From Lyle McDonald's site:
West et. al. Elevations in ostensibly anabolic hormones with resistance exercise enhance neither training-induced muscle hypertrophy nor strength of the elbow flexors. J Appl Physiol. 2009 Nov 12.
The aim of our study was to determine whether resistance exercise-induced elevations in endogenous hormones enhance muscle strength and hypertrophy with training. Twelve healthy young men (21.8 +/- 1.2 y, BMI = 23.1 +/- 0.6 kg(.)m(-2)) independently trained their elbow flexors for 15 weeks on separate days and under different hormonal milieu. In one training condition, participants performed isolated arm curl exercise designed to maintain basal hormone concentrations (low hormone, LH); in the other training condition, participants performed identical arm exercise to the LH condition followed immediately by a high volume of leg resistance exercise to elicit a large increase in endogenous hormones (High Hormone, HH). There was no elevation in serum growth hormone (GH), insulin-like growth factor (IGF-1) or testosterone after the LH protocol, but significant (P < 0.001) elevations in these hormones immediately and 15 and 30 min after the HH protocol. The hormone responses elicited by each respective exercise protocol late in the training period were similar to the response elicited early in the training period indicating that a divergent post-exercise hormone response was maintained over the training period. Muscle cross-sectional area increased by 12% in LH and 10% in HH (P < 0.001) with no difference between conditions (condition x training interaction, P = 0.25). Similarly, type I (P < 0.01) and type II (P < 0.001) muscle fiber CSA increased with training with no effect of hormone elevation in the HH condition. Strength increased in both arms but the increase was not different between the LH and HH conditions. We conclude that exposure of loaded muscle to acute exercise-induced elevations in endogenous anabolic hormones enhances neither muscle hypertrophy nor strength with resistance training in young men. Key words: testosterone, growth hormone, IGF-1, anabolism.
My Comments: For several decades now, there has been intense focus on the acute hormonal response to training. This started back in the 80’s where researchers, interested in growth did a rather cursory examination of elite powerlifters and bodybuilders, made some assumptions about muscle size, made some even bigger assumptions about how they trained, and then proceeded to reach some staggeringly poor conclusions.
Basically, what they observed was that bodybuilders were bigger than powerlifters, which is debatable in the first place. They also observed that powerlifters typically used low reps and long rest periods and bodybuilders (remember: this was the Arnold era) trained with high reps and short rest periods. Thus they concluded that high reps and short rest stimulated muscle growth and went looking for reasons why this was the case. I’d note that this is not really how you’re supposed to do science: you don’t reach your conclusion and go find reasons why it’s right. You test hypotheses and draw your conclusions from that. But I digress.
And the main focus for a while was potential differences in hormonal response to training, primarily focusing on testosterone and growth hormone (GH). The basic study design that was followed was to compare the acute hormonal response to either 3 sets of 5 repetitions with a long rest interval (3 minutes) to sets of 10 with a 1 minute rest interval. Repeatedly, studies showed that the first type of training boosted testosterone and the second GH. Entire training schemes have grown out of this but there was a problem: nobody ever bothered to see if these acute (usually less than 10-15 minute) bumps in hormones actually did anything.
Nevermind that this makes little sense anyhow for a variety of reasons. Not the least of which is that women have higher GH levels than men and get a bigger GH response to training, yet they don’t grow better. If anything, with the known impact of testosterone on muscle growth, if there was to be any benefit to this, you’d expect the lower rep/heavy work to be superior. Yet the researchers were arguing that it wasn’t. There was a logic missing in the argument (not the least of which being the assumption that powerlifters had smaller muscles than bodybuilders) that seemed to get skipped over.
In addition to the science, there is a long held belief, echoed in various places (including the comments section of another contentious article I wrote titled Squats vs. Leg Press for Big Legs) that certain movements, notably squats and deadlifts, will have full-body growth stimulating properties, generally mediated through the hormonal response.
It’s not uncommon to see people recommending things like “If you want big arms, squat/train legs.” for example. Essentially, heavy leg work is touted as being the key to overall growth. Nevermind that the same people who make this argument will often complain about “All those guys in the gym with huge upper bodies and no legs” without realizing that the two ideas contradict one another (that is, if leg training is required for growth, how can guys get huge upper bodies without training legs). But I digress again.
In any case, this study examined the issue directly with a somewhat confusing study design: twelve healthy young men trained their biceps on different days of the week under different training conditions. In what they called the low-hormone condition, the biceps were trained all by themselves; no other exercise was done. In the other called the high-hormone condition, the biceps were trained and then a large-volume of leg training was done to elevate the supposedly anabolic hormones.
Does that make sense, all subjects trained both arms, but on different days and under different conditions. And the training was far enough apart that the hormonal response from the leg training wouldn’t have impacted the low-hormone training session. This training was followed for 15 weeks and subjects consumed protein both before and after the training (so there was nutritional support).
