BrJ7Sports Med 1997;31:240-245

Exercise enhances creativity independently ofmood

Hannah Steinberg, Elizabeth A Sykes, Tim Moss, Susan Lowery, Nick LeBoutillier,Alison Dewey

AbstractObjectives-It has been widely accepted inthe literature that various forms of physi-cal exercise, even in a single session,enhance positive mood. It has also beenshown that physical exercise may some-times enhance creative thinking, but theevidence is inconclusive. Positive moodscan favour creative tinking, but theopposite has also been reported and theserelations are unclear. There is a largeanecdotal literature suggesting that crea-tive people sometimes use bodily move-ment to help overcome "blocks". The aimof this study was to establish whetherpost-exercise creative thinking was attrib-utable to improved mood.Methods-The responses of 63 partici-pants to an exercise (aerobic workout oraerobic dance) and a "neutral" videowatching condition were compared. Moodwas measured using an adjective list, andcreative thinking was tested by threemeasures of the Torrance test.Results-Analysis of variance showed alarge and significant increase in positivemood after exercise (P<0.001) and asignificant decrease in positive mood aftervideo watching (P<0.001). A significantincrease between the creative thinkingscores of the two conditions was found onthe flexibility (variety ofresponses) meas-ure (P<0.05). A multifactorial analysis ofall data failed to show a significant covari-ance of creative tinking with the twomeasures ofmood (P>0.05).Conclusions-These results suggest thatmood and creativity were improved byphysical exercise independently of eachother.(BrJ Sports Med 1997;31:240-245)

Keywords: physical exercise; mood enhancement; crea-tive thinking

The link between physical exercise and favour-able moods has now been reasonably wellestablished. Various kinds of exercise, includ-ing especially aerobic workout, jogging/running, bicycling, and treadmill, have beenshown to enhance mood, both in anxious anddepressed patients and in normal volunteers,and after both a single session and an extendedexercise programme.`-5

In addition, effects on cognitive processeshave increasingly been examined, but withmixed and on balance inconclusive results. Of

14 papers identified between 1981 and 1995from a search of PsychLIT and Medline, ninereported positive findings-that is, improvedcognitive performance after exercise-and fivenegative.'0 The positive findings usually re-sulted from a running or jogging programmelasting weeks or months,"-" although treadmillexercise'4 and dancing' consisted of only onebout lasting 20 minutes. Some of the studiesused school children and students asparticipants,""'6, while others used older peo-ple of 40 and above." 1 18 Improvements inperformance were statistically significant al-though not usually spectacular, and measuresincluded well known standard tasks such asalternate uses, obvious consequences, match-ing problems, successive and simultaneouscognitive processing tests, verbal reasoning,successive numbers tasks, digit span, and trailmaking. Interestingly, Hassmen et all9 reportedimprovements in relatively complex cognitivetasks but only if exercise was combined withthe administration of branched chain aminoacids, which are thought to cause decreased5-hydroxytryptamine concentrations in thebrain; however, the branched chain aminoacids had little effect on mood.

In the nine "positive" investigations, moodchanges did not appear to have been measuredconcurrently. It would have been interesting todetermine how far positive moods resultingfrom exercise might have been related to bettercognitive performance.Mood was assessed concurrently with cogni-

tive performance in two of the five "negative"reports,6 7 usually by means of adjective lists orsimilar self assessment methods (State-TraitAnxiety Inventory, Center for EpidemiologicalStudies Depression Scale, Affect Balance Scaleand Cook-Medley Hostility Scale). Blumen-thal et al,6 in a complex investigation of elderlyparticipants, found that participants experi-enced some improvement in mood and showedbetter performance. However, these improve-ments were not found to be correlated withchanges in aerobic power due to exercise,although greater fitness associated with re-duced depression approached statistical signifi-cance, and after 14 months of participation, nosignificant changes in mood were found. Pierceet a! also found non-significant differencesbetween aerobic exercisers and control groups(strength training and waiting list) in improve-ments in mood or cognitive performance.Despite this equivocal evidence, it is widelybelieved, as Blumenthal et at point out, thatcognitive performance is enhanced by exercise.

