these substrates are inadequately cycled - yet as failure approaches we continue our maximal effort to lift the weight. The effect that such an effort has on the neuromuscular system must be considered.
During the 1960s Dr. John Ziegler (York weightlifting team physician and co-developer of the steroid Dianabol) designed a machine that he used to monitor overtraining by sending electric currents through muscle. The 'Isotron', as he called it (cheesy '60s name), would be used to induce a muscular contraction by supplying a small electrical impulse to the muscle being tested. It was found that an overtrained or recently trained muscle would require a higher current than a rested muscle for 'strong' contraction to be achieved. He used this to determine when a lifter was ready to train again. What does this tell us? It tells us that for a period after training a higher than normal activation threshold is needed to produce contraction.
Incidently, ~75 mA was the 'normal' current required to produce 'strong' contraction. Anything over ~100 mA was considered indicative of overtraining. You may also be wondering how accurate this is given the fact that type II fibers naturally have higher activation thresholds than type Is. Well, when it comes to external stimulation (such as the kind the Isotron applied) the type II fibers are actually easier to induce a contraction in than the type Is because of their closer proximity to the surface of the muscle.
Regardless of all this, and whether signal transmission at the neuron, sarcolemma or T-tubules is responsible for the effects, this clearly illustrates that the peripheral nervous system (PNS) requires its own recovery period after training. And logically, training to failure would impose greater stresses on the peripheral nervous system and extend its recovery period longer than training short of failure.
Our nervous system arguments up to now have focused on the peripheral nervous system. But, as any experienced coach can tell you, the central nervous system has a large bearing on the failure point and the overtraining phenomenon. Taking another segment from the Muscular Fatigue During Weight Training article:
In order for a muscle fiber to twitch the central nervous system (CNS) must send a nerve impulse to the controlling motor unit. The innervating nerve cannot maintain its capacity to transmit this signal, with optimum frequency, speed and power for extended periods of time. Eventually concentrations of substrates such as sodium, potassium, calcium, neurotransmitters, enzymes, etc. may decrease to the point where muscle contraction becomes markedly slower and weaker. If high discharge rates are continued the nerve cell will assume a state of inhibition to protect itself from further stimuli. The force of contraction, therefore, is directly related to the frequency, speed and power of the electrical 'signal' sent by the CNS.
This is reflected in the fact that a trainee's motivation and emotional state can profoundly affect the discharge characteristics of the central nervous system ...though it is far from understood on a physiological basis. It is clear, however, that the central nervous system can play a pivotal role in the perception and reality of fatigue.
Furthermore, it must be considered that as the signalling of the PNS becomes impaired as muscular effort continues, increasing stress is imposed on the CNS in order to maintain the rate coding necessary to maintain sufficient muscle force production. As the PNS 'fatigues' the CNS burden increases, leading to accelerated CNS fatigue as well.
After training, during the recovery period of the muscles and PNS, intense muscular contraction imposes an increased burden on the CNS to overcome the impairment of muscle contraction caused by these insufficiently recuperated systems.
If these concepts seem vague, think of a lifter "psyching up" for a big lift, or remember some time when you thought that you couldn't possibly get another rep, but somehow managed to "dig deep" and force another one out. Both of those situations illustrate the manipulation of the central nervous system in order to allow the lifter to be stronger. Any experienced coach will tell you, however, that you shouldn't "psyche up" all the time or you'll "burn yourself out". The "old-timers" referred to this as using up too much "nervous energy". However you want to look at it, training too intensely, too often, will certainly lead to nervous system inhibition. When that happens you can forget about making good progress until you take enough of a break to allow for nervous system recovery.
NOTE: From extensive empirical evidence it can be concluded that training to failure with low reps and heavy weights is much more taxing on the central nervous system than training to failure with high reps and lighter weights. This is, most likely, due to the fact that heavy weights require the simultaneous recruitment and maximal rate coding of all available motor units (the muscle fiber and its innervating neuron). In addition, heavy loads tend to stress and deform connective tissues and joint capsules. Proprioceptors in the joint capsules relay information about joint positioning to the central nervous system. If the integrity of the joint is compromised, even slightly, the central nervous system will not allow the muscles acting on and around that joint to be recruited at full force. Maximal contractions in these muscle groups will not be "permitted" again until full joint recovery is achieved. Heavy, low-rep training therefore requires recovery periods for both the central nervous system (especially if training is taken to the failure point) and the joint structures. It has been shown that the muscles of the lower back may need up to one month of recovery time after maximum efforts before full force contractions can be achieved again. Keep this in mind when designing training programs.
As anyone who practices these lifts knows, they are extremely complex, high-skill movements. Muscular and neuromuscular fatigue quickly causes a deterioration of form on these complex lifts, so sets of the Weightlifting-style lifts should not be deliberately trained to failure. In fact, it is very rare for Olympic weightlifters to train the Olympic lifts to failure unless, of course, they miss a lift attempt. Also, because fatigue causes a deterioration of technique with these lifts, reps are kept low - typically 3 or less.
For someone who wishes to practice these lifts (or, more likely, their "power" versions) for strength development or athletic improvement, it still doesn't make sense to practice higher reps, as the very nature of these lifts require activation of the fastest of the fast twitch fibers. These fibers are, by nature, quickly fatigued. Don't forget that even the simpler "power" versions of these lifts (the Power Clean, Power Jerk, Power Snatch), or even High Pulls, still qualify as high-skill movements and, therefore, are susceptible to form deterioration with fatigue. Slightly higher reps than with the full Olympic lifts may be employed though - up to 5 reps - but they should not be trained to failure.
Clearly, training to failure imposes a longer recovery period than an otherwise identical routine but with sets stopped short of failure. Therefore, if a person choses to train to failure then training must be done less frequently than if the person did not train to failure. The question to be answered is whether it is more productive, from a muscle growth perspective, to train to failure infrequently, or to train short of failure but more often. Herein lies the difference between the two approaches.
In my experience, how a trainee reacts to specific training protocols is strongly influenced by body type:
Clearly, the effects of training to failure and personal recovery patterns have to be considered and monitored when a training approach is adopted.