Sei sulla pagina 1di 10

Published Ahead of Print on March 13, 2019 as 10.1212/WNL.0000000000007264

ARTICLE
ARTICLE

Measurement of symptoms in idiopathic hypersomnia

The Idiopathic Hypersomnia Severity Scale

Yves Dauvilliers, MD, PhD, Elisa Evangelista, MD, Lucie Barateau, MD, Regis Lopez, MD, PhD, Sofi ` ene Chenini, MD, Caroline Delbos, S ´ everine Beziat, and Isabelle Jaussent, PhD

Neurology ® 2019;92:e1-e9. doi:10.1212/WNL.0000000000007264

Correspondence

Dr. Dauvilliers

Abstract

Objective To validate the Idiopathic Hypersomnia Severity Scale (IIHSS), a self-report measure of hypersomnolence symptoms, consequences, and responsiveness to treatment.

Methods The 14-item IHSS (developed and validated by sleep experts with patientsfeedback) was lled in by 218 participants (2.3% missing data). Among the 210 participants who fully completed the IHSS, there were 57 untreated and 43 treated patients with idiopathic hypersomnia (IH) aged 16 years or older, 37 untreated patients with narcolepsy type 1 (NT1), and 73 controls without sleepiness. IHSS psychometric properties, discriminant diagnostic validity, and score changes with treatment were assessed.

Results The IHSS showed good internal consistency and content validity. Factor analysis indicated a 2-component solution with good reliability expressed by satisfactory Cronbach α values. IHSS scores were reproducible without changes in the test retest evaluation (13 treated and 14 untreated patients). Convergent validity analysis showed that IHSS score was correlated with daytime sleepiness, depressive symptoms, and quality of life in patients with IH. The IHSS score was lower in treated than untreated patients (5 8 unit dierence, without ceiling e ect). The cutovalue for discriminating between untreated and treated patients was 26/50 (sen- sitivity 55.8%, specicity 78.9%). IHSS scores were higher in drug-free IH patients than NT1 and controls. The best cuto value to dierentiate between untreated IH patients and controls was 22 (sensitivity 91.1%, specicity 94.5%), and 29 with NT1.

Conclusions The IHSS is a reliable and valid clinical tool for the quantication of IH symptoms and consequences that might be useful for patient identi cation, follow-up, and management.

From D ´ epartement de neurologie (Y.D., E.E., L.B., R.L., S.C., C.D., S.B.), Unit ´ e du Sommeil, Centre National de R ´ ef ´ erence pour la Narcolepsie et Hypersomnie rares H opital Gui-de-

Chauliac, CHU Montpellier; and INSERM 1061 (Y.D., E.E., L.B., R.L., S.B., I.J.), University Montpellier, France.

Go to Neurology.org/N for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the article.

ˆ

Copyright © 2019 American Academy of Neurology

e1

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Glossary

AUC = area under the curve; BDI-II = Beck Depression Inventory II; BMI = body mass index; CI = con dence interval; EDS = excessive daytime sleepiness; EQ-5D = European Quality of Life 5 Dimensions; EQ-VAS = European Quality of Life visual analog scale; ESS = Epworth Sleepiness Scale; ICC = intraclass correlation coe cient; ICSD = International Classi cation of Sleep Disorders; IH = idiopathic hypersomnia; IHSS = Idiopathic Hypersomnia Severity Scale; KMO = Kaiser-Meyer-Olkin; MSL = mean sleep latency; MSLT = multiple sleep latency test; NT1 = narcolepsy type 1; NT2 = narcolepsy type 2; PSG = polysomnography; ROC = receiver operating characteristic; SOREMP = sleep-onset REM period.

Idiopathic hypersomnia (IH) is an orphan central hyper- somnolence disorder, clinically characterized by severe ex- cessive daytime sleepiness (EDS), prolonged nighttime sleep, and sometimes sleep inertia. 1,2 EDS is required for IH di- agnosis. It is the most common symptom and often the most debilitating; however, EDS often occurs also in the general population, and EDS etiology is multifactorial, with numerous phenotypes, and is far more common than IH. 3,4 Besides EDS, many patients with IH complain of prolonged nighttime sleep, long unrefreshing naps, sleep inertia characterized by great diculty in awakening from sleep, both in the morning and after naps, impaired daytime alertness, cognitive di - culties, and automatic behaviors. 5 The diagnosis of IH is primarily based on a detailed medical history with the pres- ence of EDS for more than 3 months, absence of cataplexy (which would lead to a suspicion of narcolepsy), less than 2 sleep-onset REM periods (SOREMPs) on the multiple sleep latency test (MSLT) or the preceding polysomnography (PSG), a mean sleep latency (MSL) 8 minutes on the MSLT or a total 24-hour sleep time 660 minutes on 24-hour PSG. 6 Other causes of hypersomnolence (i.e., other sleep disorder, narcolepsy, insucient sleep syndrome, medical or psychiat- ric disorder, medication or substance use/abuse) need to be carefully ruled out. In contrast to its previous edition, the current International Classi cation of Sleep Disorders (ICSD-3) 6 merged the 2 forms of IH (with and without long sleep time) into one heterogeneous condition, because of the absence of clinical and PSG characteristics speci c to each subgroup. 7,8 Symptoms in IH may vary between patients and both their severity and consequences remain mostly unknown. 9 11

Several psychometric tools hav e been developed to evaluate EDS severity in the general population, including the widely used Epworth Sleepiness Scale (ESS). 12 Conversely, only the Sleep Inertia Questionnaire has been developed to evaluate sleep inertia, but it ha s been validated exclusively in patients with mood disorders. 13 A speci c instrument to assess the spectrum of IH symptoms, their functional con- sequences, and changes in response to treatment is not available.

