299BrJ Sports Med 1997;31:299-303

Original articles

Hazard of deceptive advertising of athleticfootwear

Steven Robbins, Edward Waked

AbstractObjectives-Athletic footwear are associ-ated with frequent injury that are thoughtto result from repetitive impact. No scien-tific data suggest they protect well. Expen-sive athletic shoes are deceptivelyadvertised to safeguard well through"cushioning impact", yet account for123% greater injury frequency than thecheapest ones. This study tested thehypothesis that deceptive advertising cre-ates a false sense of security with users ofexpensive athletic shoes, inducing attenu-ation of impact moderating behaviour,increased impact, and injury.Methods-Fifteen young healthy male vol-unteers confronted four surfaces: a bareforce moment platform, and three with thisplatform covered by identical shoe solematerial made to appear different andadvertised divergently. Advertising mes-sages suggested superior impact absorp-tion and protection (deceptive message),poor impact absorption and high injuryrisk (warning message), and unknownimpact absorption and safety (neutral mes-sage). Ground reaction forces were re-corded for 10 barefoot footfalls, accordingto a protocol requiring stepping forwardfrom perch to a surface 4.5 cm below.Results-Impact varied as a function ofadvertising message (p<0.001). Deceptivemessage equalled neutral message in elic-iting higher impact than the warningmessage and the bare platform. Differ-ences grew with repetitions (p<0.001).Conclusions-These data provide a plausi-ble mechanism explaining higher injuryfrequency in users of expensive athleticshoes. This is the first report to suggest: (1)deceptive advertising of protective devicesmay represent a public health hazard andmay have to be eliminated presumablythrough regulation; (2) a tendency inhumans to be less cautious when using newdevices of unknown benefit because ofoverly positive attitudes associated withnew technology and novel devices.(BrJ Sports Med 1997;31:299-303)

Keywords: sports injuries; athletic footwear; impact;running; advertising; product safety; warning labels

Research dealing with product advertising hasfocused on consumer preferences-it is notknown if advertising influences how consumersuse products after purchasing them. Advertis-ing tends to be persuasive when dealing withopaque attributes but less effective whenpromoting easily appreciated properties.'Safety provided by products designed toprotect is usually obscure to users, thereforeadvertising products as offering improvedsafety has the potential of being convincing.We consider advertising of protective devices

to be truthful when the message suggestsprotection consistent with data from objectiveproduct testing under conditions of use beforeor after marketing of a product, whereasdeceptive advertising implies protection signifi-cantly exceeding equivalent data. Suggestinggood protection against injury may convincethe user that cautious behaviour previouslyexercised is less necessary. Attenuated cautioncould affect safety differently when dealingwith products that are either truthfully ordeceptively advertised. Protective equipmentproved capable of protecting before advertis-ing, and truthfully advertised as effective, mayprove less beneficial or perhaps even ineffectiveonce safety benefits are disclosed to the public,because users may conclude that cautionpreviously exercised is less necessary. Productsdeceptively advertised as performing well, thatare actually ineffective, might attenuate cautionsufficiently to heighten injury frequency abovelevels existing before their introduction.

Protective athletic equipment represents avast group of items. Perhaps no item in thisgroup is advertised as extensively as athleticfootwear, which are all promoted as capable of"cushioning" impact via sole materials. Impactamplitude, which can be measured objectivelywith precision, is poorly appreciated byhumans when they wear athletic footwear,2therefore advertising this attribute presumablywould be convincing.No athletic footwear, regardless of manu-

facturer and price range, have ever been shownto protect well against injuries, thereforeadvertising good protection is deceptive.Expensive footwear are subject to extremelydeceptive advertising. They are advertised toimprove protection over cheaper products by

McGill University,Centre for Studies inAging, Montreal,CanadaS RobbinsE Waked

Correspondence to:Dr S Robbins, 800 ReneLevesque West Suite 20-10,Montreal, Quebec, H3B IX9Canada.

Accepted for publication22 July 1997

group.bmj.com on July 31, 2016 - Published by http://bjsm.bmj.com/Downloaded from

Robbins, Waked

incorporating new features that protect,and more advanced safety technology, yetepidemiological data indicate that users ofmore expensive shoes are injured morefrequently. Expensive shoes accounted for123% greater frequency of injury than thelowest cost models, regardless of manufac-turer, after correcting for previous injury,training mileage, and sex, factors influencinginjury frequency.`5 This suggests that impactis greater when consumers use more expensiveathletic shoes, and we can think of nothingother than the effect of advertising that mightaccount for it.As an explanation of the higher injury

frequency when humans use expensive athleticshoes, we hypothesise that deceptive advertis-ing suggesting good protection against impactpersuades users to eschew cautious behaviour,resulting in higher impact. We tested this in anexperiment in which the independent variablewas the advertising message deceptively de-scribing the same shoe sole material made toappear different as either effective protection,unsafe, or of unknown safety. The dependentmeasure was impact.

MethodsSUBJECTSA random sample of 15 healthy young menwith a mean (SD) age of 31 (4.2) years partici-pated in this study on a voluntary basis. Allwere selected from the general population andparticipated actively in sports and leisureactivities. Mean (SD) body mass and heightwere 75 (9.2) kg) and 175 (7.8) cm, respec-tively. All were free of conditions affecting theirability to walk, run or balance.

SOLE INTERFACE MATERIALThe sole interface material covering the topsurface of the force moment platform consistedof a 2.5 cm thick layer of ethyl-vinyl acetate(EVA), Shore Al0 in hardness. EVA is thepolymer material most commonly used infootwear soles. Sole conditions differed only inthe colour of the fabric encasing the EVA solematerial. This envelope ensured subjects couldnot identify the material within it.

MESSAGES FOR SHOE SOLE ADVERTISING

Deceptive messageThis message resembled advertisements usedto promote athletic footwear in periodicalsavailable to us. It included pseudoscientificgraphs and tables, and athlete endorsementsuggesting superior impact absorption and sta-bility. We used the terms "advanced technol-ogy" and "state of the art", to describe the sole.We stated that this technology is incorporatedin the most expensive athletic shoes currentlyon the market.

Warning messageThis message was based on pre-experimentalevaluation of user opinions, indicating thatusers believed that athletic shoes were effectivein protecting against injury, that expensiveshoes protected better than cheap ones, andthat new technology has resulted in more pro-

tective shoes. We concluded that contrastingnotions from the deceptive message wouldconstitute warning. The shoe sole was de-scribed as associated with frequent injuriesfrom excessive impact and instability. Subjectswere told that this material was old technologyand used only in cheap running shoes.

Neutral messageThis shoe sole material was described as neverused in shoe soles previously, and impossible topredict how it might influence either impactabsorption or stability.

FORCE MOMENT PLATFORMA unique force moment platform was used.This force moment platform is 500 mm' in areaand 150 mm high. It is capable of measuringpeak loads of up to 10 000 N with a resolutionof 1 N for forces and 0.1 Nm for moments. Themaximum acquisition/ transmission rate foreach load sensor is 500 Hz. A high speed (115kbaud) serial link transmits data from forcemoment platform to host computer, which cal-culates various performance indices andrecords data to disk. A graphical user interfaceallows operator to view time history of theforces, moments, and centre of force.

TESTING PROTOCOLWritten consent was obtained both beforebeginning the experiment and followed theethical guidelines for human experimentationset by the Helsinki Declaration. Subjects wereconfronted with four surface interface condi-tions in a counterbalanced order: a bare rigidforce moment platform, and three with thisplatform covered by identical impact absorbingshoe sole material. The advertising messageswere presented in writing and graphics. Beforeeach surface condition, the advertising messagewas read by the subject and re-read by thetechnician. No message was presented for thebare force moment platform. Impact testingconsisted of 10 consecutive barefoot footfallimpacts recorded on the force moment plat-form for each surface condition. Subjects wererequired to step forward from perch to asurface 4.5 cm below, and land on one foot, asshown in figure 1. Subjects were debriefed atthe end of the experiment to assess any subjectbias regarding the experiment and its protocol.Included in the debriefing were questions con-cerning the credibility of claims made bymanufacturers of modern athletic shoes con-cerning protection against injuries, andwhether subjects were aware of any reports oradvertisements contrary to these claims.

DATA ANALYSISOne way analysis ofvariance was performed foreach dependent measure using the three mate-rial conditions as independent variables. Posthoc tests were performed for significant F val-ues only. Differences between the bare plat-form surface condition and the three othermaterial surface conditions were assessed usingt test with Bonferroni adjustment to correct fortype I error. Statistical significance for all testswas set at 0.05.

300

group.bmj.com on July 31, 2016 - Published by http://bjsm.bmj.com/Downloaded from

Hazard of deceptive advertising of athletic footwear

1 KItS

Figure I Schematic representation of experimental set up for dynamic footfall testing. Subject stepped down onto forcemoment platform, which was either bare, or covered by athletic shoe sole material. He balanced on one foot afterfoot contactwith platform. Subjects were barefoot with gaze straight and arms to side.

ResultsImpact varied in relation to advertising claims(F (2, 28)=96.87, p<0.001). Impact was great-est with the deceptive message and lowest withthe warning message, with means of 121% and110% body weight (BW), respectively. Neutralmessage, with a mean impact of 1 17% BW, didnot significantly differ from the deceptive mes-sage, yet was significantly greater than thewarning message. Mean impact with the bareforce moment platform was 108% BW (fig 2)This was significantly lower than the deceptiveand neutral messages, but not significantlylower than the warning message. There was a

o Deceptive messageA Warning messageo Neutral messageo Bare platform

10 1 5

Trial

Figure 2 Mean footfall impact values with standard error bars for the 10 consecutivetrials for each of the three surface interface conditions and the rigid bare platform. Footfallimpact values stabilised after the fifth trial, except for the rigid platform, which stabilisedafter the third trial.

significant interaction effect between messageand trial (F (18, 252)=5.16, p<0.001). Posthoc tests indicated that with each subsequenttrial, footfall impact significantly increasedwith the deceptive message, and significantlydecreased with warning message, yet no changeoccurred with the neutral message.

DiscussionThe notion that cumulative microtrauma fromimpact experienced during running and jump-ing causes injury in humans is supported byepidemiological reports indicating that everysport involving running is associated withfrequent lower extremity injury,5-8 and experi-mental studies showing that in animals, repeti-tive microtrauma from impact of locomotionaccounts for injury in relation to cumulativeexposure.9 '0 Humans are capable of moderat-ing impact dramatically through behaviours,such as knee and hip flexion at contact of sup-port surface..'. Amplitude of impact whengymnasts land impulsively has been shown tobe directly related to hip and knee flexionamplitude."15Attempts at moderating excessive impact

during running and jumping through externaldevices, such as yielding interfaces (for exam-ple, soft athletic footwear and mats), hasproduced disappointing results. Impact hasbeen shown to be higher when gymnasts landon mats than on hard surfaces, becausehumans greatly reduce flexion when landing onsoft mats." '5 It is not clear if impact in humansis higher when wearing soft soled athletic foot-wear than when barefoot, although it is not sig-nificantly lower. This is because humansreduce flexion when soft soled shoes are usedthereby at least neutralising any advantage thatyielding materials offer in terms of impactabsorption.2"11151' It is now believed thathumans land rigidly when soft materials areunderfoot in an attempt to counter theinstability that these interfaces induce.'7 Thereis no evidence that athletic footwear differ sig-nificantly in impact absorption in relation tomanufacturer, model, or cost, when testedblindly using force moment platform or accel-erometer technology during normal running."

13

2.5 -

12 -

1.5 -

1 1 _

0.5 -

.jnI

z

0m

E

11

5

301

I I

d

1;:

1

1

group.bmj.com on July 31, 2016 - Published by http://bjsm.bmj.com/Downloaded from

Robbins, Waked

This finding is expected as all incorporate solesmade of similar materials, and produce compa-rable impact enhancing instability.'8-2' Thereare no data supporting the notion that athleticfootwear are capable of protecting againstinjury through cushioning of impact.Without any significant difference in impact

absorption between footwear, one would an-ticipate that users of athletic footwear wouldexperience no difference in injury frequency inrelation to model and manufacturer. It comesat some surprise however that Marti found thatexpensive shoes accounted for 123% greaterinjury frequency than lowest cost models,regardless of manufacturer, after correcting forprevious injury, training mileage, sex, andalmost everything hypothesised to be associ-ated with injury.5 Furthermore, runners whoexpressed strong brand preference had asignificantly greater chance of being injuredthan those who had no brand loyalty. This sug-gests that footwear products may be useddifferently in relation to cost based on expecta-tions of protection as communicated throughadvertising. As an explanation of injury fre-quency in relation to athletic footwear cost, wehypothesised that advertising affects usercaution, thereby inducing users to adjustimpact moderating behaviour in relation toperceived risk inferred by the advertisingmessage-higher impact with expensive mod-els that are assumed to be safe, and lowerimpact with inexpensive models that arepresumed dangerous.