Hormone levels were measured and while there was no significant change in hormones in the low-hormone situation, in the high-hormone situation, there were increases in lactate, growth hormone, free and total testosterone and IGF-1 with the peak occurring approximately 15 minutes after the leg work.
And, if the hormonal response to heavy leg training actually has any impact, what you’d expect to see is that one arm, the one trained along with the leg training, would grow better.
Did it happen? Guess.
Both maximal strength and muscle cross sectional area increased identically in both arms to the tune of a 20% vs. 19% increase in strength for low- vs. high-hormones and an increase in skeletal muscle cross sectional area of 12% vs. 10% in low- vs. high-hormones. These differences were not statistically significant. Quoting the researchers:
Despite vast differences in hormone availability in the immediate post- exercise period, we found no differences in the increases in strength or hypertrophy in muscle exercised under low or high hormone conditions after 15 weeks of resistance training. These findings are in agreement with our hypothesis and previous work showing that exercise-induced hormone elevations do not stimulate myofibrillar protein synthesis (36) and are not necessary for hypertrophy (37). Thus, our data ((36) and present observations), when viewed collectively, lead us to conclude that local mechanisms are of far greater relevance in regulating muscle protein accretion occurring with resistance training, and that acute changes in hormones, such as GH, IGF-1, and testosterone, do not predict or in any way reflect a capacity for hypertrophy.
I don’t think it gets any clearer than that and I’d note that another recent study titled “Resistance exercise-induced increases in putative anabolic hormones do not enhance muscle protein synthesis or intracellular signalling in young men.” by the same group found the exact same thing.
Summing Up: Leg training has no magic impact on overall growth, most of which is determined locally (through mechanisms of tension and fatigue mediated by changes in local muscular metabolism). If you want big arms, train arms. If you want big legs, train legs.
And if folks are wondering why empirically ‘folks who train legs hard’ seem to get big compared to those who don’t, I’d offer the following explanation: folks willing to toil on heavy leg work work hard. Folks too lazy to train legs hard often don’t. And it’s the overall intensity of the training that is causing the difference, not the presence or absence of squats per se. Which is why guys who only hammer pecs and guns get big pecs and guns even if they couldn’t find the squat rack in the gym: the small acute hormonal responses to training are simply irrelevant to overall growth.
A lot of this makes sense. In the Q&A below the post he brought up some good points about those with lower/upper dominant physiques including athletes in sports like speed skating or wheelchair bodybuilders.
I know that this goes against a lot of the advice posted here on WBB, and regardless of the validity of the article I would never cut compound lifts out of your regimen. This is especially true if you are looking for athletic performance, functional strength, or overall mass (size).
VERY good article!!
Bench geared 525 @ 210
Deadlift geared 650 @ 220
Captains of Crush #3
Building Mighty Mitts...
12 to me seems like a very bad sample size. I don't know about the factors that trigger growth, but I will definitely won't drop the squats. I don't know if all of this has been mental, or motivation through the roof, but my bench went up since I started squatting. Maybe like you say Tom, we just train harder than the average and that's why we start seeing results when we force ourselves to work hard---squats and deadlifts (which with no doubt are damn taxing).
I conclusion, 12 seems a very bad sample size, plus I believe that if you do compound lifts, you should never forget isolation completely. For me, isolation and compound lifts work wonders ;O)
-█--------█- Squat: 285 x 3
......\☻/...... Bench Press: 235 x 3 (stuck there)
........▌....... Skull Crusher: 110lbs ez bar
......./ \......Standing Shoulder Press: 115lbs (barbell)
"I am always doing things I can't do, that's how I get to do them." Pablo Picasso
It's not the hormonal response of squats. I agree with that. It's been seen in several studies that the hormonal response from squats is negligible. So I don't think that big arms are a function of squats producing more hormones.
I think that (similar to what Lyle says) is that those that train legs train harder but I also think that the body responds to things as a whole or a unit. It isn't going to let one arm get that much bigger than the other and it's not going to let the upper that much more disproportionate from the lower. Studies have been done on subjects with broken limbs. Those that trained the good limb saw less atrophy in the broken limb than those that didn't train. The body will respond better if everything is in place.
If you are only training arms, you'll lose the muscle gain from else where. Thus limiting your potential. Training everything allows the body to respond better. If the arms aren't trained, I'm sure you'd lose potential of possible muscle mass gained in the lower.
What is elite?
"Those who work the hardest often complain the least." -anonymous
Pain is inevitable. Suffering is optional.
This makes a lot of sense. I never quite bought the squat/deadlift heavy would increase chest/arm size but I never really said anything because everyone was for the idea. Especially considering that if these squats/deadlifts were necessary for overall growth (including arms), how can people get big arms and chests without these lifts?