Therefore, we developed the Idiopathic Hypersomnia Se- verity Scale (IHSS), a brief self-report instrument designed to measure the severity of the key symptoms of hypersomno- lence and their consequences. The aims of this study were as

e2

Neurology | Volume 92, Number 15 | April 9, 2019

follows: to assess IHSS psychometric properties, validity, and reliability; to determine the ability of IHSS to quantify the severity and consequences of hypersomnolence symptoms; to evaluate IHSS responsiveness to treatment in a well-dened population of patients with IH aged 16 years or older; and to compare the IHSS scores in patients with IH and in 2 other groups: participants without EDS complaints (controls) and patients with the other, most serious cause of EDS, namely narcolepsy type 1 (NT1).

Methods

Idiopathic Hypersomnia Severity Scale The rst step in the development of the IHSS was to de ne the domains to be evaluated by the scale based on the key symptoms of IH and their consequences. Several pre- liminary versions were developed and pilot-tested by sleep medicine experts in patients from the France-Narcolepsy- Hypersomnia Patients Association. Readability and com- prehensiveness were con rmed by the target population. A nal set of 14 items was retained for inclusion in the IHSS to assess the severity, frequency, and consequences of the key symptoms of IH (appendix e-1; links.lww.com/WNL/ A854). Five items evaluate nighttime sleep symptoms and the related sleep inertia, 4 items evaluate daytime sleep symptoms and the related sleep inertia, and 5 items evaluate impaired daytime functioning due to hypersomnolence. Symptom frequency, intensity, and consequences are rated using a 3- or 4-point Likert scale, providing a total score that ranges from 0 to 50. Higher scores indicate more severe and frequent symptoms. Most patients could complete the IHSS in approximately 5 minutes. The original IHSS was de- veloped in French and validated in a French-speaking population; forward and back translations were performed to develop a certi ed English translation (appendix e-1;

Population validation For this study, 218 consecutive participants aged 16 years or older completed the scale at least once. Patients were recruited from the Reference National Center for Narcolepsy of Montpellier-France. Overall, 102 patients with IH com- pleted the scale: 59 untreated (13 men, mean age 29.8 ± 11.0 years) and 43 treated patients (9 men; mean age 37.93 ± 13.83 years). Treated and untreated patients constitute in- dependent samples. Among the 59 patients in the untreated

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

condition, 44 completed the scale a second time: 12 (4 men, mean age 26.20 ± 10.08 years) were still drug-free and 32 (10 men, mean age 26.52 ± 7.15 years) were taking stimulants (this corresponds to the dependent sample) ( gure 1). Di- agnosis of IH was based on pathologic MSLT results (MSL 8 minutes and 2 SOREMPs), or total sleep time 11 hours during the 24-hour PSG monitoring following ICSD-3 crite- ria, 6 or total sleep time 19 hours during a controlled 32-hour bed rest protocol, as recently described. 14 Of note, all par-

Finally, 77 controls (29 men, mean age 36.04 ± 13.60 years) who were community-dwelling adults were recruited through advertisements and local association networks during the same period from the general population. They had no complaints of EDS (de ned by an ESS score below 11), were not taking any psychotropic medications, and did not have any substantial medical, neurologic, or psychiatric conditions. All participants could speak and understand French and consented to participate in the study.

ticipants had total sleep time 11 hours during the 24 hours, 70% had total sleep time 19 hours during the 32-hour bed rest protocol, and 71.3% had a MSL 8 minutes on the MSLT. Patients did not have cataplexy, were not obese, did not work night shifts or shift work schedules, were not sleep-deprived

Standard protocol approvals, registrations, and patient consents This study was approved by the institutional review board of the University of Montpellier, France.

(at least 7 hours of sleep per night as assessed by medical interview and sleep diaries), did not have severe sleep-

Other measures

disordered breathing, and did not have any substantial

A

standardized clinical evaluation was performed at the time

of

the study by a sleep expert physician to gather information

comorbid medical, neurologic, or psychiatric conditions. CSF hypocretin-1 levels were available for 42 patients, all of whom had normal levels >200 pg/mL 6 .

This study also included 39 untreated patients with NT1 (22 men, mean age 39.45 ± 18.20 years). NT1 was diagnosed according to the ICSD-3 criteria: MSL 8 minutes on the MSLT with 2 SOREMPs and a history of clear-cut cataplexy or CSF hypocretin-1 level 110 pg/mL 6 (CSF hypocretin-1 results were available for 19 patients, all of whom had low CSF hypocretin-1 levels).

related to (1) demographic and clinical characteristics; (2) body mass index (BMI) (nonobese: <30 kg/m 2 ; obese: 30); (3) diagnosis delay; (4) severity of depressive symptoms assessed with the Beck Depression Inventory II (BDI-II), categorized as none or mild (score 19) or moderate to severe symptoms (score 20 63) 15 ; (5) EDS presence and severity, assessed with the ESS and categorized as no EDS (score 10), mild (score 11 15), or severe EDS (score 16) 12 ; and (6) quality of life assessed with the European Quality of Life 5 Dimensions (EQ-5D) instrument, which includes a health

Figure 1 Study flowchart of patients who completed the Idiopathic Hypersomnia Severity Scale (IHSS)

completed the Idiopathic Hypersomnia Severity Scale (IHSS) IH = idiopathic hypersomnia; NT1 = nar- colepsy type

IH = idiopathic hypersomnia; NT1 = nar- colepsy type 1.