In this experiment impact amplitude variedas a function of advertising message, with thewarning message resulting in lower impactthan the deceptive one, 110% BW versus 121%BW, respectively (p=0.03 1). With repetitivefootfalls, impact amplitude declined and in-creased in relation to the warning and decep-tive message, respectively (p<0.001). This sup-ports our hypothesis that footwear advertisingchanges usage behaviour along lines account-ing for higher injury frequency with moreexpensive shoes and lower injury frequencywith inexpensive shoes.The bare platform resulted in lowest impact

of all conditions, though not significantlydifferent from the warning message (bare,109% BW; warning, 1 0% BW, p=O. 15). Dur-ing debriefing subjects reported that landingon the bare platform was uncomfortable,whereas the sensation associated with landingon any of the sole conditions was morepleasant. As the warning message was equiva-lent to painful consequences in inducingimpact moderating behaviour, we concludethat use of warning messages is a highly effec-tive means of eliciting cautious behaviour.The deceptive message resulted in highest

impact, though not significantly greater thanwith the neutral message (deceptive, 121%BW; neutral, 117% BW, p=0. 16). Equivalencyof neutral and deceptive message is explainedby information obtained during debriefing ofsubjects. Higher impact with the deceptivemessage is consistent with our hypothesis thatattenuated caution is based on a false sense ofsecurity resulting from deceptive advertising

messages. This is further aided by pre-existingbeliefs subjects reported during debriefing thatassociates protection against injury with expen-sive footwear and little or no protection withcheaper ones, notions based on previousdeceptive advertising of athletic footwear andfalse information disseminated by retail out-lets. In addition, many subjects harboured falseimpression about truth in advertisingregulations-most believed companies wererequired to present a truthful assessment ofproduct performance in advertisements.We believe that higher impact with neutral

message is accounted for by general irrationalfaith in new products. Subjects believed thatnew products, particularly safety products,always perform better than older ones. Thispositive attitude about novelty seemed based inpart on the mistaken impression that govern-ment regulations require all new safety prod-ucts to be tested for proof of satisfactoryperformance under conditions of use beforebeing available to the public. In addition, somepositive notions about new athletic footwearseem to be generalisations based on brand loy-alty, which in turn was based on deceptiveadvertising of other footwear products. This isunderstandable considering that humans canbe easily mislead by advertising that deals withproduct attributes that are opaque tousers.' 22 23

Practical application of our data relate to twoconditions: protective devices truthfully anddeceptively advertised. This report is the firstto imply that safety products truthfully adver-tised to protect effectively before being com-mercialised might be less effective than antici-pated after commercialisation, because of astrong tendency in humans to become lesscautious when they use products they believeprotect better. This is also the first report tosuggest that deceptive advertising of protectivedevices is a clear public health hazard, becausehumans will become less cautious and mayinjure themselves more frequently when theyuse them. We believe diminished vigilancecaused by this advertising accounted for 123%greater injury frequency in users of expensiverunning shoes that Marti reported, and prob-ably continues to be responsible for innumer-able injuries.5 Deceptive advertising of protec-tive devices must be eliminated.Our report suggests that a simple and direct

solution to this tendency in humans to be lesscautious when using new safety products lies inadequate warning labelling, as we found thatwarning message was highly effective ineliciting cautious behaviour in this experiment.Public health would be advanced if truthfullyadvertised protective devices possess a promi-nent explicit warning label affixed to themexplaining to users that protection againstinjury with this product is only realisable if theyremain as cautious with the new product asthey were with the one it replaced. The resultsofwarning labelling may be impressive, consid-ering with respect to expensive running shoes,we estimate that this labelling could result in aninjury frequency reduction of 55%.

302

group.bmj.com on July 31, 2016 - Published by http://bjsm.bmj.com/Downloaded from

Hazard of deceptive advertising of athletic footwear 303

One limitation of this experiment is thatexperimental conditions differed from naturalconditions-stepping off a perch differs fromhow footwear products are used during loco-motion. Shock absorbing materials have beenshown to influence impact similarly whethertested by humans stepping from a perch orduring actual locomotion. For example,McNitt-Gray found during landings from aplatform,"5 and Nigg during running,'6 thatimpact increases as interfaces become softer.This suggests that this limitation should notinfluence our conclusions.

In summary, pre-existing beliefs of protec-tion against injury formed through deceptiveadvertising of athletic footwear attenuates usercaution thereby amplifying impact. This mayexplain the 123% higher injury frequencyassociated with more expensive shoes.5 Accord-ingly, deceptive advertising of protective de-vices is a public health hazard and must beeliminated. Humans are less cautious evenwhen they use truthfully advertised productsbecause of excessively positive attitudes towardnew products and wrong impressions of thestandards of truth in advertising throughgovernment regulation.