School ofPsychology,Middlesex UniversityH SteinbergE A SykesN LeBoutillierA Dewey

University CollegeLondonH Steinberg

Department ofPsychology, PlymouthUniversity, Plymouth,Devon, UnitedKingdomT Moss

Charing Cross andWestminster MedicalSchool, London,United KingdomS Lowery

Correspondence to:Professor Hannah Steinberg,Queensway, Enfield EN34SF, Middlesex, UnitedKingdom.

Accepted for publication23 April 1997

240

on April 27, 2020 by guest. P

rotected by copyright.http://bjsm

.bmj.com

/B

r J Sports M

ed: first published as 10.1136/bjsm.31.3.240 on 1 S

eptember 1997. D

ownloaded from

Exercise, mood, and creativity

It seems clear from published and unpub-lished case reports collected by ourselves thatcreative individuals, especially writers and aca-demics, do use various forms of movement,mostly walking, to help them, especially whenthey are blocked. For example, a recent poem20contains the interesting line "I walk along thebeach to jog my memory". Our attention hasalso been drawn to Harry, the spiv, who in the1950's film The Belles of St Trinians (withAlistair Sim), when told by the headmistress to"stop pacing about like that", replies: "I mustpace-if I don't pace I can't concentrate, can'tthink!".There is also considerable evidence from a

large although mostly anecdotal literature thatthe "high" moods (hypomania as defined inDSM-IV') in manic depressive people can bemost productive of creative work,22.2. althoughsome contrasting reports have appeared-forexample, that depressed moods are apt to leadto creative work and that any symptoms ofmental illness are detrimental to creativity.2629Creativity is one of the most sought after butleast understood mental processes. It is mostusually defined as leading to a new product ofvalue in the broadest sense, 30 and it is difficultto test quantitatively. The best known testmethod is the Torrance test,32 first published in1966 and since much revised; its various com-ponents require "the subjects to think in diver-gent directions in terms of possibilities". Thehigh moods in manic depressive people areusually described as feelings of intense wellbeing, euphoria, confidence, fluency, andvigour," and this is similar to the descriptionsof, for example, the runner's high.'4 Eysenck2'has stressed the connection between psychoti-cism and creativity. Experimentally inducedpositive moods-for instance, showing a com-edy film-have also been shown to facilitatecreative thinking.'5 Oaksford et al,'6 however,have shown that both positive and negativeinduced mood states can suppress some kindsof reasoning performance.Mood has traditionally been assessed by self

reports of positive and negative dimensions;most standard assessment methods have beendeveloped against psychiatric backgrounds andhave been apt to concentrate on negativeaspects of mood (e.g. Profile of Mood States(POMS)3"). However, Hendrick and Lilley"devised a mood adjective list with a highproportion of positive items. The current listhas been adapted from these two, as describedin another paper.'9 This list was modified froma previous investigation' and was subjected toa principal axis factor analysis with an obliquerotation of the mood scores. The results wereconsistent with a primarily two dimensionalstructure of positive and negative effects whichwere relatively independent states. (For adiscussion of the validity and complexity ofmood scales and the status of the positive-negative dimension, see Watson et al.41)Three hypotheses were tested: that aerobic

exercise would lead to improved mood; thataerobic exercise would lead to improvedperformance in tests of creativity; that improve-

ments in creativity were attributable to im-proved mood.

MethodsPARTICIPANTSThe participants were divided into two groups:those taking part in aerobic workouts and thosedoing aerobic dance. For the first group, 31subjects, mainly students and staff at Middle-sex University, were recruited by campusadvertising with help from physical fitnessinstructors, and each was paid £8.00 forattending two sessions. There were 15 men and16 women aged 19-54 (median age range25-29). For the second group, 32 participantswere similarly recruited, and paid £7.00 forattending two sessions. There were four menand 28 women aged 19-59 (median age range20-24).

Participants from both groups were asked tostate how much physical exercise they wereinvolved in weekly. All exercised regularly for atleast two hours a week, although this did notnecessarily involve aerobics. This was an inter-esting but coincidental factor and not a pre-requisite for selection. All participants for bothexperiments declared themselves to be in goodhealth and able to participate in an aerobicexercise class. There were no drop outs.

Positive and negative mood changes wereassessed, using a selfcompleted mood adjectivechecklist containing 24 "positive" adjectivesand 23 "negative" adjectives (see table 1) inalphabetical order and classified according tothe results of a factor analysis with obliquerotation. Mood for each questionnaire wasscored as total positive score and total negativescore. Each individual adjective was scored ona scale of 1-5. The purpose of the investigationto be described was to compare changes inboth mood and performance of two forms ofthe Unusual Uses Test of Creative Thinking'2in participants who had taken part in a brief(25 minute) aerobics session, and to comparethem with mood and performance changes ofparticipants who had watched a documentary

Table 1 Mood adjective checklist containing 24 positiveadjectives and 23 negative adjectives

Positive adjectives Negative adjectives

Active AngryAlert AnnoyedCarefree AnxiousCheerful ApatheticClear headed ApprehensiveConfident BoredContented DetachedElated DownheartedEnergetic DrowsyEnthusiastic ExhaustedEuphoric FearfulFriendly FrustratedCalm IrritableHappy LonelyKindly Mentally slowLively RegretfulPleased SadQuick-witted Self-absorbedRelaxed SorrySatisfied SpitefulSociable SuspiciousStimulated TenseSympathetic WearyTrusting

241

on April 27, 2020 by guest. P

rotected by copyright.http://bjsm

.bmj.com

/B

r J Sports M

ed: first published as 10.1136/bjsm.31.3.240 on 1 S

eptember 1997. D

ownloaded from

Steinberg, Sykes, Moss, Lowery, LeBoutillier, Dewey

style video of the rock formations of the LakeDistrict of similar duration. The video waschosen to be, as far as possible, emotionally"neutral" but just interesting enough to holdattention.There were thus two experiments: the first

used conventional aerobic workout and thesecond aerobic dance. Aerobic workout waschosen because it had in previous experimentsby ourselves"9 and numerous others been foundreliably to improve mood. This is probably bestdescribed to the lay person as high impact, andinvolved fast and energetic movement. It wasdescribed by the instructor as being a warm up,with exercises designed to increase mobilityand coordination and raise the pulse rate (sixminutes), followed by six minutes of variedroutines using arm and leg movements to-gether in order to increase stamina, and finally,a cool down, with slow relaxation stretches (fiveminutes). Four different music tracks accom-panied the session.

Aerobic dance was chosen because at leastone investigator (Gondola6 ) had reported it tobe effective in improving creativity. It is bestdescribed as low impact and involved gentleexercises with a good deal of slow rhythmicstretching. It was described by the instructor asconsisting of a warm up, which was repetitiveexercises designed to loosen muscles in armsand legs (4.27 minutes), followed by rhythmicstretching (gentle exercises stretching differentparts of the body in sequence: waist (4.12 min-utes) shoulders/arms (3.35 minutes), ribs/hips(3.18 minutes) and legs (3.37 minutes)), andfinally a cool down involving winding downand relaxation exercises again with the mainemphasis on arms and legs (3.4 minutes). B"Sky" music tracks accompanied the exercises.The instructor was the same in both experi-ments, and so was the general format. It waspredicted that both kinds of aerobic exercisewould improve mood and, to a lesser extent,that concurrent improvements in creativitywould be accounted for by the mood enhance-ment following exercise. Preliminary summa-ries have been published.42 43

DEPENDENT VARIABLESBecause the psychological effects of a singlebout of exercise can be short lived, all testmethods, including the creativity tests, had tobe brief. Measures of mood were taken beforeand after each experimental session, allowing amood change variable to be calculated-that is,mood post-exercise-mood pre-exercise. In-structions were as follows: "please indicate byplacing a tick in the appropriate place, theextent to which each adjective describes howyou feel at the moment. This should not takemore than a few minutes, so please do not thinkabout your response to each word and pleasebe honest"; two to three minutes were allowed.Effects of exercise are usually shown mostclearly in improvements of positive mood,40rather than in decreases in negative aspects ofmood.

