Content uploaded by Stephen Morley
Author content
All content in this area was uploaded by Stephen Morley on Jan 15, 2014
Content may be subject to copyright.
British Journal of Psychiatry (1995), 167, 99—103
The role of hedonic tone, that is, the ability to
experiencepleasure,hasattracted attention in the
study of psychopathology (Loas & Piersen, 1989).
In the presentcentury,thesignificanceof hedonic
tone had beeneclipsed(Snaith, 1993),but attention
to it was revived by a study (Klein, 1974)which
proposedthat low hedonic tone isa central feature
of a type of mood disorder which is likely to respond
to antidepressantmedication.An alternativeview
had been proposed(Meehl, 1962), which considered
that low hedonic tone is a personality trait pre
disposing to the development of schizophrenia and
depressive disorder. Interest in the significance of
thestatewasincreasedbyitsinclusionasoneof the
central features of major depressivedisorder in
DSM—III—R(American Psychiatric Association,
1987).
The significance of hedonic tone requires further
clarification. For this to occur theremust beprecise
definition and the provision of accurate methods for
screeningand assessment.Several‘¿pleasure'scales
exist, the bestknown of which arethoseof Fawcett
et a! (1983)and Chapmanet a! (1976).There is also
a pleasure scale for children (Kazdin, 1989).
Problemswith thesescalesarise from their length
and probable cultural bias: for instance,items
include such statements as “¿Poetsalways exaggerate
thebeautiesof nature―(Chapman),“¿Whilefishing
you feel a tug on your line andwatcha six-pound
fish jump out of the water with your bait in its
mouth―,“¿Yourneighbours rave about the way you
keep up your house and yard―(Fawcett), “¿Your
teacher makes you the King/Queen for the day
(Kazdin).
Thereisa needfora simplerscale,unlikelytobe
affectedby socialclass,sex,age,dietaryhabitsand
nationality. It should be capable of ready translation
intootherlanguages.Sinceit willbea self-assessment
scale the statements must be simple and easy to
understand. The scale should cover a wide range of
domains of pleasure.
The construction of such a pleasurescalewasthe
purposeof this study.
Selection of items
Method
In order to obtain a representativesampleof items,
100 members of the general public were asked to
submit a list of five situations which causedthem
pleasure.Theywereinformed of the purposeof the
study and asked to avoid items which were unlikely to
beapplicabletomostpeople,forexamplea particular
sporting activity, alcoholic drinks, sexualactivity,
andparticulararticlesof diet.The respondentswere
of both sexes, of age range 15—80years, and from
a widerangeof socialclass.Fifty-five replieswere
returned, and from thesea provisional list of 20 items
wasdrawnup. The itemscoveredthedomainsof
socialinteraction, food and drink, sensoryexperiences,
achievement, and pastimes. Subjects were instructed
to indicate the degree to which each item caused them
pleasureon a four-pointscale;inorderto avoid
response set, some items were phrased in negative
terms.
This questionnaire was given to a number of
people from the general population and some
99
A Scale for the Assessment of Hedonic Tone
The Snaith—Hamilton Pleasure Scale
R. P. SNAITH, M.HAMILTON, S. MORLEY, A. HUMAYAN, 0. HARGREAVES and P. TRIGWELL
Background.Hedonictone andits absence,anhedonia,areimportantin psychopathological
research, but instruments for their assessment are lengthy and probably culturally biased.
Method. A new scalewas constructedfrom the responsesof a large sampleof the general
population to a request to list six situations which afforded pleasure.The mostfrequent items
were reviewed and those likely to be affected by cultural setting, age, or sex were removed.
A pilot study led to an abbreviated scale of 14 items, covering four domains of pleasure
response. This questionnaire was subjected to psychometric evaluation in new samples from
the general population and psychiatric patients.
Results. The scale was found to have a score range that would distinguish a ‘¿normal'from
an ‘¿abnormal'response.Validityand reliabilitywere foundto be satisfactory.
Conclusions. The new scale, the Snaith—HamiltonPleasure Scale (SHAPS),is an instrument
whichmay be recommended forpsychopathologicalresearch.
Generalpopulationagedunder60:men25131agedunder6O:women27411--1agedover6O:men8-11agedover60:women8Total68553--1General
populationscoring>4 onGHQ127-111Psychiatric
patientsfirstrating2-333263621363—repeatrating1923-22-1-—-1———
100 SNAITH ET AL
psychiatricpatients.It becameclearthat the negative
wording caused confusion, so these items were
deleted;theproblemof responsesetwasovercome
by varying the order of responses. The resulting
14-item scale appears in the Appendix. Instead of
using a Likert-style scoring device, it was decided to
adopt the simpler method, used in the General
HealthQuestionnaire(GHQ), in whicheitherof the
‘¿Disagree'responsesscoresI point andeitherof the
‘¿Agree'responses scores 0 points. Thus, the score
range is 0—14.
Establishment of the revised scale
This was undertaken among members of the general
public and selected psychiatric in-patients.
Thesamplefromthegeneralpublicwascomposed
of 102 members of staff, students and hospital
visitors.Itwas importantto establishscoreson
an emotionally healthy sample, so the GHQ-30
(Goldberg, 1972) was also administeredand the
responses of those scoring over 4 on the GHQ were
discarded, reducing the sample to 82. For the
purpose of a retest reliability estimate, 30
respondentsrepeatedthe questionnaireafter a few
days, without sight of the previous record. The
samplewas approximately equally divided between
menand women,coveredtherangeof socialclass,
and had an agerangeof 20—80years.
Since the purpose of the study was to establish
scalescoresfor hedonictone, clinicians were asked
to put forward names of patients who clearly
suffered from a defect of this state and who were
willing and capableof completing a self-assessment
scale. The majority, but not all, of the patients
suffered from a major depressiveillness. Forty-six
participated (18 men: 28 women range 17-81). Those
patients who during the study had undergone
considerable improvement in their clinical state
were asked to repeat the ratings, and 30 patients
did so.
The patientswere assessedby pairs of raters
using the Montgomery—Asberg(1979), not to
provide an overall depressionscore,but because
the set of 10 constructson the MADRS, all
ratedon six-pointscales,includesan ‘¿Inabilityto
feel' item. The researcherswere instructed to use
thisitemtoassesshedonictone,which isprobably
the manner for itsuse in routineapplication
of the MADRS. By summing the scoresof the
two raters,scales(0—12)of relevantpsycho
pathologicalconstructswere produced:depressed
mood, depressedappearance, inner tension, appetite,
sleep, lassitude, ability to concentrate, suicidal
preoccupation,pessimism,and hedonictone.Raters
remainedblind to the patients'completionof the
pleasure scale.
Statistics
Non-parametric statisticswere usedincluding, for
the measuresof correlation,the Spearman rank
method.A levelof 1 in 50 (P<zO.02)was accepted
as indicating statistical significance.
Validity
Results
The face validity of the pleasure scale rests upon the
wordingof its items.The contentvalidityis based
on its coverageof arangeof domains of pleasure.
The distribution of the scale scores is shown in
Table 1.
Fewrespondentsinthegeneralpublicscoredover
2, whereasthe majority of patients with reduced
hedonictonescoredmorethan2. Inspectionof the
itemswhich weresometimesselectedby membersof
Table1
Distributionofscoresinsamples
Score
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
SCALE FOR ASSESSINGHEDONIC TONE 101
thegeneralpopulationasfailingto elicit a pleasure
responsedidnotrevealapreponderanceof anyitem.
No sexdifferencewasapparentin responsesto the
items.
Criterion va!idity(i.e. that the scaleisa preferential
measure of hedonic tone rather than some other
construct) is indicated by the following correlations
with the MADRS item ratings: hedonic tone + 0.36
(P< 0.02), suicidalpreoccupation+ 0.38 (P<0.02),
anxiety + 0.34 (NS), depressedmood —¿0.04 (NS),
lassitude+0.08 (NS), appetite+ 0.10 (NS), sleep
+ 0.01 (NS), and pessimism + 0.27 (NS). The rating
of suicidal preoccupation probably represents a
measure of overall severity of illness and thus the
significant correlation is explained; the borderline
significance of the correlation with anxiety is
interestingand requiresconfirmationina different
sample and setting.The absenceof a significant
associationwith depressedmood was somewhat
surprisingbutsupportstheviewthatdepressedmood
and low hedonictonearedifferentconstructs.The
receiveroperatingcharacteristics(ROC)(Fig. 1)show
the performanceof the scaleat differentcut-off
pointsagainsttwo levelsofdiminishedhedonictone:
(a)‘¿perceptible'(ratingover2 on theMADRS item),
(b) ‘¿clinicallysignificant'(ratingover 4 on the
MADRS item).The ROC aresatisfactoryat both
levels,witha confirmationthata cut-offscoreof
0.4
1 - specificity
2 provides the best discrimination between ‘¿normal'
and ‘¿abnormal'levelof hedonictone.
The utility of a scalerefersto the easeof admii
stration and acceptanceby the population for which
it is designed.No respondentsexpresseddifficulty
in comprehension of the scale or returned incomplete
forms.
The sensitivityto changeof clinicalstatuswas
examined by inspection of the ratings of those
patients who had repeatratings of normal hedonic
tone.At initial rating, themedianscorewas5(range
0—13);themedianscoreon repeattestwas0 (range
0—2).The significanceof the change(McNemar
x2=18.05, P<0.001) indicates that the scale is
sensitive to change in clinical status of hedonic tone.
Scores on a mental state measure should vary
according to clinical status. Therefore the test
retest procedure to establish reliability in the patient
sample is not possible since clinical status may change
over a few days. Therefore this aspect of reliability
wastestedby inspectionof the scoresof members
of the general population who repeated the scale after
severaldays:of the30respondentsonly two recorded
a ‘¿normal'(2 or less)scoreon oneoccasionand an
‘¿abnormal'(> 2) scoreon the otheroccasion.The
split-halfcorrelationin thesampleof patients(first
seven with second seven items) was significant
(r= 0.74, P< 0.01). The estimation of internal
consistency of the scale was estimated by the Kuder
Richardsonformulaapplicableto non-parametricdata
(Guilford, 1954) (comparable to the Cronbach
method): for the sampleof patientsthis figure was
0.857, which shows satisfactory internal consistency.
Relationshipof the scalescoreswith agerevealed
no significance in either sample.
Discussion
The study presents a brief assessment scale, to be
calledtheSnaith-Hamilton PleasureScale(SHAPS),
for estimation of the degree to which a person is able
to experience pleasure or the anticipation of a
pleasurableexperience.The items relate to experiences
likely to beencounteredby mostpeople.It is thought
thattranslationinto otherlanguagesfor use in
differentculturalsettingswillnotaffectthevalidity
oftheSHAPS, althoughthisremainstobeestablished.
The phrasing of the items is expected to overcome
thedifficulty whena subjectisnot ableto experience
thesituationcurrently. Thereisonecircumstancefor
which the SHAPS would not be valid, and this is
its use with blind people, since four of the items
depend upon visual experience. The effect of removal
oftheitemsand useof a partialscalecannotyetbe
recommended but may prove to be valid.
1
0.8
0.6
0.2
0 0.2 0.4 0.6 0.8 1
Fig. I Receiver operating characteristics (ROC) of the SHAPS at
two levelsof diminishedhedonictone(cut-offpointof 2/3, -—
and *; cut-off pointof 4/5, -- -- and **).
102 SNAITH ET AL
The SHAPS covers four domains of hedonic 2. I wouldenjoybeingwith myfamily or closefriends:
experience: interest/pastimes, social interaction, Definitelyagree [
sensory experience, and food/drink. Each of these Agree [ I
is based on only a few items and it is not at present Disagree [
recommendedthat subscoresindicatingthesedomains Stronglydisagree [ I
should beused.Further study is now required,but it
is likely that the SHAPS will prove to be a valid 3. 1wouldfind pleasurein myhobbiesandpastimes:
measure of hedonic tone. Strongly disagree I I
It should be noted that children's scoresremain Disagree [ I
to beestablished,but for adolescentsand adult age A@ee [ I
and sexhave no major effect on scores. Stronglyagree [
Performanceof the SHAPS among physically ill
peoplemust beestablished;theability to experience 4. I wouldbeableto enjoymyfavouritemeal:
pleasure is certainly an important aspect of the con- Definitely agree [ I
cept of ‘¿qualityof life'; the relation of scoresto more A@Jee [ I
generalmeasuresof thisconceptwill beanimportant Disagree [ I
area of study and possible application for the scale. Strongly disagree [ I
Hedonictoneiscertainlyanimportantaspectof many
aspects of psychiatric disorders. It may be that 5@I would enjoy a warm bath or refreshing shower:
hedonictoneprovidesanimportant link constructbe- Definitelyagree [ I
tweendepressiveillnessandsuchstatesasobsessional Agree [ I
disorders; Andreasen (1982) recognised lowered Disagree I I
hedonic tone to be one of the components of the Strongly disagree I I
‘¿negative'phase of schizophrenia. The SHAPS may
therefore be of aid in unravelling interconnections 6. 1would fmdpleasurein thescentof flowersor thesmell
ofa freshseabreezeorfreshlybakedbread:
between different psychiatric disorders. Further
progress in research into psychiatric disorder will Strongly disagree [ I
certainlybe aidedby more accuratedefinitionand Disagree [ I
measurementofpsychopathologicalconceptsaswell Agree [ I
asby theredefinitionof diagnosticcategoriesbased Stronglyagree [ I
upon combinations of symptoms (Birley, 1990; Van 7. I would enjoy seeingother people'ssmiling faces:
Praag, 1992; Costello, 1992). Definitely agree [ I
Agree [ I
Acknowledgements Disagree I I
Strongly disagree I I
We are gratefulto otherswho assistedin assemblageof data:
Dr L. Almond, Dr T. Oakleyand Dr M. Radcliffe. 8. I would enjoy looking smart when I have made an
effort with my appearance:
Strongly disagree I I
Appendix. This scale may be reproduced under its Disagree [ I
proper title for personal use and research. Reproduction Agree
in any book or manual or for commercial purpose must Strongly agree [ I
be negotiatedwiththeBHtlshJournalofPsychiatry. 9. I would enjoy reading a book, magazine or newspaper:
This questionnaire is designed to measure your ability to
experiencepleasurein thelastfew days. Definitelyagree I I
it is important to read each statement very carefully. Agree I I
Tickoneoftheboxes[ Itoindicatehow much youagree Disagree [ I
ordisagreewitheachstatement. Stronglydisagree [ I
I. I would enjoy my favourite television or radio 10. 1would enjoy a cup of tea or coffee or my favourite
programme: drink:
Strongly disagree [ I Strongly disagree I I
Disagree [ I Disagree I I
Agree [ I Agree E I
Strongly agree I I Strongly agree I I
SCALEFORASSESSINGHEDONICTONE 103
I I . I would find pleasurein small things, e.g. bright sunny ANDREASEN,N. C. (1982) Negativesymptomsin schizophrenia:
day, a telephone call from a friend: definition and reliability. Archives of General Psychiatry, 39,
784—788.
Strongly disagree I I BIRLEY,J. L. T. (1990) DSM—III:from left to right or from right
Disagree [ J to left? British Journal of Psychiatry, 157, 116—118.
Agree I I CHAPMAN,C., CHAPMAN,i. P. & RAULIN,M. L. (1976)Scalesfor
Strongly disagree [ I physicalandsocialanhedonia.Journalof Abnormal Psychology,
85, 374—382.
I 2. 1would beable to enjoy a beautiful landscapeor view: COSTELLO,C. G. (1992)Researchonsymptomsversusresearchin
. syndromes. British Journal of Psychiatry, 160, 304—308.
Definitely agree I I FAWCETF,J., CLARK,D. C., SHEFThER,A., el al (1983)Assessing
Agree I I anhedonia in psychiatric practice. Archives of General
Disagree I I Psychiatrv, 40, 79—84.
. GOLDBERG, D. (1972) Detection of Psychiatric Illness by
Strongly disagree I I @uestionnaire.Oxford: Oxford University Press.
13. 1 would get pleasure from helping others: GUILFORD,J. P. (1954) Psychometric Methods (2nd edn). New
York: McGraw Hill.
Strongly disagree I I HARDY,P., JOUVEN,R., LANCRENON,S., et al (1986) L'échellede
Disa ree@@ plaisir-désplaisir.L'EncEphale, 12, 149—154.
g KAZDIN,A. E. (1989) Evaluation of the pleasure scale in
Agree I I the assessment of anhedonia in children. Journal of the
Strongly agree [ I American Academy of Child and Adolescent Psychiatry. 28,
364—372.
14. 1would feel pleasure when I receive praise from other KLEIN,Ii F. (1974) Endogenomorphic depression. Archives of
people: General Psychiatry, 31, 447—454.
LOAS, G. & PIERSEN, A. (1989) L'anhédonie en psychiatric: revue.
Definitely agree I I Annaks Medico-Psychologiques, 147, 705-717.
Agree E I MEEHL,P. E. (1962)Schizotaxia,schizotypy,schizophrenia.
Disaoree I I American Psychologist, 17, 827—838.
. MONTGOMERY, S. A. & ASBERG, M. A. (1979) A new depression
Strongly disagree I I scale designedto be sensitiveto change. British Journal of
Psychiatry, 134, 382—389.
References SNAITH,R. P. (1993)AlihUlOma:the forgottensymptomof
psychopathology. Psychological Medicine, 23, 957—966.
AMERICAN PSYCHIATRICASSOCIATION(1987) Diagnostic and Statistical VAN PRAAG, H. M. (1992) Reconquest of the subjective: against
Manual of Psychiatric Disorders (3rd edn) (DSM-1Il). the waning of psychiatric diagnosis. British Journal of
Washington, DC: APA. Psychiatry, 160, 266—271.
R. P. Snaith, FRCPsych,S. Morley, PhD,Division of Psychiatryand Behavioural Sciences,Clinical Sciences
Building, St James's University Hospital, Leeds; M. Hamilton, MRCPsych,A. Humayan, MBBS, D.
Hargreaves, MBChB,P. Trigwell, MRCPsych,Leeds Regional Psychiatric Rotational Training Scheme
Correspondence: Dr R. P. Snaith, Division of Psychiatry and Behavioural Sciences, Clinical Sciences Building, St James's
University Hospital, Leeds LS9 7TF
(First received 14 February 1994, final revision 20 September 1994, accepted 22 September 1994)