‘S's Aes Se isn Asc s Sep Reh Sty A New Method for Measuring Daytime Sleepiness: The Epworth Sleepiness Scale Murray W. Johns Slop Disorders Uni, Epworh Hospital, Melbourn, Vitoria, Ausra Seeman: Te deropent nef no snl, to Epwort spins sl (ESS ec. Th ¢ Spl nlfatminitrel quand wc ot pee smart fh aes sal eo Ss cp Gi eda py fle mane ES, ag men ev syn tra in om cnr uy Tot rst ‘ota angie hapermn FSC sens we sian cat wt ep nee esse ong ‘he mall dp ny Hota rng vege prmnerneyI aen win obec sp tnet ‘talane 8 sures sigan creed hh ent dirt inde nde einer Se, ‘oral oveigh BS wore of pcb nh aed i ot ir fom cots Key Words Spe (GettoazeSlerp props“ tmamain”-Oteracive sepa yrs A lass proportion of adult patints who present 9 sleep disorder centers have disorders associated with excessive daytime sleepiness. These include obsruc- tive slesp apnea syndrome (OSAS), periodic limb movement disorder (PLMD), narcolepsy, idiopathic ypersomnia and other miscallaneous disorders (I). ‘The severity of ther chronic daytime sleepiness is an Jmportantsspect ofeach patients assessment. Thus, there isa great ned for a simple standardized test for ‘measuring a patent's general level of scpines, which is independent of short-term variations in lespines, ‘with the time of day and from day to day ‘The muliple sleep latency test (MSL) is widely ‘sed and is generally believed 10 provide a valid mez ‘surement of sleepiness on the particular day ofthe test. {@,3). ts hased on the premise thatthe slepice the subject, the quicker he wl fill asleep when encouraged to do so while ying down in @ nonstimulating eavi- ronment. The MSLT basa reasonably high test-retest reliability over periods of months in normal subjects (@). Assuming the same reliability holds true for pa tients, the MSLT must be considered the standard method for measuring their chronic daytime sleep ness. However, the MSLT is very eumbersome,time- ‘pil polation Jy 191, Jes Sep Dade Unis Ewart Hana Meter ita consuming and expensive to perform. I takes all da, both forthe subject and the polysomnographer and is not easy to justly asa routine test forall patients ‘Other measures of sleepiness have been devised (5,6, In the maintenance of wakefulness test (MWT) the latency to sleep onset is measured with the subject siting ina dimly lit, warm, quiet room, tying to stay ‘awake rather than to fall aseep (5). However, ll uch tests share the disadvantage of the MSLT in being cumbersome and expensive Similar criticisms can be levelled attests of sleepiness based on pupillometry (©, of exrebral evoked potentials (8), Other asses- mens of sleepiness have involved prolonged peycho- motor performance tet, the results of which are not related in any simple or consistent way to sleepiness in diferent subject). By contrat, the Stanford slepiness sale (888) is a quick and simple west (10) Tt involves the subjects (ova reports of symptoms and felings ata particular time, Visual analogue scales (VAS) of sleepines/l- trtness have also been used inthis conten (11) low fever, these tests do no attempt wo measuee the general level of daytime sleepiness, a stint from fecings of sleepiness ata particular time. Nor, it appears, isthe Subjective sleepiness that they measure the same asthe abjective sleepiness measured by the MSLT (7. Scores on the SSS or on 2 VAS of sleepiness are not sienficantly correlated with step latency in the MSLT, soA SCALE MEASURING DAYTIME SLEEPINESS S41 TABLE 1 The Epuorth sleepiness scale Yours, Your ae ele Meena FP ow at ae you dae of i ey ne loin ‘Sopot ie ie eo ti Even yu have oat dave soma ese ‘te men uy oc ot how thy rl ne fet You. ‘ise Arig hw appro ae ode es tegen Sing, cel apt ace og tent oa aig ng don oe nthe tesonon when emma Store tone nn, Int ar we Stopped be few ius inte tic “Tank you fr er npn IT UI [lt even when measured at virtually the same time (12. “These subjective reports may be related more to tired. ess and fatigue thant sleep propensity, as manifested ‘by the tendency to fll alep. "The present report describes the development and vse of a new questionnaire, the Epworth sleepiness sale (ESS), designed to measure sleep propensity in « simple, standardized way. The sale covers the whole range of sleep propensities, ftom the highest to the Towest. The concept ofthe ESS was derived from observa tions about the nature and occurrence of daytime sleep and sleepiness, Some people whe sur from excesive daytime sepiness kep themselves busy and choose rat tole down nor to sit and relax during the day, thereby purposely avoiding daytime sleep. Others who may be bored, with spare ime ar who are socially withdrawn but who may not be very sleepy, choose to Tie down and sleep during the day. About 80% of os- tensibly healthy medical students usually slep during the day a least once in an average week (13). Among 17-22-year-od recruits entering he French army, 19% reported sleeping during the day, regularly or ooca- ‘Honaly. But only 5% complained of daytime sleep- nest (14), Thus, knowing how frequently or for how Tong subjects usualy sleep during the day wll probably ‘ot provide a useful measurement of their sleepiness. 'BY contrast, sleepy people often describe how they doze off inadvertently while engaged in activites that Involve low level of stimulation, relative immobility andrelaxation, such a sitting and watching TV. Earlier ‘uestionnzresurveyshave indicated which situations, ‘commonly encountered in daily if, ae the most sopo- ‘fe (13) A lage survey among adults in New Mexico asked about their frequency of falling asleep in five ‘Stations (16). The authors derived a score from the thre “most sleepy” questions, which referred to fling asleep while “inactive in 2 public place”, “at work”, fand "in a moving vehicle as paseager or driver" [MSLTS on 116 ofthese subjects showed a statistically significant correlation between their hep latency (SL) And their answers to those three questions (r= ~0:32, P= 0.0). ‘TheESS is based on questions refering to eght such situations, some known to be very soporic; others less ‘0, The questionnaire, which is seltadministered, is reproduced in Table 1. Subject are suked to rate on @ ‘ele of 0-3 how likely they would be to doze off oF {al asleep in the eight stations, based on thir usual ‘way of ie in recent times. A distinction is made be- ‘meen dozing off and simply fing tired. I subject Imag not been in some of the stutions recent, he Is ‘asked, noncthles, 0 estimate how each might affect hin, ‘The ESS tres to overcome the ict that people have ferent daily routines, some facilitating and others Snhibitng daytime seep. For example, the FSS does not ask how frequently the subject fills asleep while ‘watching TY. That Would depend on how frequently he watched TV as much 3s on his slepines. Instead, the subject rates the chances that he Would doze off ‘whenever he watches TV. ‘One question asks how Likely the subject would be to doze off while ying down 10 retin the ternoon ‘when citeumstancss permit. Tt was felt that normal people probably would, and sleepy people certainly ‘would tend to dove off in that situation. Never to do 0 would indicate an unusually low level of sleepiness, fa: described by some insomniacs. Some other situa. ‘ions were included nthe questionnaire because it was ‘elieved that oaly the most sleepy people would daze in them—while sting and talking to someone, and in ‘car while stopped for afew minutes in wai. These ‘suppositions proved correct “The numbers selected for te eight situations in the [ESS wero added togedher to give a score foreach sub |ec, between 0 nd 24, These ESS scores proved ca: able of distinguishing individuals and diagnostic ‘sroups over the whole range of daytime sleepiness, Slog, Yo 86199sa 1M, W. JOHNS ‘TABLE 2. The grup of experimental nue, ther ages and ESS sores “Toa number of ae ‘aun Subjetlignons ‘Kemal contl Dawa Pama sree Boo cans Sean Rares ies [eta bypemoeie em pn tater» Ba ie METHODS Sabjects A total of 180 adult subjects completed the ques- tionnaire, There were 30 comiols who were mainly hospital employees, working during the day, who gave history of normal slep baits without nosing. There ‘were 150 patents with varios sleep disorders, whose ages, sex and diagnostic categories ae show in Table 2. Every new patient who presented to the Epworth Sleep Disorders Unit answered the BSS at ther fist consultation. After investigation, all patients with the iagnoses listed in Table 2 were included in the study unt there were 150. The age of patients ranged from 18 0 78 yeas, The mean age within diagnostic groups ‘atid ftom 36 to 52 years. Men greatly outnumbered ‘Women inthe snoring, OSAS and PLMD groups. The ‘sexes were about equal in the other groups, apart fom the