Creativity was only measured post-exercise;although ideally pre-exercise creativity scoreswould have been helpful, their inclusion would

have resulted in the counterbalancing proce-dures becoming too complicated and timeconsuming. Two forms ("Tin Cans" and"Cardboard Boxes") of Torrance's UnusualUses Test were used.'2 The instructions were:"Most people throw their empty cardboardboxes/tin cans away, but they have thousands ofinteresting uses. In the spaces below and on thenext page, list as many of these interesting andunusual uses as you can think of. Do not limityourself to any size of box/can. You can use asmany boxes/cans as you like. Do not limityourself to the uses you have seen or heardabout; think about as many possible new onesas you can". Participants were given fiveminutes to complete each form of the Torrancetest. There were three scores: fluency =

number of responses; flexibility = variety ofresponses-that is, shifts from one approach toanother or using a variety of strategies;originality = unusual uses. The number ofunusual uses was expressed as the sum ofratings on a four point scale of each response.For example, listing 13 uses for a cardboardbox that all involved using the box as some sortof container or shelter would score highly onfluency but have a comparatively low score forflexibility and originality. Diverse responsessuch as shoes, toy car, television, and shield,although less fluent, would yield higher flexibil-ity and originality scores. Each participantcompleted both A and B forms of the test onseparate occasions; a maximum of five minuteswas allowed. Scoring was carried out inde-pendently by two experimenters, whose agree-ment was of the order of r = 0.89.

INDEPENDENT VARIABLEThe independent variable was type of stimula-tion with two levels: exercise and television/video watching. Participants watched a videothat was chosen to be as far as possibleemotionally neutral (the archaeology of theLake District) and took part in an exerciseclass, both of which lasted 20-25 minutes.

COUNTERBALANCINGParticipants were allocated by a randommethod to either a "video then exercise" or an"exercise then video" condition. Within each ofthese groups, half the participants completedthe Tin Cans form of the Unusual Uses Testafter the video, and the Cardboard Boxes testafter the exercise class, and half completed thetests in reverse order.The participants attended the experimental

sessions on two consecutive days, exercising onone day and watching the video on the other. Inthe video condition, participants completed aninitial mood scale, and then watched the videofor 25 minutes. They then completed the moodscale again, and finally the Unusual Uses Test.The 25 minute exercise condition was similar.The standardised written instructions for eachtest were supplemented with a verbal reinforcerof the points made on the printed instructionsheets.

242

on April 27, 2020 by guest. P

rotected by copyright.http://bjsm

.bmj.com

/B

r J Sports M

ed: first published as 10.1136/bjsm.31.3.240 on 1 S

eptember 1997. D

ownloaded from

Exercise, mood, and creativity

ResultsThe original intention of this study had beenonly to use aerobic workout as describedabove. However, because of the relatively smalldifferences on the three measures of creativitybetween the two groups and the fact that aero-bic workout requires a good deal of energy,concentration, and coordination, which mightdetract from any beneficial cognitive effects, itwas decided to use a comparable but less com-plex and gentler form of aerobic exercise(aerobic dance') for the second experiment.Although there may have been physiologicaldifferences in the two types of exercise, moodand creativity scores did not differ significantly;there was little difference from the resultsobtained by the analysis of the combinedscores, and the sample sizes of 31 and 32became too small to be meaningful. It wastherefore decided to combine the two data sets.No significant differences were found be-

tween the two demographic variables (age andsex) on any of the measures used.

Total positive and negative mood scores werecomputed from all initial questionnaires, and,when results were analysed separately for eachof the two experiments, they yielded baselineresults comparable with those obtained overall.