Neurology | Volume 92, Number 15 | April 9, 2019

e3

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

self-classi cation system (EQ-5D utility values) and a visual analog scale (EQ-VAS). 16 All psychostimulant drugs taken by patients with IH at the time of the evaluation were also recorded.

Statistical analysis Demographic characteristics and clinical data were described using means and SD for continuous variables, and percentages and frequencies for categorical variables. Independent Stu- dent t tests and analyses of variance were used to compare continuous variables in 2 or more groups, respectively. The χ 2 or Fisher exact tests were used to compare categorical varia- bles. When a signi cant relationship was found in more than 2 groups, 2 × 2 comparisons were required to determine whether groups were signi cantly di erent. A correction for multiple comparisons with the Bonferroni method was used for the 2 × 2 comparisons. Associations between continuous variables were assessed using the Pearson correlation co- e cient. The dependent t test was used to compare di er- ences between continuous variables at 2 dierent time points or 2 di erent conditions. The intraclass correlation coe cient (ICC) was used to evaluate the test retest reliability.

To analyze IHSS factor structure, a principal components factor analysis was performed using the data of the partic- ipants who completed the IHSS at the rst visit using a Vari- max rotation. The number of factors was determined on the base of the obtained factor loadings and eigenvalues. Sam- pling adequacy was assessed by calculating the Kaiser-Meyer- Olkin (KMO) index. The internal consistency (reliability) of the scores for the di erent items was estimated using the Cronbach coe cient α .

Receiver operating characteristic (ROC) curves were drawn using the IHSS total scores to identify the cuto that best di erentiated between groups. The best cutowas de ned as the point with the highest Youden Index ([speci city + sen- sitivity] 1). Statistical signicance was set at p < 0.05. Sta- tistical analyses were performed using SAS version 9.4 and Stata 11 software (StataCorp 2007; Stata Statistical Software:

Release 11; StataCorp LP, College Station, TX).

Data availability statement Anonymized data not provided in the article because of space limitations will be shared by request from any qualied investigator.

Results

Characteristics of untreated patients and controls Comparison of untreated patients with IH (n = 59), NT1 (n = 39), and controls (n = 77) who completed the IHSS the rst time showed that women were more numerous among patients with IH than with NT1, and that patients with IH were younger than patients with NT1 and controls. BMI was higher in patients with NT1 than in the IH and control groups. As expected, ESS scores were lower in patients with IH than in patients with NT1, but higher than in controls. Sleep inertia was reported more frequently by patients with IH than by patients with NT1 (78.9% vs 23.5%, p < 0.0001). Depressive symptoms (BDI-II score) and health-related quality of life were comparable between patients with IH and NT1 (table 1).

Table 1 Demographic, clinical, and polysomnographic characteristics in drug-free patients with idiopathic hypersomnia (IH) or narcolepsy type I (NT1) and controls

 

Untreated patients with IH, n = 59, n (%) or mean (SD)

Untreated patients with NT1, n = 39, n (%) or mean (SD)

Controls, n = 77, n (%) or mean (SD)

 

Post hoc

Variable

p Value

comparisons

Sex, M

13 (22.03)

22 (56.41)

29 (37.66)

0.002

IH < NT1

Age, y a

29.78 (11.00)

39.45 (18.20)

36.04 (13.60)

0.004

IH < NT1, controls

Body mass index, kg/m 2a

23.64 (5.23)

25.85 (4.13)

22.86 (3.18)

0.001

NT1 > IH, controls

Age at onset of EDS, y a

17.20 (7.13)

20.43 (9.37)

0.08

Disease duration, y a

12.58 (10.86)

17.80 (15.21)

0.07

Family history of EDS (yes)

4 (6.90)

3 (7.69)

0.88

Epworth Sleepiness Scale a

16.18 (3.31)

18.54 (2.59)

4.99 (2.64)

<0.0001

NT1 > IH > controls

Sleep inertia (yes)

41 (78.85)

8 (23.53)

<0.0001

BDI-II Score a

14.55 (9.42)

15.26 (8.74)

0.60

EQ-5D-utility a

0.77 (0.20)

0.76 (0.20)

0.86

EQ-5D-VAS a

63.61 (21.47)

60.58 (16.36)

0.30

Abbreviations: BDI-II = Beck Depression InventoryII; EDS = excessive daytime sleepiness; EQ-5D = European Quality of Life 5 Dimensions questionnaire; VAS = visual analog scale. a Quantitative variables are expressed as mean (SD).

e4

Neurology | Volume 92, Number 15 | April 9, 2019

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Questionnaire feasibility All IHSS items showed less than 2.3% missing data among untreated participants. The correlation coe cients between items (in the whole sample and in each group) were lower than 0.70, suggesting a non-redundancy of the items. The IHSS was fully completed by 210 participants ( gure 1). The validation analysis was performed among the 167 drug-free participants without IHSS missing data (i.e., 57 untreated patients with IH, 37 untreated patients with NT1, and 73 controls). Conversely, all treated patients with IH fully completed the scale (no missing data).

