Frequency of narcolepsy symptoms and other sleep disorders in narcoleptic patients and their first-degree relatives

This study aims to compare the frequency of sleep disorders between first-degree relatives of individuals with narcolepsy and the general population.

Narcolepsy, a lifelong neurological disorder, has been known for more than a century (Gelineau, 1880), but its causes are not fully understood yet.The disorder is characterized by daytime sleep attacks and manifestations of various REM sleep abnormalities (cataplexy, sleep paralysis, hypnagogic hallucinations).Its prevalence was set at 0.045% based on representative samples of the general population (Ohayon et al., 2002). Other studies using clinical samples or non-representative community samples have extrapolated the prevalence to be between 0.02% and 0.067% in North America (Dement et al., 1972) and Western Europe (Franceschi et al., 1982; Roth, 1980; Hublin, 1994). In Japan this prevalence was estimated to be between 0.16% and 0.59% (Tashiro et al, 1992; Honda, 1979).

The importance of genetic factors in narcolepsy has been addressed for more than 60 years (Krabbe and Magnussen, 1942). However, the results varied from six to 40 percent of narcoleptic individuals who have a close relative with the disease (Nevsimalova et al., 1997; Billiard et al., 1994; Hayduk et al., 1997; Guilleminault et al., 1989; Baraitser and Parkes, 1978). The risk for narcolepsy was estimated to be between 10 to 40 times higher among families with a narcoleptic member than in the general population (Nevsimalova et al., 1997). However, other factors were also cited as playing a role in the appearance of narcolepsy. This was further illustrated in twin studies (Honda et al., 2001; Partinen et al., 1994; Pollmacher et al., 1990). Among 20 pairs of narcoleptic monozygotic twins, only 25-30% were concordant for narcolepsy-cataplexy (Mignot, 1997): It is likely that narcolepsy results from a multifactorial etiology involving both environmental factors and genetic background.

METHODS

PARTICIPANTS

Five groups of subjects constituted this study:

1. A group of 157 narcoleptic patients (probands) 15 years of age and over. The probands were all diagnosed with narcolepsy and were followed at the Sleep Disorders Center of the San Raffaele Hospital (Milan, Italy) or at the Sleep Disorders Clinic of the Institute of Clinical Neurology at the University of Bologna (Bologna, Italy).
2. A group of 261 first-degree relatives. It included 43 fathers, 53 mothers, 33 sisters and 45 brothers, 39 daughters and 47 sons. All but two first-degree relatives agreed to be interviewed.
3. A group of 68 spouses of narcoleptic probands. This group constituted a comparison group in order to assess the possible role of environmental factors.
4. A group of 3,970 subjects 15 years or older representative of the non-institutionalized general population of Italy (46 millions inhabitants). Methodology and various results involving this sample have been published elsewhere (Ohayon and Smirne, 2002). This group was used to calculate the relative risk of narcolepsy among first-degree relatives.
5. A subgroup of 1,071 subjects aged 15 years or older was selected from the previous sample. These subjects were randomly chosen to match male and female first-degree relatives of probands by age and body mass index. This subgroup was used to determine if the frequencies of sleep disorders among first-degree relatives would differ from what can be expected for individuals of the general population.

A total of 4,456 subjects were interviewed and their data collected in order to realize this study.

PROCEDURES

All the 4,456 subjects involved in this study were interviewed with the help of the Sleep-EVAL System. The general population sample was done using random stratified design. The telephone numbers were randomly selected with each province of Italy. The amount of telephone numbers for each province was based on the population size of the province in relationship to the country. Subjects who initially refused to participate were called again before being classified as a refusal. Overall, 4,442 individuals were solicited and 3,970 of them completed the interview. This sets the participation rate at 89.4%.

The 157 narcoleptic patients were assessed and diagnosed in one of the two sleep disorders clinics. One of the interviewers subsequently contacted them by telephone and asked for the coordinates of first-degree family members and spouses (names and telephone numbers). After explaining the study, verbal consent was obtained before collecting any information and starting the interview. The Sleep-EVAL interviews were done by university students at the San Raffaele Hospital (Milan, Italy). The study was approved by the ethical committee of the San Raffaele Hospital.

Interviews lasted on average 72 (+-45) minutes.

DESCRIPTION AND FUNCTIONING OF THE SLEEP-EVAL SYSTEM

The Sleep-EVAL system (Ohayon, 1995, 1999) is an expert system that includes (a) non-monotonic, causal reasoning mode, (b) fuzzy reasoning driven by two neural networks, (c) a mathematical preprocessor and (d) a knowledge base. It is designed to provide homogeneous and standardized diagnostic evaluations. This simply means that the system is able to direct an interview and to make diagnostic hypotheses that will be confirmed or rejected during the interview. The system is able to build decisional trees for positive and differential diagnoses.

The interview typically begins with a standard questionnaire applicable to all subjects. This questionnaire includes sociodemographic information, medical history (illnesses, medications), sleeping habits and sleep symptoms. Each interview is adapted according to the responses provided by the subject on a series of key questions. The system uses the answers to these questions and elicits a series of diagnostic hypotheses. Once a hypothesis is identified, the system begins to build a decisional tree. During this exploration, the system may ask other questions in order to complete the information. This diagnostic exploration is pursued until a final decision is reached regarding that hypothesis (positive diagnosis process). The system then searches for another hypothesis and repeats the process until all diagnostic possibilities are exhausted (differential diagnosis process). The knowledge base of the system (Ohayon, 1995) contains the questions required for all diagnostic descriptions of the ICSD (AASM, 1997) and DSM-IV (APA, 1994) classifications. The neural networks manage uncertainty in the subjects' responses and accumulate this uncertainty. This information is then used to confirm or reject a diagnosis. Consequently, each explored object (including diagnoses) will have a degree of certainty ranging from 0.4 (completely present) to -0.4 (completely absent).

