MEDICINE AND SCIENCE IN SPORTS AND EXERCISE
Copyright 10 1992 by the American College of sports Medicine
Vol. 24, No.1
Printed In U.S.A.
LETTERS TO THE EDITOR-IN-CHIEF
Drs. Robbins and Gouw hav e chosen to raise the
issue of the conflict of interest policies of the Journal
and imply that the Journal should police letters written
to the Editor-in-Chief. In the Informa tion for A uthors
section of this issue of the Journal, the " conflict of
i nterest" regarding published peer reviewed manuscripts
has been addressed after receiving editorial board
approv al and reflects a y ear-long rev iew of the policy. I
reject Drs. Robbins and Gouw' s expectation that the
Editor-in-Chief should impose this policy on letters to
Historically, I have had to respond to a letter of
complaint from Robbins and Gouw c onc erni ng one of
their rejected manusc ripts. I n this complaint they a lso
raised t he issue of c onflic t of in te rest by t he rev iew ers,
as being the reason for the rejecti on of the ir manus cript,
even though they w ere bli nd as to whom the rev ie wers
were. Neither Drs. F re deri ck or C av anagh were reviewers
on the re jected manusc ript. Furthermore, I rev iew ed
their conc erns regarding t he ir manusc ript a nd found
thei r charges of conflict of i ntere st t o be unf ounded.
Clearly, these two l etters need to be r ead with a
critical ey e to the sci ence bei ng dis cuss ed, and one needs
to ref rain from bein g draw n into pers onal differe nce s.
I c an assur e the reader s that the se two le tters will be the
only ones publis hed o n the issues raised by Drs. Frederick
Dr. Steven Robbins has been republishing the same
editorial seasoned with a pinch of new data for the last
several years, and we are distressed to see y et another
appearance in MSSE CMSSE 23:2 17, 1991). Dr. Robbins'
main contention is that sport shoes affect the
pattern of movement in a way that is unsafe. He
presents data showing the effect of shoe sole materials
on the mechanics of mov ement. Although Robbins
wishes to extrapolate his data to actual running, his
results are collected in a laboratory setting that involves
contrived quasi-static movements and atypical limb
postures rather than actual studies of runni ng mechanics.
Further t o the poin t. Dr. Robbins' arguments re st
heavily on the notion that informati on from cutaneous
receptors are the only data that the body uses to monitor
potentially damaging ;)trC ;);)CT5.his assertion is untenable
and unproven. These facts combine d with the inappropriate
kinetics used in Robbins' c xpcr ime nte conspires
to make his results of limite d use in an understanding
of running mecnamcs, or any other vlaorous
Some runner snectnc technical Ilmitat ions we o bserved
of which we fed y our reade rs should be aware.
The spec ification of me chanical inputs in units of kg.
crr r ? is misle ading. This is presumably the pressure in
the authors' pneumatic actuators and is certainly not
the pressure applied to any part of the body . To q uote
these figures in the abst ract is meanin gless since the
result of s uch an input is s pecifi c to t he re searcher' s
apparatus. Thi s confusion seems to persist as the authors
go on t o ex trapolate about possible vertical loads
in r unni ng of 1.03 kg- cm" . T he pla ntar pressures
duri ng actual run ning are at least 10 time s this v alue
when measured on the bare foo t. T here is also some thing
peculiar about the straight lines in Figure 4 and
the das hed ext rapolations into load ra nges "s ee n in
The opening line of the discus sion is a blatant misstatement:
"This experiment relates pla ntar lo ad during
locomotion and jump ing and plantar discomfort w hen
shod and unshod." It does nothing of the ki nd.