Construct validity

Internal consistency

Reliability tests showed a Cronbach α of 0.92 for the whole IHSS, indicating a good internal consistency. The correlation

of each IHSS item with the total score was satisfactory (range

0.60 0.91).

Factor analysis

Sampling was adequate (KMO = 0.92). The eigenvalue-1 criterion retained 2 components that accounted for 63% of the total variance (table 2). Component I was composed of 5

items focusing on night/inertia (questions 1, 2, 3, 4, and 8), and component II included 9 items focusing on day/per- formances(questions 5, 6, 7, 9, 10, 11, 12, 13, and 14). Reliability, expressed by the Cronbach α value of each factor, was good for both components (0.92 and 0.84, respectively). The item loading values, which represent how strongly each item is associated with the underlying component, ranged from 0.41 to 0.85 (table 2). Communalities, which refer to the percentage of variance for each item, were higher than 0.40 (range 0.46 0.86), except for question 5 on automatic behaviors related to sleep inertia (i.e., In the minutes after waking up, do you ever do irrational things and/or say irrational things, and/or are you very clumsy? ) with

a communality value of 0.22. Concerning question 14 on

driving-related problems (i.e., Do you consider that your

hypersomnolence is a problem in terms of your driving

a car? ), 14.5% of the volunteers did not drive, and were scored as 0 (no problem). However, due to the clinical importance of these 2 items, we chose to keep them in the nal version of the scale.

Convergent validity

The IHSS total, component I, and component II scores did not change in patients with IH according to MSLT latency

Table 2 Factors structure of the Idiopathic Hypersomnia Severity Scale based on the whole sample of drug-free participants (n = 167, 57 patients with idiopathic hypersomnia, 37 with narcolepsy type 1, and 73 controls)

 

First 2 components

Questions

KMO item by item

Communalities

I

I I

1

0.89

0.68

0.79

2

0.90

0.65

0.74

3

0.91

0.64

0.77

4

0.90

0.66

0.70

5

0.85

0.22

0.41

6

0.88

0.66

0.81

7

0.91

0.51

0.58

8

0.87

0.46

0.67

9

0.96

0.67

0.79

10

0.93

0.86

0.85

11

0.93

0.82

0.84

12

0.93

0.75

0.79

13

0.95

0.78

0.81

14

0.94

0.50

0.71

Cronbach α

0.92

0.84

KMO measure of sampling adequacy

0.92

Percent of explained cumulative variance

63.23

Abbreviation: KMO = Kaiser-Meyer-Olkin.

Neurology | Volume 92, Number 15 | April 9, 2019

e5

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

below or above 8 minutes, total sleep time below or above 19 hours during the 32-hour bed rest protocol, or 14.49 hours (median of the IH population) during the rst 24 hours of the recording. In contrast, compared to IH patients without sleep inertia (n = 16), those with sleep inertia (n = 41) had higher IHSS total (33.88 ± 8.48 vs 26.00 ± 6.82, p = 0.003), com- ponent I (11.53 ± 3.43 vs 8.30 ± 3.06, p = 0.004), and com- ponent II scores (22.35 ± 6.37 vs 17.70 ± 5.14, p = 0.03).

The IHSS total score was positively correlated with the ESS score in untreated and treated patients with IH and also in controls ( p < 0.01 for all comparisons) (table 3). The IHSS total score was also positively correlated with the BDI-II score and negatively with the EQ-5D utility score in untreated and treated patients with IH and patients with NT1. The IHSS total score was negatively correlated with the EQ-VAS score only in untreated and treated patients with IH.

Temporal stability

Among the 57 untreated patients with IH, 12 fully completed the IHSS scale a second time in the same untreated condition

after a median interval of 206 days (range 29 700). Similarly, 13 treated patients with IH completed the IHSS scale twice with a median delay of 306 days (range 148 465). The total IHSS score and the score of the 2 distinct components did not di er between evaluations in drug-free (IHSS total score:

31.42 ± 9.53 at the rst visit and 31.50 ± 6.67 at the second visit, p = 0.39) and in treated patients (22.23 ± 10.31 vs 22.31 ± 11.94, p = 0.56). The ICC was 0.60 (95% condence in- terval [CI] 0.090.83) among untreated patients and 0.83 (95% CI 0.56 0.94) among treated patients.

IHSS score responsiveness to medication in patients with IH No ceiling e ect was observed in the 43 treated patients with IH who completed the IHSS the rst time. Indeed, their median IHSS score of 23 (range 439) was near the center of the possible range of scores. The stimulant drugs used to treat patients with IH were moda nil (n = 22), methylphenidate (n = 16), and pitolisant (n = 5). Comparison of the 2 in- dependent IH samples (57 drug-free and 43 treated patients) showed that IHSS total score was higher in drug-free patients

Table 3 Association of the Idiopathic Hypersomnia Severity Scale (IHSS) total score with sleepiness, depressive symptoms, sleep measures, and quality of life in patients with idiopathic hypersomnia (IH), patients with narcolepsy type 1 (NT1), and controls

Untreated patients with IH (n = 57), IHSS: total score

Treated patients with IH (n = 43), IHSS: total score

Untreated patients with NT1 (n = 37), IHSS: total score

Controls (n = 73), IHSS: total score

 

Mean (SD) or correlation coefficient

 

Mean (SD) or

Mean (SD) or

Mean (SD) or

p

correlation coefficient

 

correlation coefficient

p

correlation coefficient

 

Measurements

Value

p Value

Value

p Value

ESS total score

10

33 (1.41)

0.03

20.76 (7.02)

0.004

0.92

1116

29.29 (7.45)

22.00 (12.04)

25.75 (11.95)

16

35.64 (7.48)

34.00 (10.74)

27.81 (7.53)

ESS total

0.38

0.005

0.53

0.0002

0.26

0.12

0.57

<0.0001

score

BDI-II total

score

19

31.03 (8.48)

0.07

22.12 (10.17)

0.02

25.70 (7.21)

0.03

>19

35.63 (7.03)

34.00 (7.35)

32.82 (7.69)

BDI-II total

0.39

0.006

0.52

0.0007

0.61

0.0001

score

IH phenotype

MSLT > 8 min a

34.18 (10.87)

0.07

28.42 (9.79)

0.11

MSLT 8 min

31.32 (7.27)

22.55 (10.85)

EQ-5D-utility

0.47

0.0005

0.58

<0.0001

0.53

0.001

EQ-5D-VAS

0.37

0.01

0.58

0.0001

0.11

0.55

Abbreviations: BDI-II = Beck Depression Inventory II; EQ-5D = European Quality of Life 5-Dimension questionnaire; ESS = Epworth Sleepiness Scale; IHSS = Idiopathic Hypersomnolence Severity Scale; MSLT = multiple sleep latency test; VAS = visual analog scale. a Patients with MSL >8 minutes (MSLT) but with prolonged nighttime sleep (>11 hours; 24-hour polysomnography monitoring).

e6

Neurology | Volume 92, Number 15 | April 9, 2019

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

(32.07 ± 8.42 vs 24.19 ± 10.78, p < 0.0001), with a mean dierence of 7.88 ± 3.93. The scores of the IHSS component I and II were also higher in drug-free patients (10.82 ± 3.5 vs 8.49 ± 4.12, p < 0.006 and 21.25 ± 6.14 vs 15.70 ± 7.60, p < 0.0001, respectively). According to the Youden Index, the IHSS cutovalue for discriminating between untreated and treated patients was 26 (area under the curve [AUC] 71.2%, 95% CI 60.8 81.6; sensitivity 55.8%, 95% CI 41.0 70.7; speci city 79.0%, 95% CI 68.489.5): 64% of all patients with IH had an IHSS score >26 (78.95% in the untreated and 44.19% in the treated group).

In addition, 32 patients with IH were evaluated twice, before and during stimulant drug treatment (moda nil n = 18, methylphenidate n = 10, pitolisant n = 3, and amphetamine n = 1), with a median interval of 187.5 days (range 15 473 days) (i.e., dependent sample). The IHSS total score was higher in the untreated condition (32.13 ± 9.59 vs 27.22 ± 9.61, p = 0.0004) with a mean di erence of 4.90 ± 7.59. Similarly, the scores for component II were higher in un- treated than treated patients (21.19 ± 6.72 vs 17.28 ± 7.03, p = 0.0002) but not for component I. Among these 32 patients, 75% had a IHSS total score >26 in the untreated and 46.9% in the treated condition.

Discriminant diagnostic validity In the drug-free condition, IHSS scores were higher in patients with IH than in the NT1 and control groups (32.07 ± 8.42, 27.76 ± 7.89, and 10.49 ± 6.36, respectively, p < 0.0001) ( gure 2). These results remained unchanged after adjust- ment for age, sex, and BMI. According to the Youden Index, the IHSS score of 22 was the best cutovalue for discrimi- nating between untreated patients with IH and controls (AUC 97.1%, 95% CI 94.0100.00; sensitivity 91.2%, 95% CI 83.9 98.6; speci city 94.5%, 95% CI 89.399.7). The best threshold for discriminating between patients with IH and

NT1 was 29 (AUC 65.3%, 95% CI 54.3 76.4; sensitivity 73.7%, 95% CI 62.285.1; speci city 56.8%, 95% CI

40.8 72.7).

The score of component I was higher in the IH group ( p < 0.001), whereas no dierence was found between NT1 and control groups. The score of component II was higher in the IH and NT1 groups than in controls ( p < 0.001 for both). Similar results for IHSS total, component I, and component II scores were also obtained when comparing a subpopulation of untreated patients with IH and age- and sex-matched controls (n = 47 per group, p < 0.0001 for all comparisons).

The IHSS total, component I, and component II scores were always higher in patients with IH than in those with NT1 after adjustments for age, sex, BMI, and ESS score ( p < 0.0007, p <

0.0001, and p < 0.02, respectively). A sensitivity analysis found that the IHSS total score was higher in patients with IH and MSLT latency >8 minutes than in patients with NT1 (32.80 ±

9.31 vs 27.76 ± 7.89, p = 0.01), while the score for component

I, but not for component II, was higher in all patients with IH

(irrespective of the MSLT results) than in patients with NT1 ( p < 0.0001). Similar results were also found when comparing patients with IH with sleep inertia and patients with NT1, with higher IHSS total, component I, but not component II scores in the IH group ( p < 0.0009 and p < 0.0001, respectively).

Discussion

We provide the rst validation of a brief patient-reported questionnaire, IHSS, for idiopathic hypersomnia. IHSS assesses all IH clinical symptoms (prolonged and non- refreshing daytime and nighttime sleep, impaired daytime alertness, and sleep inertia), with a good balance between

Figure 2 Idiopathic Hypersomnia Severity Scale (IHSS) total score in the different groups

Severity Scale (IHSS) total score in the different groups Controls, untreated patients with narcolepsy type 1

Controls, untreated patients with narcolepsy type 1 (NT1), and untreated and treated patients with idi- opathic hypersomnia (IH) (independent sample, different populations of drug-free and treated patients; dependent sample, patients evaluated before and during therapy).

Neurology | Volume 92, Number 15 | April 9, 2019

e7

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

items that were selected and validated by sleep experts with patients feedback. Its psychometric properties and its re- sponsiveness to treatment indicate that IHSS is a reliable tool for assessing IH symptom severity and their consequences and to detect clinically signi cant changes upon medication. Our results also provided evidence that IHSS can di erentiate patients with IH from controls and from patients with NT1, the most serious cause of EDS in humans.

IH pathophysiology remains unknown. In the absence of

a specic biological marker, 17 23 IH diagnosis is made by

excluding other sleep disorders, such as narcolepsy type 2 (NT2), atypical forms of depression, mild forms of sleep- disordered breathing, behaviorally induced insucient sleep syndrome, and long sleeper phenotypes. 1,2 A careful history and full physical examination, sleep questionnaires, overnight PSG, MSLT, 24 hours or longer continuous EEG monitoring, or actigraphy are essential for demonstrating increased nighttime and daytime sleep and to rule out other EDS cau-

ses. 6 Research groups have used di erent diagnostic criteria, making comparisons across studies inconsistent. It remains

di cult to precisely estimate the percentage of patients with

each symptom of IH (length of nocturnal sleep, severity of EDS, unrefreshing naps, and sleep inertia), and the severity and consequences of such symptoms on daily life. The overall alteration in quality of life and psychosocial burden seem similar between IH and NT1. 9 11 Di erently from narcolepsy, for which a reliable and valid clinical tool was recently de- veloped to evaluate symptoms, severity, consequences, and response to treatment, 24 no previous instrument evaluated the spectrum of IH symptoms and their consequences.

We developed the IHSS to assess IH symptoms, their severity, and functional consequences, in order to characterize IH burden at the time of the initial diagnostic evaluation and to monitor changes in symptom severity and functional im- pairment in response to treatment. The IHSS has good psy- chometric properties with signi cant item-total score correlations and adequate internal consistency. The factor analysis indicated a 2-component solution, with a component

I (night/inertia) and a component II (day/performance),

both with good reliability. The IHSS total, component I, and component II scores were higher in patients with IH with sleep inertia compared to those without. In contrast, no dif-

ferences were found according to MSLT latencies or pro- longed total sleep time on 24- or 32-hour recording. IHSS total score is reproducible in patients with IH, without sig-

ni cant changes in the test retest evaluation in both treated

and untreated conditions. Correlations among the IHSS, ESS, BDI, and EQ-5D scores in untreated and treated patients with IH support the convergent validity of the scale.

Altogether, the internal consistency and the reproducibility analyses gave results for the IHSS total score that meet the quality standards for such an instrument. IHSS can discrimi- nate patients with IH from controls with a cutovalue of 22/ 50 with excellent speci city and sensitivity, and also from

e8

Neurology | Volume 92, Number 15 | April 9, 2019

patients with NT1 with a cutovalue of 29/50 with correct sensitivity but lower speci city. The scores of components I and II were higher in patients with IH than in controls and patients with NT1 after adjustment for age, sex, BMI, and ESS. Moreover, IHSS score dierences were more pro- nounced in patients with IH with long sleep time and non- pathologic MSLT. All these ndings suggest that IHSS may be

a relevant tool to evaluate all the main features of IH and could help during IH diagnosis.

IHSS can also quantify the symptom severity and related impairment after IH diagnosis, and showed good sensitivity for detecting changes in symptoms following treatment. Comparisons of the IHSS scores in 2 independent samples (untreated and treated patients with IH) and in the de- pendent sample (before and during treatment) showed lower scores in treated patients with 58 unit dierences between groups, without ceiling e ect. A cutoof 26 showed a good speci city (79%) in discriminating between treated and un- treated patients with IH. Changes in IHSS scores in untreated and treated groups were signi cant for components I and II in the independent sample and for component II in the de- pendent sample.

Other measures (e.g., ESS, MSLT, and Maintenance of Wakefulness Test) have been used to evaluate various aspects of sleepiness and wakefulness in IH 6,12,25 27 and to assess the treatment e cacy in clinical trials. 28 However, di erently from NT1, few randomized controlled trials have been per- formed in IH. 29 31 For future studies, IHSS could be useful for monitoring symptoms at baseline and at dierent time points during and after treatment in order to assess the treatment

e cacy, and guide decision-making concerning the choice

and modication of treatment regimens. The patient may become an active participant in the assessment/quantication of the main symptomatic complaints and in treatment choices and goals for IH management optimization, such as dose adjustment or drug switching. IHSS could help to phenotype patients with IH by assessing all IH clinical symptoms in

a semi-standardized way. IHSS could also be used to monitor

the natural course of IH. Indeed, although often its symptoms

are stable and long-lasting, spontaneous improvement may occur in up to 25% of patients. 7,32,33

