Thursday, April 4, 2013

Some thoughts on optimal swimming ‘style’ for triathletes

Alan Couzens, MS (Sports Science)

In my last post, I offered some thoughts on how the optimal running gait at Ironman pace may differ from the optimal running gait in traditional distance running events. Specifically, I highlighted that scientific investigation on the topic to date has found that the most important ‘technique’ element in maximizing economy is cadence and that stride length (& resulting ‘style’) will differ markedly between events. Finally, I brought into question teaching methods that seek to mimic the traditional ‘pretty’ run technique of high speed runners when applied to low speed events.

In this post, I’m going to investigate whether the same is true for the Ironman swimmer, i.e. should we seek to mimic those long beautiful strokes of Thorpey et al., or is there a better way?

Identifying the most economical swimming technique for an individual is a surprisingly complicated proposition. In addition to the various morphological & biomechanical factors that influence economy, there are several different types of economy or efficiency that we can measure. Are we talking about propulsive efficiency, gross efficiency or delta efficiency?

Propulsive efficiency

Propulsive efficiency is what most people think of when they think of an ‘efficient’ swimmer. It refers to how much of the swimmers movements result in forward propulsion vs those movements that are ‘lost to the water’. Huub Toussaint has done some neat studies utilizing his proprietary MAD system, which measures actual force of each swimming pull vs the ‘real world’ aquatic propulsion that comes from this force. Unsurprisingly, he has found that elite pool swimmers exhibit significantly better propulsive efficiency than elite triathletes. Elite swimmers only lose 40% of the energy that they apply to the water, while elite triathletes lose up to 55%!

This difference in the propulsive part of a pool v/open water stroke is easily visualized as the horizontal distance between when the forearm reaches vertical and purchase is made at the start of the stroke & when the forearm breaks vertical and purchase is lost at the back end of the stroke…

The figures above show that the distance between gaining the vertical forearm in the front and losing it at the back is greater with a traditional swimmer technique and less with a more circular, higher revving technique typical of a triathlete. This leads to a greater propulsive efficiency per stroke for the traditional swimmer.

Gross efficiency

Similarly, the swimmers rule when we take a closer look at gross efficiency. Gross efficiency refers to the efficiency in converting metabolic power into mechanical power. In the world of cycling, this is easily measured as the amount of metabolic energy (kilojoules or kilocalories per unit of time) vs the amount of mechanical energy produced per unit of time (i.e. watts). Typically, the gross efficiency numbers arrived at for cyclists fall in the 20-25% range. Utilizing Toussaint’s MAD system, a similar method can be used for swimmers to compare metabolic work (VO2) vs mechanical work produced. Even for the best swimmers, swimming is quite a bit less efficient than cycling as an activity, resulting in gross efficiency numbers of ~7-9%.

Again, when comparing swimmers and triathletes, swimmers take significantly less oxygen to produce the same levels of swimming power. Put another way, if we have a swimmer and triathlete with the same VO2max, the swimmer will produce significantly more mechanical power & (assuming similar propulsive efficiency) will go significantly faster at this effort level.

Delta efficiency

This begs the question, why don’t more elite triathletes swim like pool swimmers? Sure the more circular, faster revving technique offers some obvious advantages in choppy, close quarter swimming but is there more to it than that?

Returning to the world of cycling, professional cyclists have had the guys in lab coats confused for quite some time. See, the bulk of lab studies that have looked at the most efficient cycling cadence have found that a cadence of approximately 60rpm consistently results in the lowest oxygen cost for a given power output. However, we know from watching professional bike racing that elite cyclists rarely work at such a low cadence in the real world despite it, theoretically, being the most efficient. The million dollar question – why?