EXERCISE, VIDEO, AND MOODThe aim of including the video task was toassess the effects of exercise on creativityrelative to a television watching activity for asimilar period in the presence of the same par-ticipants. If mood did account for an increasein creativity after exercise, but watching thevideo did not, then a significant interactioneffect would be found on the mood ratingsbetween exercise and video before and after thetask. To test for this effect, a two way repeatedmeasures analysis ofvariance was used for eachof the two measures of mood. A significantmain effect of type of leisure activity was foundon the positive mood variable (F(1,62) =

37.16, P<0.001). This effect can be explainedby the predicted interaction effect (F(1,62) =58.21, P<0.001). In summary, positive moodincreased after the exercise condition anddecreased after the video condition (see fig 1).An approximately converse pattern of results

was found for the negative mood variable witha significant main effect of type of leisure activ-ity (F(1,62) = 19.96, P<0.001) accounted forby the overall interaction effect (F(1,62) =32.62, P<0.001). Thus negative mood de-creased after exercise and increased afterwatching the video (see fig 1).

EXERCISE AND MOODTwo repeated measures t tests were used toassess differences in mood after exercise (fig 1) .As predicted, a significant increase in positivemood (t(63) = 5.04, P<0.001) and decrease innegative mood (t(63) = -4.04, P<0.001) wasobserved for the two experiments combined.These results were mirrored in the analyses ofthe individual experiments as follows: aerobicworkout: positive mood (t(31) = 5.48,P<O.001) and negative mood (t(31) = -3.47,P<0.001); aerobic dance: positive mood (t(32)= 5.28, P<0.001) and negative mood (t(32) =-4.57, P<0.001).

90 F

80 F-

70

60

50

40

30

20 Pre- Post- Pre- Post-positive positive negative negative

Video

Pre- Post- Pre- Post-positive positive negative negative

Exercise

Figure 1 Mean positive and negative mood scores before and after the task. Positive moodincreased after exercise and decreased after video watching (P<O. 001); negative mooddecreased after exercise and increased after video watching (P<O. 001).

EXERCISE, VIDEO, AND CREATIVITYPrevious research described in the introductionhas shown improvements in creativity afterexercise, as measured by divergent thinkingtasks. It was therefore expected that theparticipants would perform better on thedivergent thinking task after exercise. Repeatedmeasures t tests were used to assess the differ-ences on three aspects of divergent thinking:fluency, flexibility, and originality. Althoughthe participants obtained better scores on thefluency measure after exercise (fig 2), thedifference between exercise and video watchingwas not statistically significant (t(63) = 1.46,P>0.05). A significant improvement was, how-ever, found on the flexibility measure (t(63) =

2.25, P<0.05). Finally, the participants scoredworse on the originality measure after exercise.However, this difference was marginal andnon-significant (t(63) = -0.46, P>0.05).

MOOD, EXERCISE, AND CREATIVITYAlthough only one of the three measures ofdivergent thinking produced significant differ-ences between the- exercise and video condi-tions, it was decided that an analysis of covari-ance would be used on all three measures totest the final hypothesis, which stated thatimprovements in divergent thinking after exer-cise could be explained by an increase in posi-tive mood and a decrease in negative mood. Allthree analyses failed to show a significant co-variance with the two measures of mood:fluency (F(2,60) = 0.41, P>0.05); flexibility(F(2,60) = 0.20, P>O.05); and originality(F(2,60) = 2.62, P>0.05), and, if anything,creativity increased when mood was partialledout. Similarly, calculations of correlation coef-ficients showed no relationship between im-provements in mood and creativity. In sum-mary, it can therefore be tentatively concluded

U,

4-Ca)r-co

C

Coc)0)a,

u0

a)0

w

a)~000E

a)

243

on April 27, 2020 by guest. P

rotected by copyright.http://bjsm

.bmj.com

/B

r J Sports M

ed: first published as 10.1136/bjsm.31.3.240 on 1 S

eptember 1997. D

ownloaded from

244 Steinberg, Sykes, Moss, Lowery, LeBoutillier, Dewey

12

11

010 -

C. 9*CD

8

7Flexibility Fluency Originality Flexibility Fluency Originality

Video Exercise

Figure 2 Mean scores on three measures of divergent thinking after exercise and videowatching. The difference in mean flexibility scores is statistically significant (t(63) = 2.25,P<O. 05).

that mood did not account for increases increativity after exercise.

DiscussionThe results of this investigation show thatmood was significantly improved (by around25%, P<O.OO1) by two kinds of aerobicexercise (workout and dance) lasting only 25minutes, as compared with a neutral videocontrol group, and that there was a consistentalthough modest increase in at least one meas-ure of creativity assessed by two parallel formsof the Unusual Uses Test, which is part of theTorrance test of creative thinking. Even wheredifferences in Torrance test performance werenot statistically significant, two thirds were inthe predicted direction of improvement,namely flexibility and fluency, while originalityworsened. In particular, one might haveexpected fluency to have shown a biggerpositive effect, since this is especially character-istic of hypomania.

Physical exercise can therefore be said tohave slightly enhanced creative thinking, butthe effect, surprisingly, appears to have beenindependent of mood changes as well as of thekind of aerobic exercise used, although dancewas perhaps marginally more effective thanaerobic workout. In other words, moodchanges did not predict creative performance;when the effects on mood were partialled out,the differences between exercise and videoeffects on the creativity measures if anythingincreased.Although therefore aerobic exercise can be

used to enhance creative performance, theconsequences as measured here are modestand not related to the degree of moodimprovement.The dissociation between creativity and

mood after exercise is puzzling. As wasindicated in the introduction, the existingliterature is of little help in explaining this. Itmay be that the improvements in mood in this

investigation were not great enough to mark-edly facilitate creative thinking, as happens inhypomania.

It is also possible that running, which hasmost strikingly led to enhanced moods(e.g.39 44), would have been a more suitableform of exercise for the present purpose,although again, previous literature throws littlelight on this. It may be that "free" rather thanprescribed exercise is more likely to release the"stream of consciousness", as we ourselveshave found in pilot experiments; but we foundit difficult to persuade subjects to carry out freeexercises. The slightly greater improvements increativity after aerobic dance could possibly bedue to the nature of the exercise, which allowsmore freedom of movement than the regi-mented aerobic workout. Or else it may be thatlonger exercise bouts are required for bettereffects on creativity. Although previousresearch'2 15 has looked at both short and longterm effects of aerobic exercise, the lengths ofthe sessions have been similar. Possibly, moresensitive tests of creativity are needed; but asfar as we can tell no other suitable short test isavailable. The exercise history of the partici-pants is also likely to be relevant. Tomporowskiand Ellis45 suggest that regular exercisersexpect exercise to have beneficial effects onmood and cognitive abilities; consequently anyresults may be biased by motivational factors.Perhaps future research should, in addition tothe mood factor, examine other robust exerciseeffects such as self esteem and "arousal"changes as predictors of increased cognitiveactivity after exercise, or perhaps habitual andinexperienced exercisers should be compared.It is also possible that exercise increasescreativity directly, without the intervention ofmood changes, although again, the publishedliterature is of little help. Once the various con-tributory factors have been identified and"titrated" it should be possible to develop exer-cise programmes that have greater benefits forcognitive performance.

We thank the Wolfson Foundation for support, Janie Turnbull,the aerobics instructor, for her cooperation, Malcolm Hart forstatistical advice and computing, and Professor Craig Sharp forlively suggestions.

1 Steinberg H, Sykes EA. Introduction to Symposium onEndorphins and Behavioural Processes: review of literatureon endorphins and exercise. Pharmacol Biochem Behav1 985;23:857-62.

2 Plante TG, Rodin J. Physical fitness and enhanced psycho-logical health. Current Psychology: Research and Reviews1 990;9:3-24.

3 Seraganian P (ed). Exercise psychology: the influence ofphysicalexercise on psychological processes. New York: Wiley, 1990.

4 Biddle S, Mutrie N. Psychology of physical activity andexercise. London: Springer, 1991.

5 Cockerill IM, Nevill AM, Byrne NC. Mood, mileage andthe menstrual cycle. Br7Sports Med 1992;26:145-50.

6 Blumenthal JA, Emery CF, Madden DJ, Schniebolk S,Walsh-Riddle M, George LK, et al. Long-term effects ofexercise on psychological functioning in older men andwomen. J Gerontol 1991;46:352-62.

7 Pierce TW, Madden DJ, Siegel WC, Blumenthal JA. Effectsof aerobic exercise on cognitive and psychosocial function-ing in patients with mild hypertension. Health Psychol1993l2:286-91.

8 Madden DJ, Blumenthal JA, Allen PA, Emery CF. Improv-ing aerobic capacity in older adults does not necessarilylead to improved cognitive performance. Psychol Aging1989;4:307-20.

9 Lawless WF. Effect of arousal on mathematics scores.Percept Mot Skills 1 988;67:3 18.

10 Blumenthal JA, Maddem DJ. Effects of aerobic exercisetraining, age and physical fitness on memory searchperformance. Psychol Aging 1988;3:230-85.

on April 27, 2020 by guest. P

rotected by copyright.http://bjsm

.bmj.com

/B

r J Sports M

ed: first published as 10.1136/bjsm.31.3.240 on 1 S

eptember 1997. D

ownloaded from

Exercise, mood, and creativity 245

11 Ismail AH, El-Naggar AM. Effect of exercise on cognitiveprocessing in adult men. JHum Ergol (Tokyo) 1981;10:83-91.

12 Gondola JC, Tuckman BW Effects of a systematic programof exercise on selected measures of creativity. Percept MotSkills 1985;60:53-4.

13 Tuckman BW, Hinkle JS. An experimental study of thephysical and psychological effects of aerobic exercise onschool children. Health Psychol 1986;5:197-207.

14 Zevras Y, Danis A, Klissouras V. Influence of physical exer-tion on mental performance with reference to training. Per-cept Mot Skills 1991;72:1215-21.

15 Gondola JC. The enhancement of creativity through longand short term exercise programs. Journal of Social Behav-iour and Personality 1986; 1:77-82.

16 Gondola JC. The effects of a single bout of aerobic dancingon selected tests of creativity. Journal of Social Behaviourand Personality 1987;2:275-8.

17 Clarkson-Smith L, Hartley AA. Relationships betweenphysical exercise and cognitive abilities in older adults. Psy-cholAging 1989;4: 183-9.

18 Clarkson-Smith L, Hartley AA. Structural equation modelsof relationships between exercise and cognitive abilities.PsycholAging 1990;5:437-46.

19 Hassmen P. Blomstrand E, Ekblom B, Newsholme EA.Branched chain amino acid supplementation during 30-kmcompetitive run: mood and cognitive performance Nutri-tion 1994;10:405-10.

20 Adcock F. Labour of love. Sunday Times, Style 1995;December 10:8.

21 Diagnostic and statistical manual of mental disorders (DSM-IV), 4th ed. Washington, DC: American PsychiatricAssociation, 1994.

22 Greene TR, Noice H. Influence ofpositive effect upon crea-tive thinking and problem solving in children. Psychol Rep1988;63:895-8.

23 Richards R, Kinney DK. Mood swings and creativity. Crea-tivity Research Journal 1990;3:202-17.

24 Richards R, Kinney DK, Daniels H, Liskins K. Everydaycreativity and bipolar and unipolar affective disorder:preliminary study of personal and family history. EuropeanPsychiatry 1992;7:49-52.

25 Eysenck HJ. Genius: the natural history of creativity.Cambridge: Cambridge University Press, 1995.

26 Andreasen ND, Glick ID. Bipolar affective disorder andcreativity: implications and clinical management. ComprPsychiarty 1988;29:207-17.

27 Eisenman R. Creativity, preference for complexity, andphysical and mental illness. Creativity Research Journal1990;3:231-6.

28 Slaby AE. Creativity, depression and suicide. Suicide LifeThreat Behav 1992;22:157-66.

29 Ludwig AM. Creative achievement and psychotherapy:comparison among professions. Am J Psychother 1992;42:307-20.

30 Weisberg RW. Creativity: beyond the myth ofgenius. New York:W H Freeman, 1992.

31 Boden M. Dimensions in creativity. London: MIT Press,1994.

32 Torrance PE. Torrance tests of creative thinking. New Jersey:Personnel Press, 1966.

33 Jamison KR. Touched with fire: manic depressive illness and theartistic temperament. New York: Free Press, 1993.

34 Sachs ML, Buffone GW. Running as a therapy: an integratedapproach. London: University of Nebraska Press, 1984.

35 Isen AM, Daubman KA, Nowicki GP. Positive effect facili-tates creative problem solving. Jf Pers Soc Psychol 1987;6:1122-31.

36 Oaksford M, Morris F. Grainger B, Williams JMG. Moodreasoning and central executive processes. Jf Exp PsycholLearn Mem Cogn 1996;22:476-92.

37 McNair DM, Lorr M, Droppleman LF. Psychologicaleffects of acute physical activity. Arch Phys Med Rehabil197 1;52:422-5.

38 Hendrick C, Lilley RS. Self-reported mood and runningunder natural conditions. Work and Stress 1970;7:161-77.

39 Steinberg H, Sykes, EA. Mood enhancement through physi-cal exercise. In: Shroder H, Reschke K, Johnston M, MaesS, eds. Health psychology, potential in diversity. Regensburg:Roderer Verlag, 1993.

40 Steinberg H, Salmon P., Morris M. The "runner's high":dimensions of mood states after running. In: Proceedings ofsport, health, psychology and exercise symposium 88. London:Sports Council and Health Education Authority, 1990;147-52.

41 Watson D, Clark LA, Tegellen A. Development and valida-tion of brief measures of positive and negative effect: thePANAS scales. J Pers Soc Psychol 1988;6: 1063-70.

42 Sykes EA, LeBoutillier N. Moss T, Steinberg H. Dissocia-tion between improvement of mood and creativityfollowing exercise. The British Psychological Society Sportand Exercise Psychology Section Newsletter 1995;4:5-7.

43 Sykes EA, LeBoutillier N, Moss T, Steinberg H. Dissocia-tion between improvement of mood and creativityfollowing exercise [abstract]. British Psychological SocietyAnnual Conference 1995;3:133.

44 Morris M, Steinberg H, Sykes EA, Salmon P. Temporarywithdrawal from regular running. J7 Psychosom Res 1990;34:493-500.

45 Tomporowski PD, Ellis RN. Effects of exercise on cognitiveprocesses: a review. Psychol Bull 1986;99:338-46.

CommentaryIf the sport and exercise sciences were everasked to agree on a corporate slogan then afront runner undoubtedly would be "menssane in corpore sano". Indeed, many seem tosupport this sentiment with a zeal that borderson the fanatical, and that can cloud the execu-tion of a reasoned evaluation of the precisepsychological benefits accruing from exercise.Hannah Steinberg and her coworkers are to

be encouraged for adding to a growingliterature that at long last is prepared toacknowledge the complexity of the relationbetween psychological wellbeing and physicalexercise. To develop this debate the first stepmust be a disaggregation of both the "mens"and the "corpora"-that is, which types ofphysical exercise (for example, aerobic oranaerobic, maximal or submaximal, massed or

distributed, short or long term) are related towhich cognitive and affective functions. As thispaper and previous work`' show there are nosimple answers here, complexity is writ large.Hence further empirical work of this nature isvital to add to our uncertainty before we canstart to piece together the psychological, physio-logical, and contextual components of this fas-cinating jigsaw.

JOHN KREMER

Reader in psychologyThe Queen's University ofBelfast

1 McMurray RG. Exercise, mood states and neuroendocrinol-ogy. In: Diamant L, ed. Mind-body maturity, psychologicalapproaches to sports exercise and fitness. New York: Hemi-sphere, 1991:237-54.

2 North TC, McCullagh P. Tran ZV. Effect of exercise ondepression. Exerc Sport Sci Rev 1990;18:379-415.

3 Willis JD, Campbell LF. Exercise psychology. Champaign, IL:Human Kinetics, 1992.

on April 27, 2020 by guest. P

rotected by copyright.http://bjsm

.bmj.com

/B

r J Sports M

ed: first published as 10.1136/bjsm.31.3.240 on 1 S

eptember 1997. D

ownloaded from