Our study has both strengths and limitations. We evaluated

the IHSS psychometric properties in a large and well-dened population of 102 adults with conrmed diagnosis of IH, in

a single tertiary university hospital (the National Reference

Center for Narcolepsy and Other Central Hypersomnia Disorders of Montpellier, France). Additional research is needed to determine IHSS comprehensiveness and applica- bility to potentially less severe populations or those with comorbid conditions. As no other scale or gold standard exists to quantify IH severity, no comparison could be done. We need to validate the optimal cutos to discriminate between patients with IH and other central hypersomnolence con- ditions, especially NT2, which is a common dierential

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

diagnosis. However, an overlap also exists between the 2 con- ditions, as similar clinical symptoms may be shared by IH and NT2, and changes in the number of SOREMPs over 2 con- secutive MSLTs exist in patients with central hypersomnia that may modify the diagnosis. 1,34 36 Moreover, CSF hypocretin-1 levels were normal in most patients with either narcolepsy without cataplexy (80% of patients, and 100% using ICSD-3 dening NT2 with normal CSF hypocretin-1 levels) or IH. 6,20 We also need to validate the minimal score change associated with successful treatment. However, comparisons of the IHSS score changes following treatment with other tools, such as ESS, actigraphy, sleep diary, Maintenance of Wakefulness Test, or long-term PSG monitoring, could be performed in future studies.

In this study, we developed a brief self-report questionnaire to evaluate symptom frequency, severity, and consequences in IH. The IHSS shows good psychometric properties and is sensitive to detect clinical changes in symptoms following treatment. Additional studies are needed to validate this in- strument in clinical and research settings, to discriminate patients with IH from other conditions with hypersomno- lence, and to conrm the score changes over time following treatment and in clinical trials.

Author contributions

Y. Dauvilliers: drafting/revising the manuscript for content,

including medical writing for content; study concept and design; interpretation of data analysis; study supervision and coordination. E. Evangelista: acquisition of data, in- terpretation of data analysis, revising the manuscript for content. L. Barateau, R. Lopez, S. Chenini: acquisition of data, revising the manuscript for content. S. Beziat: revising the manuscript for content, statistical analysis, interpretation of data analysis. I. Jaussent: drafting/revising the manuscript for content, statistical analysis, interpretation of data analysis.

Study funding No targeted funding reported.

Disclosure

Y. Dauvilliers received funds for seminars, board engagements,

and travel to conferences by UCB Pharma, Jazz, Theranexus, Flamel, and Bioprojet. E. Evangelista and L. Barateau report no disclosures relevant to the manuscript. R. Lopez received funds for speaking by UCB Pharma and Shire. C. Delbos, S. Beziat, and I. Jaussent report no disclosures relevant to the manuscript. Go to Neurology.org/N for full disclosures.

Publication history

Received by Neurology July 20, 2018. Accepted in nal form December 7, 2018.

References

1. Billiard M, Dauvilliers Y. Idiopathic hypersomnia. Sleep Med Rev 2001;5:349 358.

2. Billiard M, Sonka K. Idiopathic hypersomnia. Sleep Med Rev 2015;29:23 33.

3. Bixler EO, Vgontzas AN, Lin HM, Calhoun SL, Vela-Bueno A, Kales A. Excessive daytime sleepiness in a general population sample: the role of sleep apnea, age, obesity, diabetes, and depression. J Clin Endocrinol Metab 2005;90:4510 4515.

4. Ohayon MM. From wakefulness to excessive sleepiness: what we know and still need

to know. Sleep Med Rev 2008;12:129 141.

5. Vernet C, Leu-Semenescu S, Buzare MA, Arnulf I. Subjective symptoms in idiopathic hypersomnia: beyond excessive sleepiness. J Sleep Res 2010;19:525 534.

6. American Academy of Sleep Medicine. International Classi cation of Sleep Disorders-Third Edition (ICSD-3). Darien, IL: American Academy of Sleep Medi- cine; 2014.

7. Anderson KN, Pilsworth S, Sharples LD, Smith IE, Shneerson JM. Idiopathic hypersomnia: a study of 77 cases. Sleep 2007;30:1274 1281.

8. Vernet C, Arnulf I. Idiopathic hypersomnia with and without long sleep time:

a controlled series of 75 patients. Sleep 2009;32:753 759.

9. Ozaki A, Inoue Y, Nakajima T, et al. Health-related quality of life among drug-naive patients with narcolepsy with cataplexy, narcolepsy without cataplexy, and idiopathic hypersomnia without long sleep time. J Clin Sleep Med 2008;4:572 578.

10. Dauvilliers Y, Paquereau J, Bastuji H, Drouot X, Weil J-S, Viot-Blanc V. Psychological health in central hypersomnias: the French Harmony study. J Neurol Neurosurg Psychiatry 2009;80:636 641.

11. Ozaki A, Inoue Y, Hayashida K, et al. Quality of life in patients with narcolepsy with

cataplexy, narcolepsy without cataplexy, and idiopathic hypersomnia without long sleep time: comparison between patients on psychostimulants, drug-na¨ıve patients and the general Japanese population. Sleep Med 2012;13:200 206.

12. Johns MW. A new method for measuring daytime sleepiness: the Epworth Sleepiness Scale. Sleep 1991;14:540 545.

13. Kanady JC, Harvey AG. Development and Validation of the Sleep Inertia Ques- tionnaire (SIQ) and assessment of sleep inertia in analogue and clinical depression. Cogn Ther Res 2015;39:601 612.

14. Evangelista E, Lopez R, Barateau L, et al. Alternative diagnostic criteria for idiopathic hypersomnia: A 32-hour protocol. Ann Neurol 2018;83:235 247.

