Uncategorized

January 23, 2021

Do You Need to Train to Failure: Insights From Our New Meta-Analysis

Few topics in the field of exercise are as controversial as training to failure. Views on the topic tend to be polarizing, with some fitness pros strongly advocating the need to go all-out for optimal muscle adaptations, and others claiming failure training not only isn’t necessary, but in fact may be detrimental to gains.

Who’s right?

Our recent systematic review and meta-analysis provides some answers on the topic, while at the same time raising many more questions. The paper is open access and thus free for all to read, but I think it’s essential to delve beyond the numbers to fully appreciate its practical implications. Here’s the scoop…

What We Did
We searched the current literature to locate all randomized control trial studies that directly compared measures of strength and hypertrophy when carried out to muscle failure versus not to failure. Only human studies with healthy subjects that had a minimum duration of six weeks were considered for inclusion; we excluded studies that used blood flow restriction resistance training or concurrent training interventions (e.g., combined resistance and aerobic training). We then carried out a random effects meta-analysis that pooled results of all included studies to quantify the effects of failure training on muscular adaptations. A subgroup analysis of training status, body-region, exercise selection and training volume was performed to determine their potential influence on results.

What We Found
A total of 15 studies were identified that met inclusion criteria. A basic meta-analysis of pooled results found no statistical differences between training to failure versus stopping short of failure for both strength and hypertrophy outcomes; the trivial to small observed effect size differences between conditions in both outcomes (-0.09 and 0.22, respectively), suggest that any effects were of little practical meaningfulness.

Subgroup analysis showed a moderating effect of training volume on strength gains, whereby studies that did not equate volume favored non-failure training; the effect size differential was of a moderate magnitude (ES: –0.32). Alternatively, subgroup analysis found a moderating effect of training status on muscle growth, whereby trained individuals achieved a small hypertrophic benefit (ES = 0.15) from failure training.

What are the Practical Implications of Findings
On a general level, our meta-analysis indicates that training to failure isn’t necessary for maximizing muscular strength or hypertrophy. That said, numerous gaps in the literature preclude our ability to draw strong conclusions on the topic. The following points need to be considered when attempting to translate the research into individual program design:

• The choice doesn’t have to be binary: All failure-training studies to date have employed designs where one group trains to failure in every set while the other group does not train to failure in every set. This doesn’t necessarily reflect real-world programming. Fact is, you don’t have to take all sets to (or not to) failure. Training to failure on each set ultimately tends to compromise volume load, which in turn may impair hypertrophic adaptations. Moreover, there is some evidence that continually training to failure across multiple sets brings about markers of overtraining, which in turn may negatively impact muscle-building capacity. There are numerous strategies to employ failure training in a program. For example, you can perhaps limit its use to the last set or two of an exercise…perhaps use it selectively on certain exercises (see below)…perhaps reserve its use for higher rep sets (see below)…perhaps periodize its implementation across workouts or training cycles (see below)…the possibilities are almost endless. Thing is, no study has yet endeavored to study these possibilities, so all we have to go on at this point is anecdote and logical rationale.
• If not failure, then what is the appropriate set end point? : Assuming we take the results of the meta-analysis at face value and accept that training to failure isn’t obligatory for optimizing muscular adaptations, the ensuing question would be: “How close to failure do you need to train?” Unfortunately, there isn’t enough evidence to answer the question. Given that a muscle has to be sufficiently challenged to promote adaptation, we can logically make the case that at least some sets would need to be taken relatively close to failure. However, what value would that equate to in the repetitions in reserve (RIR) scale? An RIR of 1? An RIR of 2? An RIR of 3…or more? And consistent with what was mentioned in the above bullet point, how does this play out across multiple sets of an exercise? At this point, there is considerable room for debate on the topic.
• Does training frequency enter into the equation? : Although controlled evidence is lacking, it logically follows that training to failure increases recovery time between sessions. Assuming so, it makes sense that those training with higher session frequencies (>3 or 4 sessions per week) may not be able to tolerate as much (or any?) failure training. A case can be made that failure training not only has recovery implications for the targeted muscle groups, but on the neuromuscular system as a whole. At this point no research has endeavored to investigate the need to manage the level of effort based on how often you train.
• Is the need to train to failure load-dependent? : It has been speculated that you must train (closer) to failure when using higher rep schemes, both to recruit and simulate fast twitch fibers. Despite this speculation, there is a paucity of controlled research on the topic. The evidence we currently do have appears to support that higher rep training requires a higher level of effort, but whether all-out failure is obligatory remains somewhat equivocal. It is also important to note that failure in low repetition training is brought about by neuromuscular factors whereas failure in high-rep training is brought about by peripheral factors; are these factors associated with mechanisms that may elicit different hypertrophic responses and, if so, would failure be a modifying variable in the response? And given that higher rep sets involve a higher perception of discomfort, is true “failure” actually reached by most lifters when training with lighter loads before they simply give up due to the displeasure sensation? These questions require further investigation.
• What about training experience? : Our meta-regression showed that failure training was more beneficial in those with resistance training experience compared to novice trainees. However, there are a couple of caveats to this finding. For one, the overall magnitude of effect of was relatively small (ES = 0.15), calling into question the practical meaningfulness of the finding. Moreover, as noted in the exclusion criteria, we excluded a study by Carroll et al. (that employed trained lifters), due to the fact that it had an aerobic training component as part of the design (research indicates that aerobic training can interfere with muscular adaptations, and the degree of interference may be more pronounced in trained lifters). However, whether the aerobic training component actually influenced results remains unclear. Had this study been included, the result favoring failure training for trained lifters would have been nullified. I’d also note that a recently published study was just published showing no benefit to failure training in trained men; the study came out after publication of our paper and thus was not included for analysis, but certainly would have further reduced the observed effect. With all this said, no study to date has investigated failure training in highly trained lifters. It is conceivable that when lifters get increasingly closer to their genetic ceiling, a greater intensity of effort is required to achieve muscular gains. On the other hand, highly trained lifters also tend to be able to use heavier loads and are able to “dig deeper” to push the limits of failure training. Perhaps this means elite lifters should take fewer sets to failure because of the resultant neuromuscular stress on the body?
• Is age a consideration? : It is fairly well-established that recovery ability tends to decline as people age; all other things being equal, older lifters require more time to recuperate after a resistance training session compared to younger trainees. Given that failure training negatively impacts recovery, perhaps it should be employed more sparingly in this population? Unfortunately, there is scant research to date on the effects of failure training in older individuals, limiting our ability to draw strong conclusions on the topic.
• Does the type of exercise matter? : All exercises are not necessarily created equal when it comes to failure training. For example, taking sets of deadlifts or bent rows to failure can be highly taxing to the neuromuscular system. Alternatively, I’ve never heard anyone say they were crushed from going all-out on cable lateral raises. Single vs multi-joint…free weight vs machine…upper vs lower body…each of these variables concerning exercise selection requires consideration when deciding on the level of effort to expend. Unfortunately, the literature to date has not endeavored to investigate the complexities of this topic.