Ed Coyle has spent a lot of time crunching the numbers and investigating this question and he came upon a curious phenomenon. While absolute O2 cost may be lowest at low rpm, the change in O2 cost with increasing power is actually less for a higher rpm ‘style’ than a lower one. This is shown graphically below…


The blue line represents mechanical vs metabolic output at 100rpm, while the red represents the same relationship at 60rpm. While 60rpm is more economical at low power outputs, the difference disappears at ~400W

This difference in slope (rather than absolute O2 cost) between these 2 conditions has been termed the delta efficiency. The delta efficiency eliminates the basic (unloaded) O2 cost of the activity (200W for the 100rpm condition & 47W for the 60rpm). A practical example….

We take an athlete who uses a powertap (hub based power meter) and hook him up to a metabolic cart (VO2max test machine). We have him pedal at 200W and 90 rpm and monitor his metabolic response. Mid-way through the test, his chain breaks & (mechanical) power output drops to zero but metabolic output doesn’t. He’s spinning away, generating zero power but it’s costing him some energy.

All of a sudden he figures out what’s going on and slows his cadence. Lo and behold, his metabolic output goes down. In other words at zero watts and 90 rpm, his metabolic cost is higher than it is at zero watts and 60 rpm. If he just sat there on the bike and didn’t spin at all, his metabolic cost would be lower still. The question is, in real world athletic competition, to what extent are these non propulsive ‘spinning the pedals around’ movements limiting?

Put another way, what ultimately slows a cyclist down? Is it an inability to keep up with the high levels of whole body O2 delivery or are peripheral factors (within the muscle) the limiter? It has been argued that the limiter will vary both among different types of cyclists and among different types of events. For example, renowned sports scientist Alejandro Lucia has suggested that the difference between the high rpm climbing style of Lance Armstrong and the low rpm climbing style of Jan Ullrich was fundamentally the difference between an athlete who was ‘centrally strong’, i.e. Armstrong and an athlete who was ‘peripherally strong’, i.e. Ullrich. If we apply this back to the pool, it would be reasonable to assume that most triathletes fall on the ‘centrally strong’ side of the fence and can afford to give up a little gross efficiency in the name of delta efficiency, esp. considering the duration of their event.

This central v peripheral demands of different pedaling cadences has been investigated in a number of studies that have arrived at a similar conclusion: While absolute ‘whole body’ metabolic cost is higher in the high cadence condition, the peripheral cost within the ‘prime mover’ muscles is lower. This results in improved hemodynamics (O2 in, lactate out), decreased muscular stress, decreased metabolic stress (glycogen use) & improved peripheral endurance (e.g. Ahlquist et al., 1992, Faria, 1992, Gotshall et al., 1996, Takaishi, 1996). For example, The Ahlquist study showed 28% less total muscle glycogen used by cycling at 100rpm vs 50rpm at the same power output. The bulk of this difference was in the less economical Type II fibers, indicating less reliance on these with the high cadence approach.

Does the above also apply to swimming?

Indeed, if we analyze the data from Toussaint’s study on swimmer v triathlete economy with a view to identifying the optimal stroke length for delta v gross efficiency, we find a similar trend seems to exist in the pool, with presumably similar metabolic/peripheral consequences.

From the chart, it appears that the ‘sweet spot’ for maximizing delta efficiency in swimming is a faster revving, shorter length technique of ~0.4 x standing height, while the sweet spot for maximizing gross efficiency is a longer, more traditional technique of ~0.7x standing height. In practice, competitive triathletes split the difference and find the optimal balance between gross and delta efficiency at ~0.55x standing height. Real world stroke per length numbers for each of these conditions are shown below for different pool lengths


In summary, it is suggested that the optimal swimming style will vary with the absolute effort level. In events where VO2max is limiting (events in the 3-15min range typical of middle distance and distance pool events), the optimal technique will tend towards a longer stroke that maximizes pace for a given VO2 output (a stroke of ~0.7x height).  However, for longer events that are more limited by peripheral factors such as the lactate threshold or metabolic efficiency (i.e. most triathlon events), a shorter, faster rate stroke (closer to 0.55x height) may ultimately prove more economical.

Train smart,


No comments: