Ross Sanders¹, Brent Rushall², Huub Toussaint³, Joel Stager⁴, and Hideki Takagi⁵

While there is little doubt that advancements from a clever application of fluid dynamics has improved performance in sports such as cycling and flat-water kayak racing, the 'jury is still out' with respect to 'hydrodynamically' designed swimsuits. At the 2000 Olympics, gold medals were won with and without the new swimsuits, with varying degrees of use, and in their various forms (Figure 1).

One is left unsure of whether these suits actually improve performance. Seeking a scientific basis for why they should or should not have an effect could provide answers. This article reviews theoretical aspects, the scientific rationale underlying design, scientific evidence relating to the actual effects, and some alternative explanations for benefits or decrements in performance that may be attributable to the suits. It is not written because the answers to many questions are known. Rather, it is an exploration of what should be considered.

Progression through water depends on the interaction of propulsive and resistive forces. A swimmer can improve by increasing propulsive forces and/or reducing resistive forces that act on the body at a given speed. The physiological cost of any strategy must also be considered.

When swimmers are not creating propulsive forces of sufficient magnitude, they slow down. It is frequently observed that some individuals seem to 'slip' through the water requiring less effort than others. Some swimmers look to be swimming well at slow speeds but when they attempt to increase speed they do not improve as much as others. One of the main reasons for these differences is the amount of resistance, more commonly referred to as 'resistive drag', created by the swimmer.

Figure 1. Swimming attire worn by a selection of gold medallists at the 2000 Olympic Games.

An understanding of factors contributing to resistance is important in modern swimming and coaching. It is a topic of renewed interest and is now considered more important than previously thought. It appears that adjustments in technique to reduce resistive drag may be as beneficial as subtle adjustments to improve propulsive force.

While resistance to forward movement of limbs and body should be minimised, resistance to backward movements should be maximised as these movements generate propulsive forces. Thus, when the forearms and hands are moving backwards their orientation, position, and direction of movement should maximise drag resistance.

Form drag. The orderly flow over the swimmers' body may 'separate' at a certain point, depending on the shape, size, and velocity of the swimmer. Behind the separation-point, the flow reverses and may 'roll up' into distinct eddies. Consequently, a pressure differential arises between the front and the rear of the swimmer, resulting in forces termed 'form' or 'pressure drag'. These forces are proportional to the pressure differential times the cross sectional area of the swimmer. Form drag of a body is proportional to the square of flow velocity and so becomes increasingly important and influential as swimming speed increases. To minimize form drag, a swimmer seeks a 'streamlined' position. Thus, swimming with a 'head-up' position in backstroke increases form drag because the hips 'drop', thereby increasing the cross-sectional area presented by the body as it moves through the water. If a swimmer's action or swimming 'posture' creates an increased cross-sectional area, then progress through the water will be resisted more than if in a streamlined position.

Form drag is one of the easiest factors to control and can be minimised by adopting streamlined postures at every opportunity (i.e., the swimmer has to create the thinnest and straightest form while going through the water). A general concept for most strokes is to have the shoulder/chest area create a gap in the water and the hips and legs follow through that space. That usually translates into swimming with the body as level as possible. Many new advances in technique have aimed at maximising streamlining, that is, reducing form drag. Kolmogorov and Duplishcheva (1992) showed that swimmers of similar body size (height and weight) could have drastically different drag values during swimming. The streamlined position of Kieren Perkins probably contributed considerably to his outstanding performances in the 1500m (Rushall & Cappaert, 1994; Figure 2).

A factor that could affect streamlining, and thus form drag, is the buoyant force of the swimmer (Chatard et al., 1990, McLean & Hinrichs, 1998; 2000). The buoyant force (buoyancy) and body weight form a force couple that creates a torque that tends to disrupt streamline. If the bodysuit increases the magnitude of the buoyancy force, or shifts the point of application toward the feet, it could have a form drag reducing effect. This feature will be discussed in more detail below.

Wave drag. Speed at the water surface is constrained by the formation of surface waves. As a swimmer swims at the surface, water is pushed out of the way. Waves result from pressure variation due to differential water velocities around the swimmer. As velocity increases, the bow wave, with increased size and inertia, cannot flow out of the way quickly enough and hinders velocity increases of the swimmer. Eventually, there is an effective speed limit, which for conventional ships with a fixed displacement hull is called 'hull speed' (see, for example, Aigeldinger and Fish, 1995). Wave drag results from the increased work required to climb the bow wave and from the transfer of kinetic energy from the swimmer to the water. Wave drag increases steeply and becomes the dominant drag component as hull speed approaches. Accentuated vertical movements increase Wave drag, for example, 'flying' out of the water in butterfly and lifting the head when breathing in front crawl. Any action that produces a force that is not directed along the longitudinal axis of the body in the direction of travel will cause lateral (rotational) movements of the body, hips, or legs, unless the motion is counter-balanced by another action. Unfortunately, the human anatomy does not permit all forces to be directed along the longitudinal axis. However, some swimmers have techniques that minimise lateral movements more than others. Any bouncing or jerkiness in a swimmer's style also creates wave drag. When lateral and vertical movements are larger than necessary, performance is limited by excessive wave drag.

Lane lines are used to minimise the effect that the mostly lateral waves have on adjacent swimmers in races. Waves have sufficient energy to assist a swimmer when they are in concert with the direction of progression (one form of 'drafting'), and to slow a swimmer (when they collide with an oncoming swimmer). A stiff, streamlined body just touching the interface between air and water experiences five times as much drag as the same body at a depth of more than three times its width (Hertel, 1966). Thus, wave drag is reduced when a swimmer is completely immersed at a depth of about 0.7 m. The international governing body of swimming (FINA) at various times has instituted rules to limit the amount of underwater swimming that can be performed. Primarily for safety and to a lesser degree the spectacle of the sport, rules have been instituted to limit the underwater swimming distances, and in the case of breaststroke, the amount of immersion per stroke.

Wave drag is potentially the greatest limitation to a swimmer's performance because it increases in proportion to the cube of swimming velocity. Fortunately, a swimmer has some control over wave drag. Wave drag can be minimised by reducing unnecessary vertical and lateral movements. Attempts to over-extend forward and backward that produce even the slightest bending of the body up or down are not worthwhile because of increased wave drag. Similarly, attempts to swim 'over' the water in crawl stroke and butterfly increase wave drag.

There are some beneficial vertical movements that can contribute to forward propulsion. A wave action that travels down the body in modern butterfly and possibly breaststroke could be helpful (Sanders et al, 1995; 1998; Sanders, 1995). However, if that action is exaggerated to the point where the undulation is too large and the wave is not as fast as the swimmer's velocity, then it will actually slow the swimmer more than if no wave action was attempted at all.

It is unlikely that a bodysuit would have much effect on wave drag. Large movements, rather than the surface of the mover, cause waves.

Surface drag. Often called 'skin friction', surface drag is commonly attributed to the forces tending to slow the water flowing along the surface of a swimmer's body. The magnitude of the surface drag depends on the velocity of the flow relative to that of the body, the surface area of the body, and the characteristics of the surface. Skin roughness, body contouring, hair, and swimsuit fabric are examples of the surface characteristics that create friction as a swimmer moves through water. At the high Reynolds numbers (>10⁵) that occur during swimming (Toussaint et al., 1988a) increases in velocity cause a relatively much smaller increase in surface drag than in form drag and wave drag.

There is some evidence that shaving hair off the body and legs can reduce surface drag. The reduced resistance causes a reduction in the energy per stroke when compared to an unshaven condition (Sharp & Costill, 1990). Because the forearms are used to produce propulsive drag, there is no advantage in shaving the forearms. Wearing a latex cap provides a smoother surface than does a head of hair and thus, reduces drag. Tight swimsuits of sheer fabrics with a structure that minimizes seams and edges may reduce surface drag. Recently developed bodysuit surfaces are supposed to generate less resistance than natural shaved skin.

In summary, total resistance encountered during competitive swimming is the result of the summed effect of form drag, wave drag, and surface drag. For each component it is possible to estimate the magnitude using general formulae. However, care should be exercised since:

1. Formulae are normally based on objects with constant shape and orientation to the water flow. A swimmer's shape and orientation to the fluid flow, changes even when only in a glide position.
2. When a body of constant shape and orientation is pulled through water at constant speed, or put in constant water flow such as a flume, the patterns of water flow around the object are not consistent and change dynamically. Therefore, fluctuation in resistive forces is an inevitable natural occurrence. Even when forces are measured over a period of towing and averaged, calculated coefficients will vary (see for example the variability in calculated coefficients of a swimmer's hand reported by Sanders, 1999).
3. Coefficients and constants in formulae are not completely independent of velocity (see Toussaint et al., 1988a). Although the coefficients remain reasonably constant for particular ranges of velocities, those ranges vary according to the shape of the object. Swimmers are almost constantly changing shape. Given that the swimmer is in the interface of air and water, a change of velocity might invoke additional changes through variations in technique and position in the water. For example, an increase in speed might allow the swimmer to 'hydroplane' and reduce the surface area of the body in the water thereby reducing the surface drag and possibly form drag.
4. If velocity changes during the measuring period, there is an additional force due to acceleration of a mass of water (see for example Pai and Hay, 1988; Sanders, 1999).

However, given the above barriers, some practical estimates of resistive forces are possible. Tests of swimmers in a constant glide position yield measures of 'passive drag'. Tests of swimmers actually performing a swimming stroke yield much more realistic estimates of resistive forces (Toussaint et al., 2000; Vaart et al., 1987). This latter measure is termed 'active drag'. Active drag is difficult to determine directly because the forces that act on the swimmer must be measured without disrupting the natural swimming movement. An exact way of doing this has not been found. However, the method used by Toussaint et al is considered to yield very good estimates of active drag. To measure the effect of the suit on active drag, swimmers must be consistent in technique and effort when swimming with and without the bodysuit.

Research involving bodysuits is, at best, very difficult. The number of factors that need to be controlled (e.g., the fit of the suits, the conditions of testing, placebo effect, maintenance of constant wetness, etc.) is intimidating. Few objective scientific studies have been conducted. Manufacturers claim scientific bases and studies for their products but have failed to make such work available for independent evaluation.

A basic problem with researching the effects of bodysuits on swimming performances is the theoretical bases used. Most hydrodynamic models are based on static objects (e.g., boat hulls, hydrofoils), but swimmers are constantly dynamic. With arms and legs moving in all planes through four different swimming forms, generalizing from static to the dynamic models, and then across swimming strokes would be spurious. Many theoretical models would be inappropriate for swimming as would be testing of materials in an environment other than swimming.

Bergen (2001), a swimming coach, conducted a practical test of the effects of Speedo's Fastskin suit on swimming performance. Without any statistical analysis, simply comparing the means of groups of 25-m sprint performances between Fastskins and conventional suits, he concluded:

• The bodysuit had a significant advantage for underwater kicking and above-water swimming for both crawl stroke and butterfly.
• There was no advantage in wearing the suits for backstroke for either underwater kicking or above water swimming.
• The bodysuit had negative effects on breaststroke underwater kicking and swimming.
• The differences were only apparent when compared with a normal racing suit (not a 'jammer' or other variation) and with 'unshaved' swimmers.

Physical and mechanical advantages are gained from shaving before important competitions (Sharp & Costill, 1989). Bergen's conclusions are only valid for unshaved swimmers. It is possible that once shaved, a previously unshaved swimmer would equal or surpass the performance benefits from the suits. The benefits of bodysuits might only apply to unshaved conditions. It is also possible, that a statistical analysis might not support Bergen's 'eyeball' conclusions. A further possibility for Bergen's opinions lie in the 'conventional' suits used. Swimmers in competitions wear suits that are particularly tight, usually several sizes smaller than a 'normal' fit. However, in training, 'comfortable' suits are worn possibly contributing to greater drag because of looseness of fit. Thus, Bergen's results could be partly attributed to the slowing of the swimmer due to the conventional 'training' suits rather than bodysuit enhancements. Despite those misgivings, Bergen did conclude the bodysuits to be of no value for backstroke or breaststroke, and that the bodysuits' restriction on movements might even hinder races involving turns.

Some further understanding of the 'bodysuit-versus-conventional-suit' question was provided through a scientific investigation by Toussaint et al (in press). It was found that when compared to conventional suits on unshaved swimmers performing crawl stroke,

• the amount of drag reduction with Speedo's Fastskin bodysuits was nonsignificant,
• there was a suggestion of a subjects by speed interaction, and
• there was no consistent effect across subjects or swimming velocities.

This study suggests that Bergen's non-statistical conclusion of the suits benefiting crawl swimming was presumptuous. Toussaint et al did report an improvement in drag reduction for bodysuits, but it was insignificant and contradicted the claims of the manufacturers. It is realized that Bergen's estimates were based on swimming speeds, while Toussaint et al considered drag reductions. However, a more controlled, statistically analyzed study did not verify the practical interpretations of Coach Bergen, at least for crawl stroke. As an aside, Toussaint and colleagues opined that loose conventional suits could have increased drag, and provided an example of it contributing to a large drag reduction by the bodysuit worn by one swimmer.

A further confounding factor that has not been controlled, is the wetness of the bodysuits. In a dry state, bodysuits float very well and take a long time to sink unless forced into and moved in water. Floatation, or buoyancy force, is provided by trapped air and surface bubbles (see Figure 6).

It is possible that early swims in any study could enhance bodysuit benefits through floatation. Swimmers (see for example, Gould, 2000) reported a floating sensation when swimming with a bodysuit for the first time. As a study progresses, and each trial begins with an increasingly wet bodysuit, the floatation effect would dissipate. It is possible that floatation could affect performance for one or only a very few trials when bodysuits are dry (as in a race). Toussaint et al used the bodysuits repeatedly and did not control for wetness in the suits. They concluded that floatation was not a factor in these suits, which would be true once the suits were thoroughly wet.

Adding buoyancy does not simply reduce frontal area, and hence, form resistance. It is important to look at the couple that is formed by the weight and buoyancy forces in the swimmer. The couple equals the torque that tends to sink the feet to a lower position creating a less streamlined position by increasing form drag. A major purpose of most kicking actions in all swimming strokes is to counter-balance that torque. This is also a reason why not covering the lower legs with resistance-reducing materials is important. If that was done, a swimmer would have to kick harder, and therefore exacerbate fatigue, to remain streamlined. A counter argument would be that the floatation provided by bodysuits at the shank extremity would require less kicking.

Buoyancy could also explain the interaction of bodysuits with body types. By covering body parts on the legs side of the center of buoyancy, torque would be reduced. The further the added buoyancy is from the natural buoyancy center, the greater would be the effect. Swimmers whose feet sink easily, would gain most from covering the hips and legs with a buoyant material. Male swimmers generally experience leg-sinking more than females. A popular form of bodysuit worn by males, the 'Jammer', is not a bodysuit at all. It covers only the lower hips and thighs and therefore, provides floatation on the legs side of the center of mass. That would assist males to streamline easier as it would support the hips and legs.

Females are required to wear a torso covering for modesty. The upper body cover would cancel out the lower-body torque-reducing floatation effect. However, while the suits remain dry, the overall floatation factor would serve to reduce the frontal area of the swimmer by the swimmer floating higher out of the water.

Using buoyant materials in devices is against the rules of swimming. The international governing body for swimming, FINA, has 'approved' bodysuits for competitions and ignored the rule governing devices and buoyancy.

The changing velocities of swimmers during a stroke cycle present another research difficulty. Cappaert & Rushall (1994) illustrated the velocity curves of Kieren Perkins swimming his gold medal race at the Barcelona Olympic Games (see Figure 2). One curve represents the right hip, the other the center of mass/gravity. Since the relative velocity of the water influences all forms of drag resistance, it is invalid to imply consistent effects of the surface of a swimmer on drag because the nature of the fluid flow will vary considerably within a stroke cycle. With form drag, the point of separation would be influenced by fluid velocity and that would differ on the various parts of the body, arms, and legs in the stroke cycle.

The information disseminated by manufacturers about their devices is at best confusing. The mechanism to reduce drag used in the Speedo Fastskin, is purported to be embedded microscopic vortex generators. Vortex generators are supposed to create microturbulence in the boundary layer, which postpones flow separation. If effective, these devices would reduce form drag, not surface resistance. However, discussions and advertisements covering bodysuits allude to reductions in surface resistance through slipping through the water or 'channeling' flow. Vortex generators are useful only at those places where flow separation is imminent, usually just upstream from the point of flow separation. But, as has been stated above, if flow separation changes frequently within a stroking cycle, the position of the vortex generators would have to change. One can only assume, that to overcome this relocation need, the Speedo Fastskin bodysuits have in-built vortex generators all over them. Thus, in the majority of places before flow separation, the generators would increase resistance, and after separation would be useless. Added to this, is the effect of the generators on resistance when they move in a direction that is opposite that for which they are designed. For example, when the legs are drawn-up in the preparatory phase of kicking in breaststroke, they would increase both surface and form resistance. Perhaps, that is one explanatory factor for Bergen's conclusion that bodysuits actually hinder the performance of breaststroke swimmers. This design factor is rather 'strange.'

Figure 2. Positions and velocity curves for Kieren Perkins at 70m from the finish of his gold medal 1500m race at the 1992 Olympic Games in Barcelona. The two curves illustrate within-stroke velocity variations, which will affect the amount of drag that exists at any one moment (from Cappaert and Rushall, 1994).

Besides the position of vortex generators on each moving part of a swimmer, their size is important as well. The protrusion of the 'generator' must not be too small or large. Waring (1999) concluded that for a swimmer, the optimal height of a vortex generator would be about 2.5 mm (.1 inch). That is much larger than the embedded vortex generators in the Speedo Fastskin.

The influence of surface drag on total resistive drag is relatively small. However, the designers of the bodysuits have reasoned that in the serious competitive situation where 1/100th second may determine the difference between rankings or breaking a record, reducing drag is important. Indeed this idea is not new. We see it in the attempts to reduce frictional air resistance in competitive attire in skiing, cycling, speed skating, and track events.

A popular suit worn at the 2000 Olympic Games was a Fastskin™ released for sale in 2000 by Speedo to reduce the drag. The key feature is its fabric, which was designed to mimic the properties of a shark's skin (Figure 3) by superimposing vertical resin stripes. The designers' intent was for the stripe to produce vertical vortices or spirals of water, which keep the passing water closer to the swimmer's body and reduce the formation of separation bubbles and hence, form drag - a phenomenon known as the 'riblets effect'. However, for riblets to be effective they have to be aligned with fluid flow. When they are displaced, resistance can increase. Thus, when swimmers roll from side to side, as in crawl and backstroke, or move vertically in butterfly and breaststrokes, the Fastskin could actually hinder performance. Based on passive drag tests a reduction in total drag of up to 7.5% was claimed when wearing the Fastskin (Speedo press release: Fastskin fact sheet). As was indicated above, that claim was refuted by Toussaint et al (in press).

Figure 3. A scanning electron microscope image of a shark's skin. Credit: John Mansfield/Microbeam Analysis Society

However, the assumption underlying the proposal that riblets are performance enhancing is itself controversial. Vogel (1996) questioned that tenet.

Another feature of the Speedo bodysuits is the placement of ridges of stitching that are supposed to 'channel' water flow more effectively. There is a problem with this hypothesis. Ridges act as resistance-enhancers when they are positioned at an angle to the fluid flow. In those cases, they baffle the fluid flow. Since swimmers often change the alignment of their body segment surfaces during strokes, it would seem ridges of stitching would increase resistance rather than reduce it. It is not difficult to imagine the detrimental effect of these ridges in breaststroke when the legs are drawn up, in opposition to the fluid flow, preparatory to kicking. Generally, even with static bodies such as boat hulls, ridges are minimised except for functions such as stability. Some other manufacturers have copied the Speedo ridge concept. However, it should be noted that the Speedo brand is the only one that uses the 'shark skin' analogy. Other manufacturers use different strategies for drag reduction. Unfortunately, these strategies are not well documented. This prevents objective analysis and understanding of the claims.

In addition to the reduction in form drag by reduced separation bubble formation, it is claimed that the new whole-body swimsuit assists in reducing form drag by making the swimmer's body more 'streamlined' in shape. Furthermore, wearing a tightly fitting swimsuit has the advantage that it tends to flatten out the contours of the body and reduce the amount of water that flows between the suit and the skin surface thereby reducing form drag.

This discussion has centered mainly around the Speedo Fastskin bodysuit because of its availability. Other manufacturers, such as Adidas, Tyr, and Mizuno, have produced bodysuits often with design differences. For example, Adidas reported the following.

From the Swimming Times (September, 1999) -- Adidas claimed the following in its advertisement 'Adidas revolutionizes swimming.'

Finally, Adidas stated:

The concepts and 'science' behind Adidas' product are very different from those of Speedo.

While manufacturers cite their own research to support claims that bodysuits improve swimming performance, independent assessment of that research will not be possible until it is made available for public scrutiny. Thus, one cannot be sure the manufacturers' research was conducted scientifically, validly, and objectively.

Assuming bodysuits reduce resistive drag without affecting propulsion or increasing physiological cost, the best way to establish the effects of the suits is to measure active drag with and without a suit. Such a study was conducted recently by one of the authors (Toussaint et al, in press) using an established method for the measurement of active drag (M.A.D system, Toussaint et al, 1988b). With this system, swimmers push off pads that are instrumented with force transducers. At constant swimming velocity, the mean propelling force is equal to the mean drag force (Vaart et al., 1987). By measuring the forces applied at each pad and the velocity of the swimmer, the researchers can determine the mean drag force. Using this system, active drag was calculated for six males and seven females swimming at different velocities (1.10 up to 2 m•s^-1) with the Fastskin neck-to-ankle bodysuit and with conventional swimwear.

For the Fastskin suit, a nonsignificant reduction in drag of ~2% (p = 0.31; Figure 4) was found, a figure considerably less than the 7.5% claimed by Speedo. Drag differences varied with velocity and swimmers. In most instances, there was no clear reduction in resistance with the Fastskin.

Figure 4. Drag data for all subjects showing active drag dependent on swimming velocity wearing the Fastskin and conventional suits. Fitted curves are presented as well. The overlap of all data demonstrates the lack of difference between each form of swimwear.

For some subjects an active drag advantage seemed present. The most extreme case is shown in Figure 5. At 1.65 m•s^-1 an 11% reduction in active drag was observed, a value that was nevertheless still not statistically significant. The non-significance means that observed differences may be due to uncontrolled factors (e.g., measurement errors, variability in swimmers' postures, placebo effects, the fit of the suits) rather than to the effect of the bodysuits. It was with this swimmer that the authors opined that the 11% 'reduction" came not so much from a benefit of the bodysuit, but an increase in resistance from an ill-fitting conventional suit.

Figure 5. Drag data for the subject who appeared to gain the greatest advantage from the Fastskin compared to a conventional suit. Fitted curves are presented as well. Although the curves differ, it is invalid to assert the differences are due solely to the type of suit used.

Stager (2000) used an alternative method to evaluate the impact of bodysuits. At the 2000 US Olympic Trials, all swimmers were issued with bodysuits, from several manufacturers but mostly Speedo. If these suits improved performance as manufacturers advertised, it would be reasonable to expect a 'step-like' improvement in all performances at the trials. Such sudden and noticeable improvements commonly occur when there is a rule change that advantages the swimmer.

Using data from US Olympic Trials from 1968 to 1996, several regression equations were developed, and the power curve best-line-of-fit was used to predict 'normal progression' times for the 2000 Trials. If the suits were as effective as proposed, most recorded times would exceed predicted times. Thus, Stager's work assessed whether bodysuits contributed to a better than expected level of performance. If there were no obvious improvements, the suits would be declared as not performance enhancing, and swimmers' performances would be in accord with reasonably expected progress.

Only two results differed significantly from predicted times. The women's 200m backstroke was significantly slower and the women's 100m breaststroke faster than predicted. No improvement impact associated with bodysuits was evident.

Stager's analysis considered shaved swimmers and adds one more bit of information. For what is known with shaved swimmers, the bodysuits do not provide an advantage.

Toussaint et al. (2000) indicated that swimming exercise intensity relates to the drag coefficient times swimming speed cubed. This can be used to evaluate manufacturer's claims. If the 7.5% reduction in total drag from Fastskins claimed by Speedo was correct, a 2.5% increase in swimming velocity could be expected. A 2.5% reduction in 100-m race time (2.5% of 49s = 1.2s) would be a sensational result. Stager determined the mean difference between predicted and actual times to be approximately 1/10^th of this, a non-significant difference. It is interesting to note that the winner of the men's 100m freestyle race at the 2000 Olympic Games broke the existing world record by .34 seconds while wearing only a waist-to-ankle suit.

At this stage it has not been shown that performance is improved by a bodysuit. Clearly, some believe that the suits do improve performance. Among those, the improvement is popularly attributed to the reduction of resistive drag. However, if the suits do aid performance then there may be alternative explanations.

One explanation is increased buoyancy. In particular, when buoyancy is increased in the hips and legs, streamline is improved and frontal area is reduced. This could be a reason why the waist-to-ankle suits are popular with males, who tend to sink more in the legs than females (see, for example, MacLean & Hinrichs, 1998; 2000). Toussaint et al. (1989) found that wetsuits worn by triathletes reduced drag by approximately 15%. This was attributed to their effect of shifting the centre of buoyancy away from the head and towards the feet. Thus, if bodysuits provide buoyancy and the buoyancy is distributed rearward, they may provide an advantage in a similar way. At the 2000 Olympic Games all male crawl-stroke gold medallists, other than Ian Thorpe (400m free) wore waist-to-ankle suits. Full bodysuits were shunned by Anthony Ervin, Gary Hall Jr., Pieter van den Hoogenband, and Grant Hackett. However, in the other three competitive strokes, bodysuits and suits-to-ankles were not nearly as popular as in crawl stroke events. It would seem, at least among males, that Bergen's findings of assistance for crawl and butterfly strokes is verified by the preferences of the best swimmers.

Figure 6. Underwater photographs of the Fastskin suit on the leg of a swimmer. The light refraction is due to air bubbles trapped in the fabric. The 'spots' are surface bubbles adhering to the fabric. Both are sources of floatation.

According to FINA-rules, devices that improve flotation are not allowed in competitive pool swimming. In that context swimwear manufacturers try to optimise their products by focussing on the reduction of surface drag. However, Rushall (2000a) cited pilot study results from Australia using the latest bodysuits (Speedo's 'Fastskin' and 'Adidas Equipment Bodysuit'). There is an initial buoyancy effect from the suits because it takes considerable time for the fabric to become saturated. In the meantime air trapped within and around the suit contributes to buoyancy. Until the fabric is saturated the suits aid buoyancy. The light refracted from the surface of the suits shown in Figure 6 is due to trapped air in the micro-channels.

Aleyev (1977) identified the possibility that tight suits might reduce drag by preventing large oscillating deformations of subcutaneous adipose tissue when swimming at higher speeds. It has also been proposed that the tightness of the suit may assist venous blood return.

Rushall (2000b) hypothesised that some swimmers might benefit from an indirect mechanical effect. Some swimmers have a technique flaw that causes excessive hip sway back and forth in crawl and backstroke, or rise and fall in butterfly. Excessive hip movements usually result from faulty stroke entries. These movements cause an increase in active drag, mainly from two sources: i) increased form drag, and ii) increased wave drag. Rushall proposed that because neck-to-knee bodysuits fit very tightly, hip sway is reduced. The reduction in movement range reduces these two forms of resistive drag, resulting in faster swimming for the same amount of energy expenditure. If there is, in fact, an improvement in performance due to the suits in the case of some swimmers, then it may be for this reason rather than due to a reduction on surface or form drag.

Rushall has also hypothesised that the suits might help some swimmers by having an added 'ergogenic' effect. The suits might maintain good posture and alignment to reduce form drag. If this were the case, then a swimmer would expend less muscular energy to maintain a streamlined posture. The energy freed from such a task could then be used to generate propulsion.

From observation of the various preferences of swimmers, it is clear that some swimmers favour the suits, but the body area covered by the suits varies among swimmers. Some swimmers believe that they are better off without the bodysuits. Thus, swimmers themselves are not convinced that the suits yield an advantage. It seems that swimmers need to be sure about what 'suits them'. Rushall (2000c) provides several interesting facts that have emerged from observation of various swimmers in different swim meets. For example, Michael Klim discarded a Speedo neck-to-ankle suit in favour of a Speedo waist-to-ankle suit to break his world 100m Butterfly record twice in three days.

Some swimmers complain that the suits remove 'feel' for the water. The loss of direct proprioception of the water could cause imprecise movements and a departure from their optimal technique. For this reason most swimmers have now rejected full versions of the suits, particularly the sleeves. Another reason for opting for versions without arms was the feeling that the suits restricted arm actions, making a high recovery more difficult. Tightness has been found to restrict movements around the shoulders and arms in all strokes, and around the knees in backstroke and breaststroke.

Although there has been insufficient research, there is a view that the muscle compression caused by suits may improve performance through increased muscle fibre recruitment. However, Rushall has identified that because some muscles would be affected and others not affected, the total movement pattern may be disrupted and efficiency reduced.

Professor Horacio Vielmo of the Federal University of Rio Grande do Sul, Brazil (personal communication, July 20, 2000), a prominent hydrodynamics scientist, evaluated the assistance promoted by Speedo's 'Aquablade' suit (1996 vintage). While Speedo promoted distinct advantages and figures for that suit, most of which were theoretically derived, in practical circumstances the 'old' Aquablade provided negligible assistance.

Rushall cited several reasons why the texture that is supposedly effective in reducing drag in sharks may not be effective for humans. The most profound of these is that a shark's body is streamlined with few protuberances to disrupt the flow of water during movement. In contrast, a human has a very irregular body surface. Thus, as it moves through the water it produces turbulent flow. Waring (1999) found that strategic placement of vortex generators can reduce form drag by minimizing the size of the separation bubbles behind the buttocks and in the small of the back of a submerged swimmer (Figure 7). However, the tests conducted in Waring's study applied only to a submerged swimmer in a constant glide posture such as that maintained for very short periods following starts and turns. The tests do not apply to swimmers actively using arms and legs, their movements creating irregular flow. The tests were also unsuitable for surface swimming where the effect of the air-water interface must be taken into account.

The above review indicates that there remains much to be learnt about whether bodysuits provide an advantage. It is clear that the transfer of manufacturers' 'results' to competitive performances has not occurred. The claims have not been vindicated by the performances of moderate to highly skilled performers. The advent of bodysuits has not resulted in a performance 'revolution' or any noticeable performance increase in any class of event.

While there is a rationale underlying these suits, whether the suits are effective in a real swimming situation is not yet established. The science underlying the design and production of bodysuits is particularly spurious. Similarly, the science refuting their value is sparse. Neither the case for or against has a solid footing. In general, scientists are sceptical of manufacturers' claims, and emerging studies seem to be siding with the scientists.

However, coaches and scientists do seem to agree on several aspects of bodysuit use. Those agreements lead to the following conclusions and recommendations.

1. When swimmers are unshaved and wear normal training swimsuits, freestyle and butterfly performances might improve in some individuals if they wear bodysuits.
2. It is unlikely that bodysuits will enhance racing performances in championship meets when swimmers are shaved and wear tight conventional racing suits.
3. Backstroke and breaststroke swimming is not enhanced by bodysuits, whether a swimmer is unshaved or shaved.
4. Some individuals will be assisted by bodysuits. The determination of assistance should only be made after careful testing.
5. Once bodysuits become wet, they contain more water than do tight conventional suits, and are likely to cause a swimmer to go slower, rather than faster, particularly in races longer than 200 m.
6. Some bodysuits, because of coatings on their fabrics, will take longer to "get wet" than others, but even they will eventually suffer the same problems as those that wet quickly.
7. Particularly tight bodysuits could hamper the range of movement at important joints, such as the shoulders, hips, and knees, and consequently, will hinder the correct execution of turns and dives.

If you are a swimmer contemplating purchasing one of these devices, suit yourself, but think about it first and be sure that the bodysuit suits you.

1. Aigeldinger, T. L., & Fish, F. E. (1995). Hydroplaning by ducklings: Overcoming limitations to swimming at the water surface. Journal of Experimental Biology, 198, 1567-1574.
2. Aleyev, Y. G. (1977). Nekton. Dr W. Junk BV, Den Haag.
3. Armitage, P. (1977). Statistical methods in medical research. Blackwell scientific publications, Oxford.
4. Bergen, P. (February, 2001). Coach Paul Bergen's tests of bodysuits. Bodysuits: The Serious Threat to the Very Nature of Competitive Swimming. (https://www-rohan.sdsu.edu/dept/coachsci/swimming/bodysuit/bergen.htm).
5. Cappaert, J., & Rushall, B. S. (1994). Biomechanical analyses of champion swimmers. Spring Valley, CA: Sports Science Associates.
6. Chatard, J. C., Lavoie, J. M., Bourgoin, B., and Lacour, J. R. (1990). The contribution of passive drag as a determinant of swimming performance. International Journal of Sports Medicine, 11, 367-372.
7. Gould, S. (2000). Testimonies about bodysuits. Bodysuits: The Serious Threat to the Very Nature of Competitive Swimming. (https://www-rohan.sdsu.edu/dept/ coachsci/swimming/ bodysuit/testimon.htm).
8. Hertel, H. (1966). Structure, Form and Movement. Reinhold, New York.
9. Imhoff, F., & Pranger, L. (1975). Boat tuning for speed. Lymington, England: Nautical.
10. Johns, R. A., Houmard, J. A., Kobe, R. W., Hortobagyi, T., Bruno, N. J., Wells, J. M., & Shinebarger, M. H. (1992). Effects of taper on swim power, stroke distance, and performance. Medicine & Science in Sports & Exercise, 24, 1141-1146.
11. Kolmogorov, S. V., & Duplishcheva, O. A. (1992). Active drag, useful mechanical power output and hydrodynamic force coefficient in different swimming strokes at maximal velocity. Journal of Biomechanics, 25, 311-318.
12. McLean, S. P., & Hinrichs, R. N. (1998). Sex differences in the centre of buoyancy location of competitive swimmers. Journal of Sports Sciences, 16, 373-383.
13. McLean, S.P., & Hinrichs, R.N. (2000). Buoyancy, gender, and swimming performance. Journal of Applied Biomechanics, 16, 248-263.
14. Pai, Y., & Hay, J.G. (1988). A hydrodynamic study of the oscillation motion in swimming. International Journal of Sport Biomechanics, 4, 21-37.
15. Rushall, B. S. (2000a). The bodysuit problem: What the scientists report. World Wide Web: www.rohan.sdsu.edu/dept/coachsci/swimming/bodysuit/science.htm.
16. Rushall, B. S. (2000b). The bodysuit problem: How bodysuits help some swimmers. World Wide Web: www.rohan.sdsu.edu/dept/coachsci/swimming/bodysuit/helpsuit.htm.
17. Rushall, B. S. (2000c). A matter of concern: What is known about these devices: Emerging knowledge and concerns about the use of 'condom' suits. World Wide Webwww.rohan.sdsu.edu/dept/coachsci/swimming/bodysuit/knowsuit.htm.
18. Sanders, R. H., Cappaert, J. M., & Pease, D. L. (1998). Wave characteristics of Olympic breaststroke swimmers. Journal of Applied Biomechanics, 14, 40-51.
19. Sanders, R. H. (1985). Can skilled performers readily change technique? An example, conventional to wave action breaststroke. Human Movement Sciences, 14, 665-679.
20. Sanders, R. H., Cappaert, J. M., & Devlin, R. (1995). Wave Characteristics of butterfly swimming. Journal of Biomechanics, 28, 9-16.
21. Sanders, R. H. (1999). Hydrodynamic characteristics of a swimmer's hand. Journal of Applied Biomechanics, 15, 3-26
22. Sharp, R. L., & Costill, D. L. (1989). Influence of body hair removal on physiological responses during breaststroke swimming. Medicine and Science in Exercise and Sports, 21, 576-580.
23. Sharp, R. L., & Costill, D. L. (January, 1990). Shaving a little time. Swimming Technique. (pp. 10-13).
24. Speedo Press Release. (2000). Fastskin fact sheet. https://www.speedo.com/.
25. Toussaint, H. M., Groot, G. de, Savelberg, H. H. C. M., Vervoorn, K., Hollander, A. P., & Ingen Schenau, G. J. van. (1988a). Active drag related to velocity in male and female swimmers. Journal of Biomechanics, 21, 435-438.
26. Toussaint, H. M., Beelen, A., Rodenburg, A., Sargeant, A. J., Groot, G. de, Hollander, A. P., & Ingen Schenau, G. J. van. (1988b). Propelling efficiency of front crawl swimming. Journal of Applied Physiology, 65, 2506-2512.
27. Toussaint, H. M., Bruinink, L., Coster, R., Looze, M. de, Rossem, B. van, Veenen, R. van, & Groot, G. de (1989) Effect of a triathlon wet suit on drag during swimming. Medicine and Science in Sports and Exercise, 21, 325-328.
28. Toussaint, H. M., Hollander, A. P., Berg, C. van den, & Vorontsov, A. (2000) Biomechanics of swimming. Exercise and Sport Science (Edited by Garrett, W. E. and Kirkendall, D. T.), pp. 639-660. Lippincott, Williams & Wilkins, Philadelphia.
29. Toussaint, H. M., Truijens, M., Elzinga, M-J., Ven, A. van de, Best, H de, Snabel, B., & Groot, G. de. Effect of a Fast-skin 'body' suit on drag during front crawl swimming. Sport Biomechanics, 1(1) (in press).
30. Vaart, A. J. M. van der, Savelberg, H. H. C. M., Groot, G. de, Hollander, A. P., Toussaint, H. M., & Ingen Schenau, G. J. van. (1987). An estimation of active drag in front crawl swimming. Journal of Biomechanics, 20, 543-546.
31. Vogel, S. (1996). Life in Moving Fluids: The Physical Biology of Flow (2nd ed.). Princeton, NJ: Princeton University Press, (p. 153).
32. Waring, J. (1999). Reduction of drag of a submerged swimmer using vortex generators. Carleton University, Canada.

Return to Table of Contents for The Bodysuit Problem.