Despite the limited usefulness of his data, researcbers
like us, who study footwear biomechanics, ar e generally
intrigued by many of Dr. Robbi ns ob serv ations about
footwear-induced kinematic changes, bUI we s trongly
disagree w ith his as se rtion tha t sport shoes hav e been
shown to be unsafe, or less injurious than barefoot
It is impor ta nt to point out that his conclucion that
wear ing ehoee i s " unsafe" is not base d on hi s own data,
but is largely bncsd on ~ selective and rather c areless
reading of the literature on me ernuermotoav oroveruse
injuries in sport. F urthermore, Robbins bla tant ly miscites
sev eral biomec hanica l st udies as supporti ng his
LETTERS TO THE EDITOR-IN-CHIEF
contention that sport shoes are unsafe. No data are
presented nor are any conclusions drawn. in any of the
biomechanical studies cited by Robbins on the question
of the safety of sport shoes.
Dr. Robbins misrepresentation of the literature he
cites is almost universal. A few of many examples:
Robbin s cites Casperson et a l. (Med. Sci. Sports Exerc.
16:113, 1984) vs Marti et al. (Am. J. Sports Med.
16:285-294, 1988) to show a trend toward an increase
in running injuries over the y ears, when in fact these
studies cover overlapping time periods. Casperson's
work (a very brief abstract that did not mention footwear)
was published in 1984 and covers participants in
a 1983 race. And, although the Marti research Dr.
Robbins cites in support of his contention that shoes
are uns afe was published in 1988, in fact, the surv ey
was done in 1984 and was based on running injuries
experienced during 1983-84.
Further to the point, Marti' s research presents no
data to show that shoes are unsafe, but he did find a
statistically signific ant "preventive effect" from the use
of shoe inserts-a point that clearly contradicts Robbins'
assertions, but which he conveniently overlooks.
Dr. Robbins further cites Marti as showing that runners
who wear more expensive shoes (>$40) are more frequently
injured, but Marti c learly points out that this
is an " incorrect" conclusion. Only 2.7% of his sample
w ore shoes under $40 in cost, raising the suspicion that
the finding of fewer injuries in this small group is an
artifact due to age and income differences (i.e., Marti
reports that runners wearing these less expensive shoes
tended to be young people and therefore more free of
This kind of inat tention to detail and a willingness
on Dr. Robbins part to overlook evidence that does not
support his point of view is a thread that runs through
this paper and his other recent publications. This is not
an acceptable practice, and we believ e we would have
hoped that such careless and irresponsible science
would have been uncovered during the review process.
Furthermore, it is our opinion that the title of Dr.
Official Journal of the American College of Sports Medicine 145
Robbins' article was misleading and should not have
gone unchallenged during the rev iew pro cess. He refers
to athle tic shoes in the title as "unsafe." His report,
however, has nothing 10 do with safety, has little to do
with athletic shoes, and presents no data that can be
interpreted as even indirectly supporting s uch a conclusion.
No original data on sports shoe safety are presented,
only data on artificial motor behaviors that
hav e not been shown to occur in actual running and
which have not been shown to be unsafe by him or any
other researcher. In light of the absence of any data to
support Dr. Robbins' conclusions, we are troubled by
his outrageous allegation that manufacturers of athletic
footwear are "irresponsible" for representing their products
as offering improved protection. In fact, he further
emphasizes the absurdity of his a llegations by attacking
existing "safety standards" for footwear, when in fact
everyone who works in this area knows that there are,
as yet, no such standards.
The biomechanical interactions between the body
and the shoes and surfaces ove r whic h we run is an
important topic for study . It promises to lead us to new
ideas about the causes of overuse injuries, and it may
result in new strategies for their treatment and prevention.
But such careless and nonscientific soap-box oratory
will not bring us any of these positive benefits, and
in fact may result in more harm and misunderstanding.
As a result of his research, Dr. Robbins makes the
outrageous and untenable recommendation of "barefoot
activity-where practical and socially acceptable."
As readers of this widely read journal, we simply can
not let stand such a radical and potentially harmful
conclusion in the advance of any data to support it.
E. C. Frederick, Ph.D.
Exeter Research, Inc.