1 Snyder R. Misleading characteristics of implied superiorityclaims. Journal ofAdvertising 1989; 18:54-61.

2 Robbins SE, Hanna AM, Jones LA. Sensory attenuationinduced by modern athletic footwear. Journal of Testing andEvaluation 1988;16:412-6.

3 Gross TS, Bunch RP. Material moderation of plantarimpact stress. Med Sci Sports Exer 1989;21:619-24.

4 Hamill J, Bates BT. A kinetic evaluation of the effects ofin-vivo loading on running shoes. J Orthop Sports Phys Ther1988;10:47-53.

5 Marti B. Relationship between running injuries and runningshoes - Results of a study of 5000 participants of a 16-kmrun - The May 1984 Berne "Grand Prix". In: Segesser B,Pforringer W, eds. The shoe in sport. Chicago: Year BookMedical Publishers, 1989: 256-65.

6 Brunet ME, Cook SD, Brinker MR, Dickson JA. A survey ofrunning injuries in 1505 competitive and recreational run-ners. J7 Sports Med Phys Fit 1990;30:307-15.

7 Clement DB, Taunton JE, Smart GW, McNicol JL. A surveyof overuse running injuries. Physician Sportsmed 1981 ;9:47-58.

8 James SL, Bates BT, Osternig LR. Injuries to runners. Am JSports Med 1978;6:40-50.

9 Dekel S, Weissman SL. Joint changes after overuse and peakoverloading of rabbit knees in vivo. Acta Orthop Scand1978;49:519-28.

10 Hellman DB, Helms CA, Genant HK. Chronic repetitivetrauma: a cause of atypical degenerative joint disease. Skel-etal Radiol 1983;10:236-42.

11 McNitt-Gray JL, Yokoi T. The influence of surface charac-teristics on the impulse characteristics of drop landings.Proceedings of the 13' Annual Meeting of the American SocietyofBiomechanics. Vermont: American Society ofBiomechan-ics, 1989: 92-3.

12 Robbins SE, Hanna AM. Running related injury preventionthrough barefoot adaptations. Med Sci Sports Exer 1987; 19:148-56.

13 Robbins SE, Hanna AM, Gouw GJ. Overload protection:avoidance response to heavy plantar surface loading. MedSci Sports Exer 1988;20:85-92.

14 Robbins SE, Hanna AM, Gouw GJ. Running related injuryprevention through innate impact moderating behaviour.Med Sci Sports Exer 1989;21:130-9.

15 McNitt-Gray JL. Landing strategy adjustments made byfemale gymnasts in response to drop height and mat com-position. J Appl Biomech 1991 ;9: 173-90.

16 Nigg B. Biomechanics of running shoes. Chicago: HumanKinetics Publishers, 1986.

17 Robbins S, Waked EG. Humans amplify impact to compen-sate for instability caused by shoe sole materials. Arch PhysMed Rehabil 1997;78:463-7.

18 Cook SD, Kester MA, Brunet ME. Shock absorption char-acteristics of running shoes. AmJ7 Sports Med 1985;13:248-53.

19 Robbins SE, Waked EG, Gouw GJ, McClaran J. Athleticfootwear affect balance in men. Br J Sports Med1994;28:117-22.

20 Robbins SE, Gouw GJ, McClaran J. Shoe sole thickness andhardness influence balance in older men. JAm Geriatr Soc1992;40: 1089-94.

21 Robbins SE, Waked EG, Allard P, Krouglicof N, McClaranJ. Aging in relation to optimization of footwear in oldermen. jAm GeriatrSoc 1997;45:61-7.

22 Marks LI, Kamins MA. The use of product sampling andadvertizing: effects of sequence of exposure and degree ofadvertizing claim exaggeration on consumers' beliefstrength, belief confidence, and attitudes. Journal ofMarketResearch 1988;25:266-81.

23 Smith RR. Integrating information from advertizing andtrial: process and effects on consumer response to productinformation. Journal ofMarket Research 1993;30:204-19.

group.bmj.com on July 31, 2016 - Published by http://bjsm.bmj.com/Downloaded from

footwear.Hazard of deceptive advertising of athletic

S Robbins and E Waked

doi: 10.1136/bjsm.31.4.2991997 31: 299-303 Br J Sports Med 

http://bjsm.bmj.com/content/31/4/299Updated information and services can be found at:

These include:

serviceEmail alerting

the box at the top right corner of the online article. Receive free email alerts when new articles cite this article. Sign up in

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:

group.bmj.com on July 31, 2016 - Published by http://bjsm.bmj.com/Downloaded from