insomniaes where women outaumbered men. ‘Atoll of 138 patients had overnight polysomnog- raphy, bu another 12 who were clearly sfering from cither chronic psychophysioloical or idiopathic in- Somnia did not ‘The later diagnoses were made onthe basis ofeach patients history, using the criteria et out inthe Intemational Classification of Sleep Disorders (), Other insomniaes, wath mood disorders or drug offects, were excluded ‘Twenty-seven paticts had MSLTs afer overnight polysomnography. They had four naps, at 1000, 1200, 1400 and 1600 hours Slep latency was measured fom the time lights were switched off until the onset of stage 1 sleep of atleast 1 minute duration, or the onset of ther stage 2 oF rapid eye movement (REM) sleep. ‘The carly onset of REM slep was indicated by the occurrence of REM sleep within 20 minutes of sleep ‘onset. Of the 27 patents, 11 had narealepsy diagnosed. from the patent's history, particularly of cauapexy, associated with an SL of fess than 10 minutes and the carly onset of REM sleep in two or more naps (10 tints orn one nap patient withcataplery) Four {cen of the 27 patients hed idiopathic bypersomaia, ‘agnosed from their excessive daytime sleepiness in the absence of cither eatalexy or the cary onset of [REM sleepin the MSLT. The remaining two patiens Te ra Bios (en 55) (oas'= 50) arse pia aii en B33has 332% rie had excessive daytime sleepiness due to OSAS. The ESS scores forthe 27 patients who had MSLTs ranged fom 11 to 24, ‘All patients wit primary snoring had presented ini- tilly Because ofthe intensity and persistence of their snoring, on most nights atleast, Many had been ob- Served at home to pause in their breathing at night, sugzesting that they may have had sleep apnea, but this was found not 10 be of clinical significance by polysomnography. The respiratory disturbance index (RD) was caleuisted ae the number of spaeas and Ihypopneas causing a drop of >39 in the arterial ox- yen saturation per hour of sleep. The RDI fer primary Snorers was 5. The 55 patients with OSAS were di- ided into three subeategories according to their RDI, regardless oftheir complaints about daytime sleepiness ‘or insomala (Table 3), The RDI for mild OSAS was ‘Within the range >$-15; for moderate OSAS the range vas > 15-30, and for severe OSAS it was >30 "A diagnosis of PLMD mas made only if there were ‘at leas 90 separate movements in one of bot legs per right. The mean periodie movement index for these subjects, calculated asthe numberof movement events ‘per hour of sleep, was 43.6 + 30.4 (SD), Patients who had both PLMD and OSAS were excluded from this study. However, 9 ofthe 18 subjects with PLMD snored ‘ducing polysomnography without having OSAS. Statistical methods ‘The ESS scores of male and female contol subjects ‘were compared bya Student's test Difleences in ESS Scores between the diagnostic groups were teed by foneeway ANOVA and then by posthoe Scheff tests A separate ANOVA and posthoc Schell tests were TTABLE 3. ES9 sores in ml, moderate and sere OSAS vo, SS Sms ito RH SETS nee mae een gan ee oe Be. G58 Ga EE Bees TG BBR EE ERA SCALE MEASURING DAYTIME SLEEPINESS ‘sed to test the diflrences in ESS scores between pri- ‘mary snorers and the three categories of OSAS. The Scheff testis conservative and is suitable for groups with unequal numbers of subjects (17). The dstribu- tion ofslep latencies, measured in minutes, was highly skewed positively and was normalized by log. tans- Tormation. The relationships becween pairs of contin- ‘uous varibles, such as RDI and slep latency during ‘overnight polysomnography, were tested by Pearson correlation coeficientsand near repression, Statistical significance was accepted at p < 0.05 in twortiled RESULTS ‘The mean ESS score for control subjects was 5.9 = 2.2 (SD) and their modal score was 6. There was 10 Signiicantdiference in the scores between male and Female controls (males ~ 5.64 + 256; females ~ 6.05, 1.84, ¢= 0520, p= 0,607). Consequentiy, a0 dit ‘inciion was made between the sexes in other groups Patients suring from disorders known to be a5- sociated with excessive daytime sleepiness reported the [ikelinood of doing under circumstances that were not conducive to sleepin normal subjects. For example, ‘toof the patents with ether narcolepsy or idiopathic lhypersomnia reported some chance, and often a high chance, of doing while