Peter R. Cavanagh, Ph.D.
Professor and Director
Center for Locomotion Studies
Pennsy lv ania State University
T his response to Frederick and Cavanaan's le tter
provides !l lone awaited opportunity to deal with c on·
tltct of interact M it pertains to scientific; publications.
which M!: concerned us for some time, We first will
de!!] with specific points they raise.
It should be made clear that to in validate our position
that the modern running shoe is unsafe, two conditions
must be met. First, data must be produced suggesting a
lower injury frequency with modern shoes, lower impact
during running with these footwear, or showing
expensive shoes (advertised as offering increased OfOtection:
e.g., increased cushioning; "pronation corrccnon")
protect better than inexpensive ehocs. Frederick
and C avanagh render no references suggc;:,tingthis be cause
Second, they must provide conv incing arg uments
that our citations and data suggesting that these footwear
are hazardous, and our causality theory are in
146 Official Journal of the American College of Sports Medicine
error. We provided 15 references indicating safety COncerns,
and 16 that support our causality theory (6).
Despi te their inflated rhetoric, they question only two
(1,3) of these 31 citations. Evidently they found no fault
with the remaining 29, because if further data were
available to fortify thei r anemic arguments they surely
would have used them. For example, they concede
impact during running is at its highest when runners
use the softest soled shoes (5), and impact is invariably
20-25% greater when gymnasts land on thick soft mats
when compared with hard surfaces (4). We encourage
readers to seek out these citations.
As for the two references they question, they contend
we mislead because the time interval between Caspersen
et al.'s (1) and Marti et al.'s (3) collection of data was
shorter than what might be assumed from the publication
dates. No statement in our report specified this
interval because Caspersen et al.'s report, which we
(and they) refer to does not mention dates (6). Since
Frederick and Cavanagh provide no reference offering
this date, this claim can be dismissed.
They argue that we do not report references that are
contrary to our position-an audacious allegation COnsidering
no example follows. We (and perhaps they) are
aware of no published report that does not support our
Frederick and Cavanagh's assessment of Marti's work
is fiction (2, 3). Of the 5,026 respondents in his study,
inexpensive shoes «US $40) were worn by 311
(6.2% -not 2.7%), and 260 (5.1%) wore expensive ones
(>US $95). Their respective frequency of injuries per
year corrected for training mileage was 14.3 and 31.9-
123% greater injury frequency with expensive shoes.
Differences of this magnitude (P < 0.05) are not easily
dismissed as the work of statistical gremlins as they
unconvincingly attempt. Obviously, features present in
expensive footwear are injurious. If this relation between
footwear cost and injury frequency were extrapolated
upward linearly, athletic footwear costing more
than US ' $300 could come with a guarantee of at least
one injury per year. This adds novel meaning to the
notion of getting value for one' s money. We believe
Marti's work, as well as the 14 other references, support
our conclusion: "Based on the above data, notwithstanding
unsupported claims by footwear manufacturers
of improved protection with their products, it seems
appropriate to consider expensive running shoes from
major manufacturers (and perhaps less expensive shoes)
In re lation to footwear safety standards, Frederick
and Cavunagh contend, "everyone who works in this
area knows there are, as yet, no such standards." This
is nonsense. Our group has voting representation on
the following committees: American Society for Testing
of Materials (ASTM) Group F08 Sport Equipment and
Facilities; ASTM F08. 54 Athletic Footwear; ASTM F-
13 Safety and Traction in Footwear; American National
MEDICINE AND SCIENCE IN SPORTS AND EXERCISE
Standards Institute (ANSI) Z41 Performance Requirements
for Protective Occupational Footwear. Each has
advanced footwear safety standards, but, as stated in
our text, "they are inadequate because they do not
account for the discomfort-impact illusion." We have
always thought that injury incidence can be reduced
through footwear modifications (6-11) and have
worked toward this goal through the above organizations.
