One of the goals of a strength and conditioning program is to enhance an athlete’s power. Certain exercises, like the Olympic lifts, do a better job of this than others. Having said this, there is an inherent limitation to even the Olympic lifts, which is that fatigue has a negative impact on power. Think about this, an athlete is “freshest” during the first repetition, but after that first repetition fatigue begins to set in which decreases the amount of power the athlete is able to produce on subsequent repetitions.
Cluster sets are a way around this limitation. Cluster sets are an approach to strength training where there is a defined amount of rest between each repetition (rather than performing them continuously and then resting). The idea is that this helps the lifter maintain their power output during the set, which can have beneficial adaptations and potentially improve the athlete’s power production over the long term.
Now this is one of those things that sounds good in theory. How do cluster sets compare to traditional training (i.e. no intraset rest)? If cluster sets are better, could you improve on them by varying the weight within the set? For example, doing the first repetition at 70%, the second at 65%, etc.
In a study in the Journal of Strength and Conditioning Research, Dudley et al investigated this. They did this using counter-movement jumps and four different experimental protocols using the hang power clean. Their subjects were seven male and three female weightlifters ranging in age from 18-35. On average the males could do the hang power clean with 112kg (almost 1.3 times bodyweight) and the females could hang power clean almost 71kg (almost 1.05 times bodyweight).
All subjects did four experimental training sessions in random order. All sessions began with a warm up, three counter movement jumps, the experimental protocol, then another three counter movement jumps. The experimental protocols were:
- Traditional: Three sets of five reps at 70%, no rest between reps
- Basic Cluster: Three sets of five reps at 70%, rest 30 seconds between each rep
- Low/high: Three sets of five reps at 65%, 70%, 80%, 70%, and 65% with 30 seconds rest between each rep
- High/low: Three sets of five reps at 75%, 70%, 60%, 70%, and 75% with 30 seconds rest between each rep
Results:
- Peak power was almost 7% lower on the fifth repetition in the traditional protocol compared with the basic cluster protocol.
- When comparing the different cluster protocols, peak power was almost 5.5% higher in the third repetition of the Low/High cluster compared to the basic cluster protocol. Peak power was more than 4% higher in the fifth repetition of the high/low cluster compared to the basic cluster protocol.
- When looking at peak velocity, significant differences were between the low/high protocol and the basic cluster protocol (between -5% and 3.5% differences on the 1st , 3rd, and 5th repetitions).
- Significant differences in peak velocity were observed between the high/low and basic cluster protocols on the first and third repetitions (-4 to 8%).
- There was a similar pattern for peak force, with the low/high and high/low cluster sets being both reduced and greater than the basic cluster sets on different repetitions.
- There were no changes in jump height across the protocols.
With the high/low and low/high protocols, essentially the heaviest repetitions had the greatest decrease in velocity and the greatest force output. The greatest power outputs were also at 80%.
Basically what this shows is that the 30 seconds rest between each repetition is superior to no rest in terms of maintaining power production across a set. However, there is no advantage to manipulating the load within the set (e.g. doing the first rep at 75%, the second at 70%, etc.).
Some care has to be taken with the practical applications of a study like this. This type of training protocol is going to be difficult to implement in a high school or collegiate team setting where the training groups are pretty large. For example, if I have 50 athletes training and only ten platforms, each athlete taking 150 seconds per set is going to eat up a huge amount of time. This would work in a smaller team setting or an individual sport setting (like with the weightlifters) or in a small group/1:1 training environment.
Dudley, C.E.P., Drinkwater, E.J., and S.A. Feros. (2022). Different cluster-loading protocols have no effect on intraset and interset power expression. Journal of Strength and Conditioning Research, 36(7), 1763-1769.
