Sunday 30 March 2014

The Slower, The Stronger, Or A Take On Time Under Tension

Greetings, Brethren of the Iron Cult.

Today, it is my intention to take you out of your comfort zone. Nah, no groping, it'll be worse.

I know you think you work hard, but... How hard are you working? Yeah, I know. You bench 100kg! How comes everybody always guesstimates his strength at a 100kg bench?
Truth is, if you really go to the gym with the hopes of ripping sleeves, you may need to worry a little less about the weight and a little more about the execution of the lift.



This guy definitely leaves the ego at the door. Do you?


We all know hypertrophy requires a good diet (a minimum amount of protein, a positive balance energetic state, etc), and a good routine. But the latter is seldom paid enough attention. What frequency are you using? What volume? Fortunately, people are more aware about the former as of late, and the latter was always higher than necessary for most gym-goers. But what about execution?

Within this we can consider reaching failure and Time Under Tension, henceforth referred as TUT. Most people aiming for hypertrophy usually reach failure alright (they've watched Pumping Iron I guess?), but not many pay attention to their tempos. And I'm sorry to bear bad news, but throwing the weights around without giving two fucks about form or TUT is not benefitting you.

We all know that slowing tempos can help to avoid momentum, but slowing the concentric can turn the reps into grinders and make you use less weight than you could. That's bad, mmmkay?
What about slowing it especially in the eccentric phase? You can still do fast, even explosive concentrics, and thus use heavier weight, and slow it down in the eccentric, enough to produce the metabolic stress you're seeking.

Tanimoto et al tested 36 lifters, and had two different groups do 5 basic compounds, with group 1 using 55-60% of 1RM with tempos 3-0-3-0, and group 2 using 80% of 1RM with tempos 1-0-1-1. Both aimed for 8 reps, trained twice a week, and followed this protocol for 13 weeks. The results? Group 1 increased muscle thickness by 6.8 +/- 3.4% and 1RM strength by 33.0 +/- 8.8%, whereas group 2 did by 9.1 +/- 4.2%, and increased their 1RM strength by 41.2 +/- 7.6%. Pretty significant, given group 1 was probably working below their full capacity (they were using only 55-60% of 1RM after all).

Nicholas et al had 8 test subjects perform three sets of unilateral knee extension exercise at 30% of one-repetition maximum strength with either 6-0-6-0 or 1-0-1-0 tempos. Protein synthesis, which was checked through vastus lateral is biopsies, showed an increase of 114% in the slow tempo group, and 77% in the fast one. In this one, we can see better results due to not having such a high discrepancy between the intensities used. We must note, however, that both groups worked to failure.

Cameron et al split 18 test subjects in 3 groups, with leg trainings of 3 sets of 30% of 1RM, 3 sets of 80% of 1RM, or 1 set of 80% of 1RM. Working to failure, 3 times a week, and 10 weeks. The usual jazz. Well, apparently no significant differences were found between the two 3-set groups in terms of muscle protein synthesis, and only the group which had done a single set had had worse results.



I like numbers. They make me feel smart and shit.



What can we make of all this?

When test subjects reached failure, intensity was slightly influential, but nowhere near decisive, in comparison to total volume. Total work load determined the highest group MPS when the rest of the variables were equal in tested groups.
When test subjects reached failure using slow tempos instead of faster reps, they showed higher MPS.

What can you do?


  • This can't be stressed enough. If you're working for strength, you should work it rather than test it. Other way to say it is "leave your ego outside the gym". This is especially important for hypertrophy lifters.
  • Working to failure is good yo.
  • Working beyond failure is even gooder.
  • Forced reps, or having your routine include negatives, is goodest.
  • "Both 3-set groups had the same MPS, but I'd rather lift heavy". Yeah, well, guess what group could aim for higher volume without frying their CNS.
  • We know volume and TUT are cornerstones for hypertrophy, and high frequency trumps over everything in my experience. Aiming for high volume while maintaining frequency 2 or 3 is a great way to work for hypertrophy, and strength. Full body, upper/lower, and L/P/P are great ways to train. Probably you don't need to do over 8-12 sets per MG and day. Forget about your 50 sets old chest day which took you nowhere. Keep your reps moderately high. MPS was shown equal for both groups, but the one with higher reps will be able to do higher volume without taxing their CNS as much. Now that's quite a thing eh?


Stay around,

J



Sources:


Manimoto, Sanada K, Yamamoto K, Kawano H, Gando Y, Tabata I, Ishii N, Miyachi M. Effects of whole-body low-intensity resistance training with slow movement and tonic force generation on muscular size and strength in young men. Journal for  Strength Conditioning Research. 2008

Nicholas A Burd, Richard J Andrews, Daniel WD West, Jonathan P Little, Andrew JR Cochran, Amy J Hector, Joshua GA Cashaback, Martin J Gibala, James R Potvin, Steven K Baker, and Stuart M Phillips. Muscle time under tensiontension during resustance exercise stimukates differential muscle protein sub-fractional protein responses in men. J Physiol, 2012

Cameron J. Mitchell, Tyler A. Churchward-Venne, Daniel D.W. West, Nicholas A. Burd, Leigh Breen, Steven K. Baker, and Stuart M. Phillips. Resistance exercise load does not determine training-mediated hypertrophic gains in young men. Journal of Applied Physiology, 2012

7 comments:

  1. "Nicholas et al had 8 test subjects perform three sets of unilateral knee extension exercise at 30% of one-repetition maximum strength with either 6-0-6-0 or 1-0-1-0 tempos. Protein synthesis, which was checked through vastus lateral is biopsies, showed an increase of 114% in the slow tempo group, and 77% in the fast one"

    Bueno, teniendo en cuenta que los del tempo lento tuvieron que trabajar 6 veces más tiempo (y sufrir más) para obtener un 48% más de mejora con respecto al otro grupo... seguramente habrá quien prefiera usar su "tempo natural" (más rápido) que estar contando segundos, aunque gane menos. La ganancia será menor, pero es más sostenible en el tiempo.

    Aparte, en ese estudio usaban la misma carga. Eso bajo mi punto de vista implica que los del segundo grupo acabarán menos cansados, por tanto podrán entrenar más o con más frecuencia, quizás obteniendo similares resultados al primer grupo ¿no?

    Un abrazo fiera

    P.D: Perdona que no escriba en inglés, pero lo mío es traducir :)

    Maokoto

    ReplyDelete
  2. Cuidado, ten en cuenta que ningún grupo trabajó más, ambos fueron al fallo. El hacer un TUT más largo en parte mejora la activación muscular al eliminar la inercia. 6 segundos es una exageración, pero un tempo lento (aunque más rápido) sigue siendo positivo :)

    ReplyDelete
  3. Vaya pues ahí si que me has matado, es verdad, no lo tuve en cuenta.

    Saludos

    ReplyDelete
  4. And what if reaching muscle failure is to blame for most of the CNS taxing? Wouldn't it be nice to see a similar study but with a predetermined set/rep scheme that avoids failure?

    ReplyDelete
    Replies
    1. How is that relevant to longer TUT triggering a higher MPS?

      Wat

      Besides, e=mc⅝

      Where's your God now.

      Delete
  5. Relevant for the frequency you can use due to CNS recovery issues, I think. More sets per week leads to more total TUT.

    By the way, the study in which you show errors is not really significant (statistically speaking). The other two only give significances in terms of p-value, and I *hate* when they do that instead of giving errors.

    My goddess is about to finish a snack watching spongebob squarepants before we go for a bike ride. Well, she rides, I run. That's about all the cardio I do in the week.

    ReplyDelete
  6. Yes, reaching failure accelerates CNS overloading. I guess it wouldn't be very bold to extrapolate results and assume that, given the same workload, better MPS could be triggered by a longer TUT. I assume they simply had the test subjects reach failure to make sure the work had been similar. And I fear I lack the experience you have in statistics to be so picky as to the methods used.

    ReplyDelete