Their lack of knowle dge of these groups, let alone
not being participating members, indicates that their
passion concerning athletic footwear user safety is less
They argue that plantar perception during locomotion
bears no relation to plantar sensibility at rest, yet
provide no data or reference to support this. It is
preferable in physiological research to simulate as
c1os~ ly as possible natural conditions, which we mentioned
in our text as a limitation. However, this shortcomi ng
is shared by all experimental research, which is
usu~l ly performed in a laboratory, and controls variables
not controlled in nature. Surely they do not
belier ve all experimental research-including their
own-is invalid. Furthermore, we do not believe it is
feasible to accurately measure perception of plantar
load during actual running. Without providing a
method, their comment is no more than empty rhetoric.
They misquote us. It should read (6): "Barefoot activity
when practical (no need for thermal insulation;
no risk of crush injuries; social acceptability) deserves
consideration since plantar sensory mediated protective
adaptations seem optimized for this condition. Although
this may run counter to notions prevalent in
economically advanced countries recounting the dangers
of barefoot activity and necess ity of footwear even
when barefoot activity is feasible, supporting data are
lacking, and many have c oncluded that footwear design
is guided by fashion rather than health considerations."
We stand behind the above statement, and Frederick
and Cavanagh provide no reference or data suggesting
that barefoot activity, when practical, is either "outrageous
We will now briefly deal with conflict of interest
guidelines. MSSE is among a small group of medical
journals that have no published guide lines that require
authors (also editorial staff and rev iewers) to disclose
potential conflicts of interest related to work s ubmitted
for publication. Here are extracts of published guidelines
from two medical journals:
"The Journal asks authors of research articles to
disclose at the time of submission any financial arrange,
mem they may have with a c ompany w hose products
figure nrnmlncntlv in the submitted manuscript, or
with a comnanv making competing products ...
Because the es senc e of revie ws and editorials is selection
and interpretation of the literature, the Journal
expects that authors of such articles will not have any
financial interest in a company (or its competitor) that
LETTERS TO THE EDITOR-IN-CHIEF J
makes a product discussed in the ar ticle." (Information
For Authors, The New England Journal fMedicine).
" ... The letter should inc lude any inti, rmation not
given on the title page that might be elevant to a
possible conflict of interest, e.g., consultancies, stock
ownership or patent licensing arranzements." (Infor mation
For Authors, The Journal of the American
Individuals with financial interest in products that
they investigate have difficulty in maintaining the detachment
required to objectively interpret data. This
conscious or unconscious bias results in commercial
advertising-not science. Scientific publications have
these guidelines to help assure reliable reporting and
interpretation of research.
Being late at developing conflict of interest guidelines
has advantages. MSSE can introduce up-to-date guidelines-
we believe that these standards in many journals
could use modernizing. For example, despite disclosed
financial interest in a product, authors are occasionally
allowed to publish, with perhaps certain revisions to
distance themselves from these products. We see nothing
improper with this, other than that the reader is not
informed about it. We believe that disclosed potential
conflicts of interest should be included in published
reports. The Editor-in-Chief is then freed to publish
work without later risking charges regarding authors
w ith conflicts of interest. Potential conflicts of interest
could be reported following acknowledgments.
In the case in point, Dr. Frederick was the previous
director of research for a footwear manufacturer using
the brand Nike. He has commercial relations (consulting
and research contracts) with a number of athl etic
footwear manufacturers through Exeter Research Inc.,
a personal corporation. Dr. Cavanagh has similar commercial
relations with athletic footwear manufacturers
using brands: Nike, Puma, Etonic, and Avia. It should
1. CASPERSON, C. J., K. E. POWELL, J. P. KOPLAN, R. W. SHIRLEY,
C. C. CAMPBEll, and R. K. SIKES. The incidence of injuries and
hazards in recreational and fitness runners (Abstract). Med. Sci.