15. Beck A, Steer R, Brown G. Beck Depression Inventory II. San Antonio: Psychological Corporation; 1996.

16. EuroQol Group. EuroQol: a new facility for the measurement of health-related quality of life. Health Policy Amst Neth 1990;16:199 208.

17. Mignot E, Lammers GJ, Ripley B, et al. The role of cerebrospinal uid hypocretin measurement in the diagnosis of narcolepsy and other hypersomnias. Arch Neurol 2002;59:1553 1562.

18. Kanbayashi T, Inoue Y, Chiba S, et al. CSF hypocretin-1 (orexin-A) concentrations in narcolepsy with and without cataplexy and idiopathic hypersomnia. J Sleep Res 2002; 11:91 93.

19. Kanbayashi T, Kodama T, Kondo H, et al. CSF histamine contents in narcolepsy, idiopathic hypersomnia and obstructive sleep apnea syndrome. Sleep 2009;32:181187.

20. Dauvilliers Y, Baumann CR, Carlander B, et al. CSF hypocretin-1 levels in narcolepsy, Kleine-Levin syndrome, and other hypersomnias and neurological conditions.

J Neurol Neurosurg Psychiatry 2003;74:1667 1673.

21. Rye DB, Bliwise DL, Parker K, et al. Modulation of vigilance in the primary hyper- somnias by endogenous enhancement of GABAA receptors. Sci Transl Med 2012;4:

161 151.

22. Dauvilliers Y, Delall´ee N, Jaussent I, et al. Normal cerebrospinal uid histamine and tele-methylhistamine levels in hypersomnia conditions. Sleep 2012;35:1359 1366.

23. Dau villiers Y, Evangelista E, Lopez R, et al. Absence of γ -aminobutyric acid-a receptor potentiation in central hypersomnolence disorders. Ann Neurol 2016;80:259 268.

24. Dauvilliers Y, Beziat S, Pesenti C, et al. Measurement of narcolepsy symptoms: the Narcolepsy Severity Scale. Neurology 2017;88:1358 1365.

25. Littner MR, Kushida C, Wise M, et al. Practice parameters for clinical use of the

multiple sleep latency test and the maintenance of wakefulness test. Sleep 2005;28:

113 121.

26. Mitler MM, Gujavarty KS, Browman CP. Maintenance of wakefulness test: a poly- somnographic technique for evaluation treatment e cacy in patients with excessive somnolence. Electroencephalogr Clin Neurophysiol 1982;53:658 661.

27. Carskadon MA, Dement WC, Mitler MM, Roth T, Westbrook PR, Keenan S. Guidelines for the multiple sleep latency test (MSLT): a standard measure of sleepiness. Sleep 1986;9:519 524.

28. Evangelista E, Lopez R, Dauvilliers Y. Update on treatment for idiopathic hyper- somnia. Expert Opin Investig Drugs 2018;27:187 192.

29. Mayer G, Benes H, Young P, Bitterlich M, Rodenbeck A. Moda nil in the treatment of idiopathic hypersomnia without long sleep time: a randomized, double-blind, placebo-controlled study. J Sleep Res 2015;24:74 81.

30. Philip P, Chaufton C, Taillard J, et al. Moda nil improves real driving performance in patients with hypersomnia: a randomized double-blind placebo-controlled crossover clinical trial. Sleep 2014;37:483 487.

31. Trotti LM, Saini P, Bliwise DL, Freeman AA, Jenkins A, Rye DB. Clarithromycin in γ -aminobutyric acid-related hypersomnolence: a randomized, crossover trial. Ann Neurol 2015;78:454 465.

32. Bassetti C, Aldrich MS. Idiopathic hypersomnia: a series of 42 patients. Brain 1997; 120:1423 1435.

33. Kim T, Lee JH, Lee CS, Yoon IY. Di erent fates of excessive daytime sleepiness:

survival analysis for remission. Acta Neurol Scand 2016;134:35 41.

34. Lopez R, Doukkali A, Barateau L, et al. Test-retest reliability of the multiple sleep latency test in central disorders of hypersomnolence. Sleep 2017;40.

35. Trotti LM, Staab BA, Rye DB. Test retest reliability of the multiple sleep latency test

in narcolepsy without cataplexy and idiopathic hypersomnia. J Clin Sleep Med 2013;9:

789 795.

36. Baumann CR, Mignot E, Lammers GJ, et al. Challenges in diagnosing narcolepsy

without cataplexy: a consensus statement. Sleep 2014;37:1035 1042.

Neurology | Volume 92, Number 15 | April 9, 2019

e9

Copyright © 2019 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.

Measurement of symptoms in idiopathic hypersomnia: The Idiopathic Hypersomnia Severity Scale Yves Dauvilliers, Elisa

Measurement of symptoms in idiopathic hypersomnia: The Idiopathic Hypersomnia Severity Scale Yves Dauvilliers, Elisa Evangelista, Lucie Barateau, et al. Neurology published online March 13, 2019 DOI 10.1212/WNL.0000000000007264

This information is current as of March 13, 2019

Neurology ® is the official journal of the American Academy of Neurology. Published continuously since 1951, it is now a weekly with 48 issues per year. Copyright © 2019 American Academy of Neurology. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.

year. Copyright © 2019 American Academy of Neurology. All rights reserved. Print ISSN: 0028-3878. Online ISSN: