Resting as Long as You Think Fit Reduces Training Time W/Out Reducing the Workload & (Hopefully) Your Gains

The study provides only preliminary evidence, but it is evidence...

You will certainly remember the SuppVersity article about the beneficial effects of long(er) rest times on strength and size gains from November last year (read more).

Now, after I posted this article, a discussion evolved about whether you actually have to wait that long (3 min) after exercises that don't leave you as winded as barbell squats; and if resting less than the "optimal" time wouldn't yield the exact same results if volume and weights you lift during set and individual rest time workouts were identical.

As the authors of a recent study from the Universidade Federal do Rio de Janeiro (Freitas de Salles. 2016), point out, there was, until now, no study that compared the "consequences of applying selfsuggested with fixed intervals in upper and lower body exercises performance". With said study, however, there's some data on what happens if you compare the effects of fixed (2 min) versus self-suggested rest intervals (RI) between sets in lower (squat and leg press) and upper body (bench press and biceps curl) exercises on experienced trainees' performance.
With at least 6 months of training experience at a frequency of three intense strength workouts per week, the subjects in Freitas de Salles's study are allegedly not as experienced as those in the previously discussed study by Schoenfeld et al. (2015), but they are by no means absolute rookies. If the scientists' initial hypothesis "that selfsuggested RI would result in similar performance to
the 2-min RI, with a shorter session" (Freitas de Salles. 2016) was confirmed in the study, there would thus be at least a good chance that the same results would be observed for better-trained athletes when they perform a workout the scientists describe as follows:

"Before the 1RM test, subjects performed standardization and familiarization set in each exercise, cor recting possible mistakes in the performance of exercises. Between different exercises a 5 min rest was given, and up to three attempts for each exercise with a three to 5 min RI were performed. Before each test a warm-up set in each exercise with 10 repetitions at 50% of 1RM self-perceived load was performed. After the 1RM test and retest a minimum interval of 48 h before ST sessions were adopted. All 1RM tests and all sets in ST sessions were closely super vised by an examiner to avoid facilitative movements or incomplete range of motion applied.  

The 75% 1RM load was stipulated, aiming to keep sets in the 8–12 repetition range. The participants were tasked to perform three sets to concentric failure and no specific velocity was determined to perform the exercises. Both groups attended the weight room on two separate days with an interval of 48 h between days for 2 min and self-suggested RI protocols. The order of different RIs applied was random and only the examiner had knowledge of what RI would be held before the sessions. The exercise sessions were accompanied by an experienced examiner who performed the count of the number of repetitions and controlled the RI duration without the participants knowing their rest time" (Freitas de Salles. 2016).

If you are now looking for the word "weeks" in the previous quote, you have already spotted a major problem with the design of the study at hand: Instead of evaluating the actual strength and size gains as it was done in the Schoenfeld study, Freitas de Salles et al. studied only the number of reps and the total volume (weight x reps) the subjects lifted on two different training days with either set or individual / selfdetermined inter-set rest periods.

Figure 2: Effective rest times (in s between sets) during the selfselected rest trial (Freitas de Salles. 2016).

Any statements about the effects on strength and size gains are thus necessarily based on the (scientifically not unwarranted) assumption that the total training volume (within the "optimal", i.e. "not overtraining" region) was the primary determinant of the adaptive response to exercise.

Figure 1: Repetition numbers for each exercise on three separate sets for 2 min and self-suggested RIs. ∗p < .05 from first set; #p < .05 from second set (Freitas de Salles. 2016).

So, if we assume that this stipulation was correct, the data in Figure 2 says: It doesn't make a difference, if you rest for the prescribed 2 minutes or start your next set after feeling recuperated - and since the selfdetermined rest times were >100 + X seconds, the study doesn't necessarily conflict with the previously cited study by Schoenfelt et al. (2016), because Schoenfeld's study tested "only" 60s and 180s. The 100 seconds + X of the selfdetermined rest in the study at hand is thus not so much off of the 3 minutes of rest in the Schoenfeld study or the 2 minutes, of which Willardson, et al. found in 2006 that they are sufficient to maximize the adaptive response to exercise.
References:

• De Salles, Belmiro Freitas, et al. "Effects of fixed vs. self-suggested rest between sets in upper and lower body exercises performance." European Journal of Sport Science (2016): 1-5.
• Ratamess, Nicholas A., et al. "The effect of rest interval length on metabolic responses to the bench press exercise." European journal of applied physiology 100.1 (2007): 1-17.
• Willardson, Jeffrey M., and Lee N. Burkett. "The effect of rest interval length on bench press performance with heavy vs. light loads." The Journal of Strength & Conditioning Research 20.2 (2006): 396-399.