Soares Exerc. 16: 113, 1984.
2. MARTI, B. Relationships between running injuries and running
shoes: results or a study or 5.000 participants in a Iti-km runthe
May 1994 Berne Grand-Prix etudy. In; The Shoe in Spurt.
W. rtornnzer and B. Se[!esser (Eds.). Chicago: Yearbook Medical
Publiehere, 1989, pp. 256-265.
3. MARTI, B.• J. P. VADER, C. E. MINDER, and T. MELIN. On the
epidemiology of running injurie~; the Bern Grand-Prix study.
Am. 1. Sports Med. 16:285-294,1988.
4. McNITT-GMY, J. L. and T. YOKOI. The influence of surface
characteristics on the impulse characteristics of drop landinos.
rroceennas ortne mn Annual Meeting or tne America n S ociety
on Iliomeehanics, Auo' 23-25, 1999, BuriinEton, VT, 1989, pp.
). NIGG, B. M., A. H. BHILSON, J. DIiNOTll, S. M. LUlITm, and A.
STACOFF. Factors influencing kinetic and kinematic variables in
Official Journal of the American College of Sports Medicine 147
be noted that these footw ear manufacture rs produce
products, presumably in part based on these consultants'
recommendations, that our res earch has shown
c re ates t he discomfort-impact illusion, thus promoting
chronic overloading. These conflicts of interest, which
were not disclos ed in their letter, change its meaning.
For example, the ir defe nse of footw ear manufactur ers
can be v iewed as a pretext to protect their financial
position as consultants to them. What makes this example
of conflict of int erest sad and cogent, is that
these authors are senior researchers w ho act as role
models for y ounger inv es tigators.
Clearly, Frederick and Cav anagh are not happy that
science is progressing along a path that they did not
anticipate. Their letter is little more t han an advertisement
for the current running shoe, in which they have
financial interest, and for w hich they are in part answerable.
As science, it is without substance, because
after removing the camouflage of feig ned self- righteous
indignation, hyperbole, platitudes, and jargon from
their letter, there is not one refer ence (or ev en unpublished
data or personal communication) suggesting that
these footwear are safe, nor do they counter data we
present indicating they ar e dangerous.
Frederick and Cavanagh blame their displeasure on
devious misleading authors, incompetent reviewers,
blind editors, and decepti ve statistics. A more plausible
and ethical explanation of their discontent is their lack
of objectivity, and their self-interest.
Steven E. Robbins, M.D.
Gerard J. Gouw, Ph.D. (eng)
Human Performance Group
Department of Mechanic al Engineering
running. In: Biomechanics of Running Shoes. B. M. Nigg (Ed.).
Champaign, IL: Human Kinetics Publishers, 1986, pp, 139-1 S9.
6. ROBBINS, S. E. and G. J. Gouw. Athletic footwear and chronic
overloading: a brief review. Sports Med. 9:76-115, 1990.
7. ROHIHN~S, .E. and G. J. Gouw. Modern athletic footwear: unsafe
due to perceptual illusions. Mad. Sci. Sports Exo»: 2J:217-224,
SI. Ronnnrs, S. h. and A. M. YAI'JI'JA. Running related injury prevention
through barefoot auaptauons, Mea. sa. soons sxeic.
9. RQHIHN~,S. E., A. M. HANNA, and O. J. oovw. ov enoac
protection: avoidance of'hoavy plantar surface loading. Mad. Sci.
Stxms tixerc. 20:1\)-92. 19M.
10. ROBBINS,S. E., A. M. HANNA,and G. J. Gouw. Running related
injury prevention through innate moderating hshnvior. M uti Sci.
Spons Exctc, 21;130-139. 1959.
II. ROIlIlII'JS, S. b., A. M. lIANNA, and L. JONES. Sensory attcnuauon
induced by modem athletic footwear. J. Test Eval; 16;412-416.
The Publications >