sting and talking to someone, frinacar while stopped fora ew minutesin theta. ‘Only 6% of contol reported a sight chance of doing Patients with persistent psychophysiologial or iio- pathic insomnia reported ither « complete inability for ony slight chance of dozing while lying down to rst in the aflemoon when cicumstances permited By contrast, 94% of controls reported some likelihood cof dozing then. One-way ANOVA demonstrated significant difer- noes in ESS scores between the seven diagnostic groups In Table 2(F= 50.00, d= 6,173; p <0.0001) Posthoc tests between paired groups showed tht the ESS sores for primary snorers didnot der fom coats (p= 0.998), Scores for OSAS, narcolepsy and idiopathic ‘nypersomnia were significantly higher than forcontrls (= 0.001) or primary snorers (> < 0.001). The is omriscshad significantly lower seores (9 <0.01) than all groups other than contol, for which the diffrence id not quite reach statistical signlicace (p~ 0.063). ‘The ESS scores of patients with PLMD did not die significantly from controls (p= 0.149). 'A separate one-way ANOVA flr the ESS scores of| ‘primary snorers and the three subeategoris of OSAS. ‘Showed significant dierenoes berween these groups (F = 25.11; af = 3,82; p < 0.001). Posthoc tests then owed that ESS scores foreach level of OSAS were 343 significantly higher than for primary snorers (p= 0.035 Tot mild OSAS, p < 0.001 for moderate and severe (OSAS). Scores for severe OSAS were higher than for moderate OSAS (p 0.001), but the diference be ftween mild and moderate OSAS did not reach stats tial significance (p ~ 0.083). Considering 3S patents with OSAS together, there was sgnifcant correlation, onthe one and, eiwean ESS scores and RDI (r= 0.550, p< 0,001) and on the other hand, between ESS scores and the minimum ‘a0, recorded during apneas overnight (r = ~0.457, ‘p< (001), The RDIand the minimum overnight a0, “luring apncas were also signifcanly correlated (F = “0.687, p = 0.001). The linear regression equations for these thre relationships, in the form Y= a+ bx, ‘were a fllows (RDN = -0.674 + 2.006(88S stom) (¢ninimum $80,3) ~ 86.47 ~ LOSSIESS sore) {minimum $40,%) = 84.15 ~ 0.440(RD), ‘Among the 138 patients who had overnight poly- somnography there was a significant correlation be- ‘eon ESS store and (a) sleep latency at night (7 = 0.379, a= 138, p< 0.001. In the smaller group of Datenis who had MSLT, the coreaton between In (GL) during the day and ESS score was aso statistically Senieant = ~0.814, 2 ~ 27, p < O01. The linear regression equation fortis relatonship was In (SL) = 3.353 = 0.091€88 scor). Individual ESS scores of 16 or more, indicating & high level of daytime sleepiness, were found only in patients with narcolepsy, idiopathic hypersomnia or OSAS of atleast moderate severity (.e. RDI > 13). All patients with ether narcolepsy or idiopathic hy- persomnia had higher ESS scores than the controls (Le. [ESS > 10) as did 12 of 13 pationts with severe OSAS. The remaining patient in the later category had an. SS score of 8 and was clinically not much affected by his sleep apnea, ‘Within the sroup of pations with PLMD, the pe- Fiodic movement index, which ranged from 16 0 122 ‘oveneats per hours of sleep, was not significantly correlated with BSS scores (r= 0.089, a = 18, p > on. DISCUSSION ‘These results provide evidence that a qvestionnare- ‘based scale as bref and as simple asthe ESS can give valid measurements of sleep propensity in adults. ESS score significantly distinguished groups of patients who ‘are known from other investigations to have dife- ences in ther levels of sleepiness, as measured by the MSLT (218) ESS scores were significantly correlated ‘Sel Ne 19815aa ‘with sleep latoncy measured during the day with MSLTs and at night with polysomnography. This is despite any effect ofthe fst night inthe laboratory. Others have found « signicant postive correlation between ‘the SL at night and during the day in the same subject (9). BSS scores greater than 16, indicative ofa high evel ‘of daytime sleepiness, were encountered only in pa- tients with moderate or severe OSAS (RDI > 15), narcolepsy oridiopthic hypersomnia. These disorders fare known to be associated with excessive daytime sleepiness a8 measured by the MSLT (2,18). Never- ‘theles, high ESS scores, by themselves, are not diag- nostic ofa particular sleep disorder, any