Take-Home Conclusions:

So where does this leave us from a practical standpoint? As with most applied exercise-related topics, research can only provide general guidelines into application for program design. Hence, here is my evidence-based take on the topic that synthesizes the current research in combination with insights from personal experience.

First, muscular adaptation requires a stimulus that challenges the body beyond its present capacity. In novice lifters, this can be achieved stopping quite a ways away from failure; even cardio is sufficient to cause appreciable muscle growth in this population! As you gain training experience, the need to train closer to failure becomes increasingly more important. Although it’s difficult to provide specifics, I’d say that at least some sets need to be within a rep or so of volitional failure. I’d also speculate that for highly trained lifters (e.g. competitive bodybuilders), there is a need to take some sets to failure to optimize muscle-building. Along these lines, as you get older, failure training should be employed more sparingly to allow for adequate recovery.

Second, when failure training is warranted, it should be applied somewhat conservatively, erring on the side of caution. A good rule-of-thumb is to limit its use to the last set of a given exercise; other sets should employ an RIR of 1 to 3. Moreover, you may need to further limit failure training with higher frequency routines. Periodizing failure training is a viable option, whereby more sets are carried out to failure prior to a peaking phase, potentially followed by a tapering phase. I’d note that numerous research studies show robust strength and hypertrophic gains when multiple sets are carried out to failure over short-term interventions (~8 to 10 weeks); however, continuing to train in this fashion likely will bring about negative consequences (i.e. overtraining). Thus, alternating periods using very high levels of effort with reduced levels of effort potentially may promote supercompensation of gains without devolving into an overtrained state.

Third, failure training should be prioritized in single-joint movements. These exercises induce less stress on the neuromuscular system, and thus don’t tax your recuperative abilities as much as multi-joint movements. Alternatively, limit the use of failure training on compound movements, particularly structural exercises using free weights (e.g. squats, bent rows, etc.). Machine-based exercises, in addition to being somewhat less taxing from a neuromuscular standpoint, provide a degree of safety when training to failure if you don’t have a spotter.

Finally and importantly, how all these considerations play out in practice will be specific to individual needs and abilities. Both genetic and lifestyle factors have a major role in program design. Ultimately, continued experimentation is required to optimize individual response over time.