To Fail or Not to Fail - 5x10 or 10x5? The Energetic Demand of Your Workouts Doesn't Depend on Workloads, Alone

Image 1: PCr, ATP, whatever as long as there was energy left, Arnold kept pumping iron (pun intended ;-)
"To fail or not to fail", this question is probably about as ancient as the hilarious idea to engage in physical activity that is not in one way or another directly related to one of our two most fundamental needs, survival and procreation. Researchers from the Physical Education Department at the Sport Sciences University of the Basque Country and the Department of Health Sciences at the University of Navarra in Spain have recently examined this question from a slightly different angle than most of the articles you have probably seen in and on the various muscle mags and bodybuilding related websites on the Internet. What Esteban M. Gorostiaga and his colleagues wanted to know was:

Are There Significant Differences in Energy Metabolism When you Train to Failure?

Or, put another way: Does it make a difference if you fail from a molecular energetic point of view or is the mere number of reps the most fundamental determinant of the changes in muscle adenine nucleotides, inosine 59-monophosphate (IMP), phosphocreatine (PCr), creatine (Cr), lactate and energy charge during a workout. To answer this question the researchers recruited 6 healthy male volunteers (age 28-40y; BMI 23.3kg/m²; 1-RMmax on unilateral leg press 199+/-43kg) and had them perform a total of 50 repetitions with the same initial load (83% of 1-RM) on two separate occasions, either
  • performed to failure, as a quintette of 5 x 10 (sets x reps), or
  • stopped before failure, in a 10 x 5 fashion.
On both occasions the subjects rested 2 minutes between the sets. Furthermore, Gorostiaga et al. tried to eliminate "confounding factors", by equating the values of several variables such as initial load and total number of repetitions between both exercise sessions and making sure that whenever a "subjects could not lift the initial load during the following sets due to fatigue" the load was decreased by 15kg until the respective subject was able to complete all 50 repetitions (Gorostiaga. 2012).
Figure 1: Peak power output profiles (average for n=6 subjects) for each exercise during the two experimental conditions: when exercise was 5 sets of 10 repetitions to failure (10REP; open circles), and when exercise was 10 sets of 5 repetitions not to failure (5REP; filled circles; adapted from Gorostiaga. 2012)
It is plain to see from the data in figure 1 that performing all sets to failure (open circles) lead to a significant reduction in total workload (the area under the peak powder curve):
During 5REP all the subjects were able to complete all the repetitions with the initially load assigned (154+/-31 Kg; 83+/-8% of 1RM). During 10REP, however, most of the participants were unable to complete all the repetitions with this starting load, due to failure. The load had to be reduced by 7.2+/-3.8% after 27+/-16 repetitions and was progressively reduced, reaching 85+/-12%(P,0.05) of the initial load at the last repetition. Average load during the 50 repetitions of 10REP was 6.1+/-6.3% lower (P<0.05) than during 5REP. (Gorostiaga. 2012)
If we examine the graph further there are a couple of other interesting things to observe, though:
  • the 2nd-3d rep was the one with the maximal power - so much about the value / validity of 1-RM maximum strength tests, then ;-)
  • the power progressively declined from the 3rd rep on (35-45%) and that with an astonishing dip after the 5th rep - maybe because subjects are used to do 5 reps, so that this could also be a psychological factor
  • while not training to failure with 10 sets of five reps allows to maintain almost identical average peak power on all sets, training to failure with 5 sets of 10 reps resulted in a net reduction of 33% from 812Watts on the first to 569 Watts on the last set
  • the average peak power per set was accordingly 28% lower, when the participants trained to failure
What's the "energy charge"? The energy charge was calculated as the quotient of (ATP + 1/2 ADP)/(ATP + ADP + AMP) and does thus quantify the ratio of usable to used energy in the muscle samples.
Yet while the average mean power output changes paralleled those of peak power output in both experimental conditions, the opposite was the case for the aforementioned muscle metabolites - muscle adenine nucleotides, inosine 59-monophosphate (IMP), phosphocreatine (PCr), creatine (Cr), lactate and energy charge - the scientists measured by high-performance liquid chromatography from the muscle biopsies they had taken from the right legs of the subjects on each occasion.

Is light training an option, at all?

Very much in accordance with the subjective experience of many trainees, the data in figure 2 appears to confirm that working out to failure does induce muscular exhaustion, which manifests in the form of physically quantifiable changes in muscle metabolites in the training to failure group
  • almost depleted PCr stores in the failure group (85% fall, P<0.05), and 
  • reduced ATP (-21%), energy charge (-4%), and  
  • reduced adenine nucleotides pool (-20%; ATP + ADP + AMP), in the presence of
  • increased IMP (+8600%) and lactate (+1400%) levels
Which stand in stark contrast to the mediocre decrease in of phosphocreatine, the almost unchanged muscle ATP, IMP, energy charge and adenine nucleotide pool and comparatively marginal elevations in blood lactate in the 10 x 5 non-failure group.
Figure 2: Changes in ATP, ADP, AMP, phosphocreatine (PCr) inosine monophosphate (IMP) and lactate from pre to post workout in the 5 x 10 (failure) and the 10 x 5 (no failure) session (data calculated based on Gorostiaga. 2012)
As far as the correlation between these markers and the actual power output during the workout are concerned, the scientist say that they observed a ...
Figure 3: Relationship between muscle lactate concentrations and average peak power (from Gorostiaga. 2012).
  • significant linear negative correlation (R²=0.59) was observed between the average changes in peak power output observed during the last two repetitions (expressed in percent of the initial two repetition values) and the decreases in ATP levels (expressed in percent of initial value).
  • significant curvilinear negative correlation between the average peak power output changes observed during the last two repetitions of the first and last sets (expressed in percent of the initial two repetition values) and the corresponding levels of muscle lactate. 
From the curvilinear nature of this relationship (see figure 3) Gorostiaga et al. conclude that "when muscle lactate levels do not exceed the upper limit of 10–15 mmol/kg wet muscle, power output changes little from maximum values". The exact opposite is yet the case, when the lactate values exceed this critical upper value and the power output begins to decrease sharply.
Image 2: While training like a sissy will at best produce suboptimal results, maxing out on every set of every workout will work for max. 2-3 weeks until you will not just lose the gains you may have made but end up weaker and with less muscle in the hospital, when an injury or total burn-out forced you to finally see reason. Going to failure on one the last set of a selected exercise for each body group may be a way smarter, safer and more productive way. Combine that with planned 3-RM tests to gauge your strength progress and you should see some nice gains and can keep track of your strength gains without risking burn-out or injury and trust me this has little to do with being a sissy!
Implications: At first sight you may certainly argue that the study does not provide much novel information. If you do however compare the main results to common wisdom about various strength training regimen, the total depletion of the phosphocreatine stores in the "higher" rep group and the increase in IMP levels, of which Gorostiaga et al. rightly argue that they reflect the failure of ATP resynthesis to match ATP hydrolysis rates and eventually feed into the uric acid cycle (as reflected by the 19% increase in the failure group) put an emphasis on the often underestimated energetic and metabolic demand of training to failure. In this context, the pronounced loss of purines from the muscle, as it has also been observed by Hellsten et al. subsequent to profoundly lowered ATP levels in the course of a one-legged HIIT protocol on a cycle ergometer (Hellsten 1999), and the subsequent extraction of urate from the blood by the muscle to restore intramuscular urate levels (remember: urate acts as a free-radical scavenger during intense exercise; cf. Hellsten. 1997) may well be an overlooked factor, when it comes to assessing exercise recovery.

Still, while the former would suggest that you better avoid training to failure altogether and simply hit the gym for a "light" 10 x 5 workout everyday, the minor reduction in PCr (-15% vs. -80%) and the non-existent rise in plasma lactate and urate (no stress = no adaptation?) do indicate that frequent light training session will probably not result in the desired, or at least suboptimal muscular adaptations, which are the physiological bases for the strength and size gains, you are looking for (read all about "The Physiology of Building Muscle"). Moreover, the almost unchanged ATP/ADP ratio (see figure 3), which is the gauge by which AMPK works (cf. "Zoning in on AMPK"), could be the reason why many of the "sissy workout" studies report that strength training would not have the same / any beneficial effect on glucose tolerance, lipid levels and all the other standard parameters of metabolic health scientists usually measure in those trials.

Bottom line: While it is almost certain that you will out-train your own recovery potential by going to failure on every set of every workout, the results of this study put an emphasis on the fundamental difference between physical workloads and their immediate physiological effects (just a reminder, the workload, i.e. weight lifted x reps was identical for both groups). What we are still lacking to derive concrete reliable workout tips from data like this, though, are clear-cut mechanistic or at least probabilistic relationships between the short term effects of and the long-term adaptation to different workout regimen and their respective energetic demands... ah, and by the way, this goes for the incredible popular measurements of post-workout protein synthesis, as well. Until now, no-one can say how much predictive value temporary increases in fractional muscle protein synthesis actually have in terms of long(er) term muscle gains.

References:
  • Gorostiaga EM, Navarro-Ame´zqueta I, Calbet JAL, Hellsten Y, Cusso R, et al. (2012) Energy Metabolism during Repeated Sets of Leg Press Exercise Leading to Failure or Not. PLoS ONE 7(7): e40621. 
  • Hellsten Y, Tullson PC, Richter EA, Bangsbo J. Oxidation of urate in human skeletal muscle during exercise. Free Radic Biol Med. 1997;22(1-2):169-74.
  • Hellsten Y, Sjodin B, Richter EA, Bangsbo J (1998) Urate uptake and lowered ATP levels in human muscle after high-intensity intermittent exercise.Am J Physiol 274: E600–E606.