more than is pL of 5 minutes in an MSLT, TESS scores were correlated with both the RDI and ‘the minimum Sa, recorded during polysomnography {in patents with OSAS of dilfering severity. In the past, ‘these measures ofthe severity OfOSAS have been found to be elated tothe SL in MSLTs in some, but notin all investigations (18,20). The finding that ESS scores ‘an distinguish patients who simply snore from those ‘with even mild OSAS is evidence forthe sensitivity of the ESS. The questionnaire shoud be useful in eluci- dating the epidemiology of snoring and OSAS, and any associated cardiovascular of cerebrovascular risk. reviousinvestigations of thiskind have tended o blur the distinction betwoea primary snoring and OSAS ep. mn the patients with PLMD, the nding ofan almost zero correlation between ther periodic movement in- dex and ESS scores sugests that whatever level of Saytime sleepiness associated with PLMD, it snot related simply tothe fequency oflimb movements. It ‘ay be more closely related t the frequency of those movements producing arousal rather than those that dono, Ths distinotion was ot made here snd farther {investigation is required to clarify this relationship. ‘The low ESS scores of patients with idiopathic or psychopiysiological insomnia are consistent with ev- ‘ence that suc patients have low sleep propensity, toven when they are able to relax (22). Te must not be assumed, however, that this is necessarily so for other Kinds o insomnia, such as with mood disorders. The relatively wie range of ESS scores inthe control subjects [2-10] i consistent with evidence that some hnaltny adults, without recognizable sleep disorders, ‘emainslepier than others during the dry (23). Such diferences persist in MSL, even after extending the hours of noctumal sleep to overcome possible sleep Aepeivation (24), The sleep propensity ofa subject on a particulae day would be influenoed by the quality and ‘duration of prior sleep or ofslep deprivation, the time fof day, the presence of various sleep disorders, drug fects, the level of interest and motivation induced by (M. W. JOHNS the situation at hand, as well as longer-teren paysio- logical diferences. The ESS does not distinguish the nature of long term physiological or pathological pro- ezses that produce a particular level of ep propen- sity. Other investigations, including overnight poly- somnography, are required for that. TheESS assumes that subjectscan remember wheth- er or not and under what circumstances they have doze off during the day s pat oftheir “usual way of. life in recent times”. The present results suggest that ‘most patients can give meaningful self reports about this aspect of their bebavior and that their ESS scores provide a measurement of their general level of day- time slepines, ftom low o very high levels. This has rot been echieved previously by any other published ‘questionnaire Acknowledgement Irene shel assed with th amin- stration of querionnares to he conrl subj 1, Ager Seep Dorr Aston, he raion ch 2. Rtas Galo Me Fa WV en Ht J.D ‘net W, Mie Macs aie Sept ma {paeep mtn eararmein mela coe a es Barercyloy Ce NewophlsToASe2t Geen A, Dona WC. Thema ep ey et hat oe mane! Sy 198553673 +4 TugpersDooris A Roc eae Roh Tes iat alay of ee MLS onsen s MOG A Gajrany Ke trowmes Chainer ‘bes eit aemnegapi ean a euntng et ‘vin pit wihoneute romans Hacronspapy Gin naopiat bans Seba «6. Emin Mi Bee B Gare DA, Roar HP. The no Esa faites UNS Say Ra IE 1, Heston MR, Pry JM, Reson of wsconienerausns. Seep cai spoend ron Sy OE ronpionR, Aire, Dona W-Asopaonf male ic St necy ad el wk pes 0) rok, Se (1 Ia 9. Iatean Spc Gomes Shy Mata. Dayine hpi, stormed nocturnal se. he et of ‘eedaovaneaoren tr inh Say 9t 10. Hod F, Zaone ¥, Smythe H, Pips, Deneat WC. ‘Geen dep sew pac. Potosi 1. Hae Senn, DF am or Nel Petal vast 1? Seotaice of notin dey an Seton ete td Pave movements fs on he cere apd seine ‘Seo of ane somata ath apo Sley erivan 798 ‘es BE. Gay THA, Goyer MDE. Masten IP. Sep {ets fhe sun fe fey acl H.R, Apa A, Pet Aen i ow, Rath 7, Caradon Dement W. Dati xp tha eco fr Ree MAL ou, Banot WoeA SCALE MEASURING DAYTIME SLEEPINESS sas and mete Seu ent W Nine Mui G, Deen pe ad rte of le ns Papi WG; ef caine seins 16, SaldeNowar WW, Wigs CL, Wal IK Baus Per. Sirona op Sp Rs nn ‘Saal fay’ asp Steph 99 20, Gaenina Pare MGuew'e MA, Hayes De hp Yo Ha 1981