I also remember asking my anatomy teacher who was a weightlifting coach (not one of those dumb ones) if the body liked to grow as a unit, and he said no. He said you can have big arms and little legs just as long as you train arms and not your legs. If I asked him if squats/deadlifts increased muscle gain in the areas they don't directly work, i'm pretty sure he would have said no.
They're still important though, obviously for the growth of the muscles they work/Glutes/Quads/Hams
Then: 134 lbs November 12, 2008
Now: 195 lbs
In my personal experience, after adding more intensity/frequency to my squats/DLs (following SS) my size, weight and strength shot up dramatically.
I was working hard on all my lifts and eating more. This may have been the main contributing factor. However, it's hard for me to dismiss the correlation between gains in strength/size when I added heavy squats/DLs.
With this study, what lower body HH inducing lifts were they doing?
There's a big difference between heavy back squats/DLs and leg presses. Namely, the spine and its supporting structures are dramatically more loaded with the squats/dls.
This difference (the spinal loading) in the exercises may be significant.
From my guess, a stronger core/spine will allow you to lift more weight in the lifts that transfer energy through it. Therefore, allowing you to load certain exercises heavier, upping the stress and in turn upping the growth/strength adaptation.
Furthermore, the acute response to the GHs may not have had much of an anabolic effect on its each individual session... But did they test to see how a chronic elevation of these hormones (3+ x per week) affected hypertrophy/strength?
It also seems unclear if they separated the group in half, and had each group doing the LH work and the other group doing the HH work...
I may be mistaken, but it seems they did LH work with 1 single arm on one day, then HH work on the other arm and the other day.
If the experiment was conducted that way, there are many holes/unknown variables in the assumptions made.
I hope all that made sense.
Here's my theory, as unscientific as it is:
squatting or deadlifting heavy leads to increased appetite. Increased appetite leads to increased intake of food. The increased intake of food leads to the gain of bodyweight. The increased bodyweight leads to increased strength in the squat and deadlift. The increased weight used in the squat and deadlift. The increased weight in the squat and deadlift leads to even greater increase in appetite. See where this is going? That's right, squatting and eating will increase the body's size all over. Simply training upper body won't give the same hunger response as heavy lower body compound movements.
Also, let's not forget that any training will induce growth for a while. You could put a new lifter on the worst routine and he'll grow for a while. So, I'm not suprised that the non-squat group showed equal growth.
And then we can discuss the CNS engagement or the Grow-or-Die response of squats vs curls...It has been proven time and time again in the gym, not in the science lab, that squats have a tremendous growth effect where other lifts have failed.
I can agree with squats, but how can people say deadlifts don't help build bigger arms....wtf are you holding the weight with?
I've seen people with gigantic arms but absolutely no lower body before so I don't doubt what was written. I happen to want tree trunk legs so its not really going to affect my training.
Also the increased leg/lower body strength can help with certain lifts that are primarily upper body lifts, such as OH press and bench which will in turn lead to higher loads, in turn leading to increased strength.
I'll trust the scientists that increased hormones may not help, if thats what their study found, but I never even new that was considered a reason why growth was stimulated from lower body work...I always thought it was simply what offroad just said.
True, I have meet lots of guys with huge upper bodies and little pin legs.
I do agree with the appetite thing.
Bench geared 525 @ 210
Deadlift geared 650 @ 220
Captains of Crush #3
Building Mighty Mitts...
One of my favorite weightlifting sayings is, "Train the appetite."
If hunger is boosted, you are doing something right. If hunger is decreased, you are doing something wrong.
As you can see from my sig, my bench is pretty disproportional from my squats/dead. And true to what I heard on here, I started squating/deadlifting and I have hit a PR every week since I started my routine, including my upperbody lifts, which had stalled for quite some time prior to training lower body as well.
I am a believer that good nutrition and continuous stress cause muscle to grow. I believe that squats are a great exercise and one that I love and am thankful that I can do.
Now I am another one that does not know if it causes any type of hormonal response or not ... I just do them.
I do know there are some darn big wheel chair bodybuilders and they sure as heck don't do squats.
Great article. Pretty much sums up what I've always bought into, and I'm glad there's finally some scientific proof backing it.
ain't nuttin but a peanut.
You will have gotten stronger when the weight that feels heavy is actually heavier than the weight that feels heavy now. Then the weight that feels heavy now will be a warmup for the weight that feels heavy then. But the weight will always feel heavy or you're not lifting enough weight. Clear? -Rippetoe
It has been proven time and again both scientifically and empirically that GH and testosterone increase strength and size. Or is he saying that the hormonal response from exercise is too minor to have an effect? Even if that's the case, over time it will surely have some cumulative effect. I'd like to know what the anabolic threshold of these substances is compared to the amount produced from deads and squats.
Last edited by J.C.; 12-04-2009 at 04:59 AM.
I'm no expert but from my experience, squatting and deadlifting have really strengthened my core and to an extent with my back which has definitely helped, me do bigger loads on my upper body which in turn has helped me to put on muscle. Also defintely agree on the squat and deadlift helping you train harder.
Deadlift:162.5kgs x3 (357.5lbs)
Dips: BW+ 30kgs x6 (66pbs)
Not gone that heavy with my bench or squat yet but will update them soon!
I studied this subject pretty extensively when I did my Master's thesis, so here are a couple points to keep in mind when discussing this topic.
1) Even the scientists who study this area for a living don't agree on whether this acute hormone response has any significant cumulative effect on training adaptations.
2) It's not entirely clear whether exercise actually increases hormone output at all. Blood concentrations of testosterone and GH do increase, but this has been shown to be primarily due to hemoconcentration due to plasma volume shifts during exercise. Nevertheless, most think it is still relavant because muscle and other tissues are exposed to higher concentrations of these hormones.
3) The increase in hormone concentrations are very short lived with hormone levels back to baseline within 15-60 minutes after exercise.
Still a very interesting study though.
In this study they are not using Squats!!!!! They use leg extension, leg curls, and leg presses. Caught your attention? well read on.
First off, I'm a big fan of Lyle McDonald. He does a very good job of conveying scientific information relative to a trainee's interest in an easy to understand manner. I would definitely recommend his books, they are worth every penny.
I took a few minutes this morning to locate the full article and read it carefully. If i'm given an abstract with results, I'm ALWAYS skeptical unless I can fully understand the experimental design, statistical analysis, etc. I cant copy/paste the PDF (doesnt format right) so here are some thoughts/comments/excerpts.
- They used 12 young men in the study.
- Each individual trained both arms in a different manner. For example, the left arm was trained using a curl exercise only, then 72 hrs the right arm was trained with a curl exercise followed by leg training, then repeat.
- 18g of protein was given immediately before exercising and 90min after exercising
My comments: Why on earth they chose to use the same individuals for both treatments confuses me. We are dealing with two treatments: LH and HH protocols. Since both treatments are performed on the same individual, our samples are no longer independent. This is going to have implications for statistical analysis later. Also, they don't explain why 12 individuals was the magic number chosen. In a good study, usually some sort of power calculation is performed to determine sample size (power is basically the probability of rejecting certain types of errors; more power is better, and more power usually comes from larger sample size). Its not clear if it was a budget issue, or if 12 was an arbitrary selection. Seems small to me.
- trainees were only recreationally active and did not have extensive lifting experience
- LH treatment consisted of 3-4 sets of 8-12 reps of arm curls
- HH treatment of the LH treatment immediately followed by 5 sets of 10 reps leg presses and 3 sets of 12 reps of leg extension/leg curl supersets
- between-set rest was 120s for arm exercises and 60s for leg exercises
Squats weren't even performed!!!!!!!!!!!! WTFBBQ!!111!11one Starting to get really skeptical now . . .
Kinda complicated here, but basically they ran analysis of variance (ANOVA) on a variety of factors. Hypertrophy and Strength data were analyzed using two factor ANOVA (training x condition). They used Tukey's HSD post hoc analysis.
My comments: First off, one of the basic assumptions of running ANOVA is that your treatments are independent. In this study, THEY ARE CERTAINLY NOT INDEPENDENT, since the same individual was given both treatments. In that respect, ANOVA is not the proper analysis of variance model. The Kruskal-Wallis test would have been more appropriate.
Tukey's HSD post hoc analysis is an EXTREMELY conservative test method. Before going into this I guess I should explain a little bit...
Whenever you are performing an ANOVA, what you are doing is testing to see if either
1. All treatments are equal or
2. A number of differences exists between treatments.
ANOVA can only tell you if number 1 or 2 is happening. It does not tell you which specific groups differ from one another. This is where you run post-hoc analysis using a variety of mathematic methods.
But here's the thing... When comparing multiple treatments, there's a problem in statistics known as the multiple testing problem, which basically means that errors in inference are more likely in certain situations in other (specifically, you as a researcher must decide if a Type 1 statistical error is more important than a Type 2 statistical error, etc. There is a tradeoff between both with multiple comparison methods)
They chose to use Tukey's HSD testing method which is ultra conservative. It controls for Type 1 error but results in much fewer significance differences.
I would have much rather seen the researchers perform whats called the SNK method, which falls right in the middle (not too conservative but not too liberal). I'm willing to be more differences would have been seen.
I'm still a bit skeptical at the results. Too few individuals, NO SQUATS, fuzzy statistical analysis...... I'm not 100% convinced.
Last edited by Cmanuel; 12-04-2009 at 10:05 AM.