From: Lyle McDonald Subject: Here's one for you to chew on for a while: Is Failure Necessary Date: Fri, 9 Feb 96 23:39:45 -0500 Ok, it's time for some more musings from the guy who never hangs out on m.f.w any more. So, here in all it's glory is hopefully the answer of the following question. -------------------- Why go to failure? The question of how much stimulation (and what type) is sufficient to cause maximal/optimal muscle growth is one that does not have an easy answer. Many groups and individuals feel that going to the point of momentary muscular failure (or beyond with certain techniques) is the key to causing muscular adaptation. That is, taking the muscle to the point of attempting the momentary impossible is the key ingredient to muscular failure. There have been several schools of thought as to why going to failure is necessary. One of those is the simple microtear theory wherby the muscle literally undergoes physical tearing. Various individuals feel that going to the point of ischemic rigor (where the muscle essentially locks up) causes minute tears to occur in the muscle during the eccentric phase (the lowering phase of a weight training movement) and that is the stimulus for growth. If true, this is incedentally why the eccentric point of the movement is both critically necessary for growth (for the most part) as well as the cause of the majority of muscle soreness. That is, since the tearing occurs during the eccentric portion, it seems reasonable to conclude that one must perform an accentuated negative to get increases in size and strength. Proponents of this theory offer as evidence the fact that muscles stressed in a concentric only method do not undergo growth consistently and that the eccentric portion of the movement has been shown to be the stimulus for growth. However, they also feel that negative only movements (which are often used to increase the intensity of training since more weight can be lifted with eccentric vs concentric contractions) do not work as well as combined concentric/eccentric lifting as the concentric is necessary to 'prime' the muscle in some way for the above mentioned microtears. Now we do know that heavy training (especially eccentric contractions) cause an increase in biochemical markers of muscle damage which lends some support to the theory that muscle damage is a key stimulus for growth. But, even this brings up the question of just how much muscle damage is needed to stimulate optimal growth. This is not a question that anyone is even close to answering at this point and I have a feeling that it may depend on the person and their genetics (which might explain why some individuals can grow from greater amounts of training while others overtrain with anything but the least amount). Now, at this point in time, there is not adequate data to say exactly what it is about lifting a weight X number of times that causes it to grow. Various other theories have been offered instead of the above including ATP depletion (which, at least during high intensity cycle ergometry has not been shown to occur), CP depletion (which, if correct would argue against creatine loading), decreased blood flow (which occurs as a result of near maximal muscular contractions which cause capillaries near the muscle to collapse), increased blood flow (i.e. the pump theory of growth), muscle ischemia (oxygen deprivation but we don't see huge muscles in individuals who spend lots of time at high altitude) and the simple tension/metabolic work theory (covered in great detail in a seminal review article by Goldspink et. al.) that argues that forcing the muscle to do high intensity work is the prime stimulus for growth. Now we also know that involuntary high intensity contractions (like with electrical simulation) does not cause growth except in very untrained or injured inviduals. So, not only does there seem to be a need to perform high intensity muscular work, it has to be generated by a person's own nervous system to be effective. Ok, so why failure? Is there anything special about going to muscular failure which might be the primary stimulus for growth. Other than the microtear theory which mandates failure so that the tears can occur, none of the above theories seem to require going to failure. And, it may be that tearing can occur without going to failure seeing as it does occur with downhill running (which forces the muscle to contract eccentrically as well). But, we know that long distance running doesn't spur muscle growth so there must be something else going on. Let's say you're lifting a load that puts you above the threshold to recruit 100% of your motor units (about 8RM for upper body movements and 15RM for lower body movements). And, let's further say that you are performing an upper body movement with your 8RM. Well, strict proponents of the failure theory would argue that you must perform 8 reps to achieve growth and that stopping short of this would not generate any growth. But, if you were to stop this set at 7 reps (knowing with 100% accuracy that it was your 8RM) you would achieve almost 100% of the (take your pick here) ATP depletion, CP depletion, decreased blood flow, increased blood flow, oxygen deprivation or time under tension. So, the question still remains: Why failure? Let's take as an assumption that the critical component to muscle growth is simply the time spent under high tension (supported by ample evidence as presented by Goldspink et. al.) and that other factors (those listed above as well as hormonal factors) are secondary in nature but may increase the adaptations seen. Several groups suggest specific set times like 60-90 seconds (HIT advocates although the times change >from source to source), 20-60 seconds (strength coach Charles Poliquin), Superslow (generally 60 seconds per set in 4 slow 15 second reps) which lends at least anecdotal evidence that some minimum time under high tension may be a pre-requisite to simulate size and strength increases. I don't think we can say with complete accuracy what that time is for optimal strength or size gains but let's take for granted now that some minimal time is necessary. Or, put a better way, slamming out 8 reps in 8 seconds with your 8RM will in all likelihood not achieve the same level (or type) of adaptation as doing 8 reps in 48 seconds with your 8RM. Although the rep count is the same, the total time under tension (and presumably other factors like ATP depletion et. al.) will not be the same. Ok, so still why failure? Assuming that stopping an 8RM set at 7 reps will achieve most of the time under tension that doing the final rep will, why push to 8 reps? I mean, that 8th rep hurts like hell and in the case of movements like squats and deadlifts, it may cause injury due to form breakdown so why not stop just short of that point if we can get similar results from it? Let me digress before I answer that question. Is there any evidence to the contrary in terms of the need for failure to spur muscle growth in either the scientific or anecdotal world? Yes, there is. We have at least one excellent example of how growth can occur without going to failure (or even including an eccentric motion in your lift). And that is the Olympic lifters. While many individuals will bitch and moan about how useless the Olympic lifts are and how dangerous they are, you cannot deny that they are some massively muscled individuals. Having seen Wes Barnett (one of the current US Heavyweight lifters), I can vouch for his extreme muscularity. Not that he's as big as even the smallest pro bodybuilder but he's built considerable muscle with the Olympic lifts. Now, Olympic lifters can't go to failure in their lifts as it will disrupt their technique. Also, the primary Olympic lifts (clean and jerk, snatch, etc) do not contain an eccentric movement. And, even on movements like squats and such, most Olympic lifters move rather quickly so there is no accentuated negative movement in their training. Now, I don't want to give everyone the impression that Olympic lifting is the most effective, most efficient, or safest way to get bigger muscles since I don't think it is. But, the fact that these individuals (who again lift very quickly, don't go to failure since it's not feasible with the types of lifting they are doing, and don't perform slow eccentric movements) show muscular hypertrophy throws a bit of a wrench in the simple theory of "You must go to the point of muscular failure in X seconds with a slow eccentric to achieve growth." To achieve optimal growth? Well, that's a different question entirely. Additionally, if you look at tradesmen who perform heavy manual labor, you often see large scale muscular hypertrophy caused by much lower than maximal work. However, their work requires large amounts of submaximal work (time under tension) which also seems to stimulate growth. So, we have at least two data points that show growth to occur without muscular failure occurring. And, any theory which can't adequately explain all data points needs to be revised. So, I'm revising it here. Now, there is one more interesting observation that can be made from Olympic lifters which is their generally large total training volumes (at least when compared to systems like HG and HIT and such). Since they rarely perform more than 3-5 reps per set and the reps are very short (less than 1 second generally), they tend to perform lots and lots of sets. We've all heard of the Bulgarian's training 3-6 times per day but each session was very short, these individuals were genetically superior, and they were most likely taking steroids so they are not the best example. But, at the Olympic training center, the Junior Olympic team lifters frequently train twice daily. So, although each set is minimal in length and there is no accentuated eccentric, it may be possible that these lifters make up for it with a large total time that their muscles are under tension. In any event, it does make quite a big hole in the theory that failure is the primary stimulus for growth since it's obviously not. The story, as they say, thickens. Now, I hate to bore you with this but Dave told me I better back up the above argument with some numbers rather than just give a hand-waving "OL's may perform similar amounts of total work" argument. So here goes but we have to make some major simplifying assumptions or the math will be impossible. Let's compare 'Typical HG Training' over the course of a year to a 'Typical OL Training' in terms of total time under tension. Let's assume a fairly advanced HG routine made up of squats, a pull and a push (ignoring abs and calves in terms of total training volume). Let me make a few simplyfying assumptions here: 1. All exercises are worked for 2 sets of 8. Yes, I know it's not the optimal range for legs and many won't do two work sets all cycle but let me go with it to simplify the calculations. 2. The individual trains twice per week all year round. Which is not going to happen with movements like squats if you're training hard as we all know. 3. The individual's rep speed is a constant 5 seconds total for concentric and eccentric (since I have never seen anyone really lift with a 2 up, 4 down as suggested by HIT. I lift slowly but I can't even do 2/4). 4. We ignore warmup sets below 70% in terms of total volume (i.e. most of them). 5. All sets are to failure. Again, this discounts the runup in most HG cycles but I don't want to deal with the math. So, You've got 3 exercises/workout *16 total reps/exercise which is 48 reps 240 seconds/workout. In 52 weeks, at two workouts/week, we have 104 workouts. So, total time under tension for this HG workout is 104 workouts * 240 sec/workout = 24,960 seconds per year that the muscles are kept under high tension (above 80% of max) which may stimulate growth. Ok, onto the OL's. A couple of simplifying assumptions (which are most likely incorrect). *BTW, the 10,000 rep number came from a friend of Dave's who says he uses that value with his high school athletes. I have no idea what kind of total volume Elite OL's actually use. The 10,000 value also does not include warmups below 70% of 1RM.* 1. Rep speed is 1 second for the Olympic lifts themselves and related movements (C&J, snatch, hang clean, power clean, jerk, etc.) 2. Rep speed is 2 seconds per rep for accessory stuff like squats, front squats, SLDL, overhead press, etc. 3. Total training volume is divided 50/50 between primary and accessory lifts. So, 10,000 reps of which 5,000 are primary at 1 sec/rep = 5,000 seconds. 5,000 reps of accessory reps at 2 sec/rep = 10,000 sec. Even still, this only total 15,000 total seconds of time under tension and even that's not accurate as all the time during the primary lifts is not spent with the muscles under tension due to momentum. But, I don't want to dig out my calculus book to figure this out with total accuracy. So, for a HG workout, we have 24,960 total seconds of work/year and for OL's we have 15,000 total seconds of work/year. Are these values close enough to give similar results? I would posit that the HG workout will give greater mass gains in the short run due to more time under tension. But, 15,000 secs is a lot of work no matter how you cut it and could explain why OL's, despite breaking all the other sacred rules of gaining muscle do so: their total time is fairly significant. Also, in all likelihood, the value for the HG trainee is overestimated (since it would be rare for a trainee to do 2 work sets every week of every cycle for a year) and the value for the OL is underestimated (if we assume that elite OL's, who probably aren't indicative of the average lifter anyway, would do more than 10,000 total reps per year). So, the numbers may be closer than they seem. Ok, back to the original gist of this article. I'm going to take the stance here that the primary stimulus for growth is the time under tension that a muscle undergoes (or total metabolic work performed or whatever. They are ultimately identical in concept but differ semantically) as I think it's best supported by the data. Hopefully I've made the point that failure is not the critical component to growth although it may be a component with the example of the Olympic lifters. So, why failure? Two reasons I think. #1: One of the big criticisms of periodization schemes is the rather large time spent working submaximally. I happen to agree with this criticism. Strictly periodized training programs (by that, I mean programs that lay out the weight, sets and reps an athlete is to do in advance) leave out one critical component which is daily variation. That is, let's say I've measured my 10RM in the squat some time ago and, based on that, I'm scheduled to do some percentage above or below it for a certain number of reps (based on the relationship between percentage of max and the number of reps one should be able to do). Well, what if I'm feeling really good one day and can get 12 reps with a weight I could previously only do 10? Or I'm feeling really bad and can only get 8 reps? I'm screwed is what. I'll either be working far below my maximal potential or forcing myself to work outside my current limits which could cause overtraining or, worse, injury. And, as much as I gripe about HIT, I think that the double progressive system has a lot going for it in that it avoids this problem entirely. If instead of saying I'm to do 10 reps and stop no matter what I set a range of 8-12 reps (or 4-8 or whatever rep range I'm in at the time), I can simply accomplish whatever I'm maximally capable of at the time ensuring I'm working at that intensity range (if desired) regardless of daily variations. This is how I periodize my programs by the way. HIT and periodization are not mutually exclusive. I might have clients of mine do the following: 12-15 reps double progressive to failure for 4 weeks or more 8-12 reps double progressive to failure for 4 weeks or more 4-8 reps double progressive to failure for 4 weeks or more rest and repeat. Low volume periodized HIT. And they said it couldn't be done. So, one good reason to work to at least concentric failure is to ensure that you are working at maximum capacity instead of at some pre- determined level which may or may not reflect daily variations. #2: Going to failure maximizes time under tension/total metabolic work which I argue is the primary stimulus for growth. When you consider that failure may occur at any one of 7 sites along the path from brain to muscle, I find it awfully naive to say that failure (whose cause we can't even identify and which may be different for different set times) is the primary stimulus for growth. Obviously, going to failure in 2 reps (about 10 seconds at 5sec/rep) will most likely have a different cause than going to failure in 15 reps (75 seconds assuming 5sec/rep). But, let's say we're working towards size gains and let's assume for now that 8-12 reps (40-100 seconds or so per set) is the optimal range (for the record, I don't think it is but that's another dicussion for another day). Going to failure within that range (irrespective of how we feel that day of training) will maximize the time spent under tension. It's great for me to say in theory that stopping a 10RM set at 9 reps will stimulate growth. I think it would. But, that's only useful if you know your 10RM weight on any given day. Since we can't know without testing it every single day to take into account variations in physical ability (and which would negate the whole point anyway since you've already done a set to failure) all we can do to maximize time under tension (also assuming here also that maximum time under tension = optimal time under tension) is take a given set to failure during any given training session. (Oh, before I forget, some individuals who are in very good touch with their muscles are able to pretty much know when they are getting close to failure. Considering that certain movements (deadlifts come jumping to mind) cannot be trained to complete failure safely, for those movements stopping the set just short of failure (i.e. if you 'know' that you've got only 2 more reps in you you just do 1 more) is probably not a bad idea and will still net you some growth. You might make up for the lack of failure by performing a second set (to increase total time under tension) or not, it just depends on the person, their genetics, etc.) or not. And, that, my friends, hopefully answers the question of "Why failure?" To be honest, I remain unconvinced that going to momentary muscular failure is the only way to get growth as we have examples of individuals who have rather large muscles without every going to failure. Might it be the most efficient way to get growth? That's a horse of a different color as the saying goes and not one I'm ready to tackle right now. But, it's certainly not the only way. Lyle McDonald, HB (which stands for Human Being which I am, you and everyone else are. So forget about putting all those letters after your name as they don't really impress anyone but you and your mom anyway.) P.S. Please send any comments on this piece via email a I don't log on here often. Thanx. From lylemcd@delphi.com Sun Feb 11 21:00:14 PST 1996 Article: 27562 of misc.fitness.weights Path: news.u.washington.edu!uw-beaver!uhog.mit.edu!news.kei.com!newsfeed.internetmci.com!news-feed.mci.newscorp.com!news.delphi.com!usenet From: Lyle McDonald Newsgroups: misc.fitness.weights Subject: Re: Here's one for you to chew on for a while: Is Failure Necessary Date: Sat, 10 Feb 96 17:04:52 -0500 Organization: Delphi (info@delphi.com email, 800-695-4005 voice) Lines: 84 Message-ID: References: <5XNpZBR.lylemcd@delphi.com> NNTP-Posting-Host: bos1d.delphi.com X-To: Lyle McDonald Well, thanks to some excellent comments I received via email, I realized a couple of major screwups in my argument regarding time under tension and failure, etc. So, here's an attempt to correct those screwups. #1: While I frequently used the term 'time under tension' in my article, it was semantically a bit incorrect. Instead, it would have been better to use 'Time under high tension'. That is, no amount of low tension work (like walking, etc) will spur muscle growth for various reasons. First and foremost of which is that you won't be recruiting the High Threshold Motor Units (Type II muscle fibers) below about 80% of max. Now I did mention that my assumption was that we were working above 80% of max for all my example but that was apparently lost somewhere in the discussion. I would probably set the cutoff for high tension to be 60% at the minimum and I think most will get better results using higher tensions. So, why not take this to it's ultimate extension and just work with 1RM loads or higher to subject the muscle to even higher tensions? See #2 below. #2: One individual made the comment that tension (and more specifically high tension overload) was the key stimulus. While I agree with the specific case, I must take issue with the first. Simply subjecting the muscle to high tension for an arbitrary time will not be sufficient to stimulate growth. That is, at the core of my argument was the idea that maximizing total time under high tension (or metabolic work, see correction #3) above 80% of maximal voluntary contraction (MVC) was the optimal way to get growth. I think anecdotally (and experimentally if you take the time to dig out the research) you will find greater results IN THE SHORT TERM using a weight that allows more work (say an 8RM which would allow 40 seconds of high tension time overload vs a 1RM which might allow 5 seconds of high tension time overload). Obviously in the long run (assuming progressive overload) either system (or a system of 20-50+ reps as used by Dr. Ken) will work. I guess it would be more accurate to use the area under the curve of muscle torque (which correlates with load lfited but takes into account biomechanical factors) vs. time (again, assuming a load equal to or greater than 80% of 1RM) would correlate most highly with growth. #3: A glaring oversight I made while talking about time under tension was in not mentioning the need for mechanical work (i.e. cross-bridge cycling in the muscle) to occur. We know that isometric work does not reliably increase muscle size compared to standard isotonic weight training. So, even though you could conceivably equal the same amount of time under tension with isometric work, the lack of cross bridge cycling would lead to poor results as far as growth were concerned. Additionally, studies into electrical muscle stimulation seem to point out the need for voluntary muscular contractions to stimulate growth. #4: In my attempt to argue time under tension over failure as the primary stimulus, I overlooked what is actually THE prime stimulus for any type of adaptation which is progressive overload. I assumed that was taken for granted but you know what happens when you assume thing. Obviously, regardless of how much total time under tension (or total time vs. muscle torque or however you want to describe it), if you don't progressively overload the system (by adding weight to the bar, or performing more reps, or moving more slowly, or doing more sets or whatever) you will not get further adaptation. Which brings me to another oversight. Reason #3 why going to failure in a double progressive system is useful: it gives you an easy barometer of when to add weight. Simply, if you are in a 4-8 rep range, you know it's time to add further tension overload when you can accomplish 8 reps. If you can only accomplish 4 reps with the new load (tension), you would then accomplish overload by attempting to add reps (or by doing another set but that's another complicated discussion) up until 8 and then add more weight. Again, in contrast to strict periodization schemes, this type of system makes it easier for the individual to know when to add weight to the bar which, in the long run, is the key to reaching your musular potential. Even those individuals in great touch with their muscles may not know when to increase the overload as it comes down to a subjective analysis of how difficult performing 7 reps at an 8RM load is or whatever. Some can get away with this but others cannot. Sorry about the mistakes in this one. And, thanks to the individuals who helped set me straight. And, if you're wondering who Dave is in the article, this was not originally intended to be posted to misc. fitness.weights but rather elsewhere. Dave is the guy (some of you know who I'm talking about) who keeps me straight and thinking critically about things so I don't write stupid things like the above. Lyle McDonald lylemcd@delphi.com From: James Krieger Subject: Re: Training to failure - why do it Date: Mon, 7 Oct 1996 17:33:36 GMT maxx@datasync.com (Maxx) writes: > James Krieger wrote: > >You're correct in that overload is one component of a successful > >training protocol. However, your overload analogies do not work > >when it comes to the human body. For example, how can you tell > >if you've overloaded a bone? It breaks. However, is this the optimal > >way that you try to make a bone bigger and stronger? No. Bone > > And that analogy is just as bad. Muscles are not the same as bone, so > should you train them the same? I'm using this analogy because, in general, all of the body's methods of adapting are basically the same. You provide the body with a stress, you give it some time to recover, and the body responds by adapting to better handle that stress. However, if the stress is too great, the body will not be able to handle it, and you will not get the effect you desire. Other examples other than my bone analogy: 1. Short periods of sun exposure will give you a tan. However, stay out in the sun too long, or go out in an area where the sun's rays are very intense (like near the equator), even for a short period of time, and you get a burn. 2. Sprinters often show large amounts of muscular hypertrophy. However, they never sprint to "failure", where they can't sprint any more. Many powerlifters and Olympic lifters also do not train to failure, and they show large amounts of muscle mass. What about Olympic gymnasts? They have huge upper bodies, and yet they don't do anything to failure. > > >Your definition of overload is "beyond what is currently > >capable." You can do 100 reps to failure with a light weight > >and you'll be going beyond what is currently capable. However, > >will you experience muscular hypertrophy? No. > > Yes, in fact you will. Will it be on the same scale? No..... and did > I say anything about number of reps? No, that is out of the > conversation. Besides, you will gain a significant amount of > endurance from that... Who cares about endurance? When you're interested in muscular hypertrophy, endurance is not what you're looking for. Anyway, 100-rep sets may cause a slight increase in the size of Type I fibers, but this difference would be barely noticeable. Also, endurance training can cause Type II to Type I fiber conversion, which is not something you want if you're looking to get bigger and stronger (I'm not saying that you're recommending 100-rep sets; I'm just making a comment here). So, even if you get a slight increase in Type I fiber size, you may get a decrease in overall muscle size since you can get the Type II to Type I conversion (Type II fibers are much larger than Type I fibers). > > >There is no scientific evidence out there that training to failure > >is better than not training to failure. Actually, a comparison > >of the two methods was recently done in a recent issue of the > >NSCA Strength and Conditioning Journal. Most of the studies > >out there have shown that training to failure may not be as > >beneficial as not training to failure. > > For one thing, there isn't that much science out there proving > ANYTHING other than "overtraining is bad, and explosive movements > cause most injuries." There's alot more science out there than you think. Read the journals; I'm not just talking NSCA; there's also the Journal of Applied Physiology, Sports Medicine, etc. Also, I'm not saying studies prove anything; they just show strong evidence pointing one way or the other. > > As for the NSCA thing, two problems: > > 1) Were both programs exactly the same except for failure? There's a > problem. Have you read the article I'm talking about? It wasn't a study. It was a review of many past studies done on the subject. > > 2) Since the NSCA doesn't like going to failure, why would they > publish info against what they recommend? Read the article. The NSCA is not against training to failure; it simply points out that it is a tool that should not be overused. No tool should be used for all applications. Read Charles Staley's home page for an excellent discussion of failure vs. nonfailure training. > > >Go back to my bone analogy. You don't need to break a bone > >to make it stronger. Also, if you were a distance > >runner, and you wanted to get better at distance running, would > >you run until you couldn't run anymore? Of course not. This is > >not the optimal way to train. > > Those analogies are just as bad as the ones you got me on. Are bones > muscles? Do you lungs work the same as the muscles? No, lungs don't work the same as muscles, but, again, I'm pointing out that in any type of exercise, pushing the body beyond its limits is usually not the optimum way to achieve adaptation. > > Of course not, these are all different areas of training. > > >I'm not saying training to failure is bad. It is simply a tool > >that should not be overused. I train to failure often and use > >it as one way of measuring my progress. It should not be, however, > >the main goal in one's training. > > I ask again, on a bicep curl WHY NOT GO TO FAILURE? Not the bone > analogy, that has no bearing on this, think of it in muscular terms. I am thinking of it in muscular terms. There are many reasons why you may not want to go to failure: 1. Going to failure will dramatically reduce training volume, and science has demonstrated a direct correlation between training volume and muscular hypertrophy. If one is doing multiple sets to failure, the weight will have to be significantly reduced on each set due to fatigue, reducing the amount of overload that can be placed upon a muscle for each successive set. 2. Fatigue is not the main component of the stimulation of muscular hypertrophy. The main mechanism behind tissue remodeling is the damage caused by the eccentric portion of a muscular contraction. Again, sets to failure will dramatically increase fatigue and cause weight reductions on sets, reducing the overload and eccentric-induced damage that can be inflicted upon muscle fibers. Some people may try to argue that muscle damage will increase when a muscle is fatigued. However, a study presented at the annual ACSM convention a few years back showed that muscle fatigue will actually reduce eccentric-induced damage on following sets. 3. Exercise physiologist Paul Ward states that training to failure will cause oxygen deprivation followed by oxygen perfusion. This results in extreme damage to cell membranes and DNA. 4. Heavy sets to failure can cause a dramatic drain upon the CNS. This dramatically increases recovery time and may reduce training efficiency. You asked for evidence why you should not go to failure, and I gave it to you. I have yet to see hard evidence that training to failure is more beneficial than not training to failure. Failure is NOT what causes hypertrophy.