Questions are read aloud by an interviewer to the subject as they appear on a computer monitor, along with the answer's choices. The interviewer then enters the answer into the computer. The answer's choices can be closed-ended (e.g., yes/no, five-point scale, multiple choice) or open-ended (e.g., duration of symptom, description of illness). All questions include "does not know," "does not understand" and "refuses to answer" in the answer choices in order to identify which questions need to be reworded.

The Sleep-EVAL system was tested within several designs. Validation studies performed in sleep disorders clinics (Stanford University, Regensburg University and Toronto Hospital) testing the diagnoses of the system against those of sleep specialists using polysomnographic data gave excellent results with the diagnosis of Obstructive Sleep Apnea Syndrome (kappas of .93 and .92) and very good results with insomnia (kappas of .78 and .71) (Ohayon et al., 1999; Hosn et al., 2000). A study using the Sleep-EVAL system was done with 96 narcoleptic individuals. They were all diagnosed and blood tested by sleep specialists. The Sleep-EVAL system had a nearly perfect recognition of narcoleptic individuals: kappa of 0.96 with a sensitivity of 94.7% and a specificity of 100% (Black et al., 2001). According to the sleep specialists, there were only 4 cases of narcolepsy without cataplexy. The agreement calculated between Sleep-EVAL and the gold standard gave a kappa coefficient of 0.85. However, the number of subjects was very small and should be interpreted with caution. The sensitivity was 75% and the specificity 100%. Answers on narcolepsy symptoms provided during the Sleep-EVAL interviews were also compared to those provided on the Stanford Sleep Inventory (SSI). Data on both instruments were available for 82 narcoleptics and 202 family members. Sleep-EVAL's cataplexy questions had a sensitivity of 75.5% and a specificity of 95.8% and a correlation of 0.75 with the SSI (Okun et al., 2001).

VARIABLES

Information collected by the system included a description of narcolepsy symptoms: daytime sleepiness, cataplexy, hypnagogic and hypnopompic hallucinations, and sleep paralysis. Frequency, age of onset, time since the last episode were collected for each symptom. Information was also collected on sleep disorders diagnoses according to ICSD-97 classification.

The Sleep-EVAL system was previously validated on its ability to diagnose narcolepsy on its positive and differential diagnosis. The agreement between the Sleep-EVAL system and three sleep specialists was tested on 90 randomly selected participants. The Kappas between Sleep-EVAL and each sleep specialist were .89, .93, and 1.0.

DATA ANALYSES

Odds ratios were calculated with 95% CIs. Because in most cases CIs overlapped between the different groups of relatives, data are presented for all relatives but differences across relatives are noted when appropriate. Most comparisons were done using chi-squares with the Fisher's exact test for small groups.

RESULTS

The characteristics of the study sample are described in Table 1.

Probands, brothers, sisters and control subjects were comparable in age, education and occupation. Male probands had a significantly higher Body Mass Index (BMI) than their brothers and sons (Table 1). Female probands had a significantly higher BMI than their daughters but comparable to that of their mothers and sisters. The female matched general population group was comparable to narcoleptics' female first-degree relatives in terms of age (mean age of 44.02+- 16.2 years; female first-degree relatives mean of 44.2+-16.0 years) and BMI (24.3+-2.7 kg/m²; female first-degree relatives mean of 24.7+-6.3 kg/m²). The selected female matched general population group was significantly younger (p<0.05) and was also significantly heavier (p<.0001) than the non-selected women of the general population sample. The male matched general population group also was comparable to the male first-degree relatives both for age (mean age of 41.4+-16.8 years; male first-degree relatives mean age of 42.1+-17.8 years) and BMI (24.5+-1.2 kg/m²; male first-degree relatives mean of 24.7+-3.1 kg/m²). The selected male matched general population group was significantly younger (p<0.05) and was also significantly slimmer (p<.0001) than the non-selected men of the general population sample.

1) GENERAL SLEEP CHARACTERISTICS

A) PROBANDS

As seen in Table 2, probands were comparable to the spouse group for sleep latency, sleep duration, extra sleep on weekends and days off. Probands reported more frequently dreaming and having nightmares nearly every night than the spouse group.

They also reported more frequently nocturnal awakenings, early morning awakenings and having a non-restorative sleep than the spouse group (Table 3).

B) FIRST-DEGREE RELATIVES

Male first-degree relatives significantly differed from the matched general population in sleep duration and extra sleep time. Female first-degree relatives were dissimilar only on sleep duration (Table 2). First-degree relatives were comparable to matched general population groups on most of sleep characteristic variables. Male first-degree relatives reported more frequently dreaming almost every night than the matched general population (Table 3). Female first-degree relatives more frequently reported difficulty initiating sleep than the matched general population group (Table 3).

2) NARCOLEPSY

A) PROBANDS

Table 4 shows the frequency of narcolepsy symptoms among the probands, their spouse, their first-degree relatives and their matched general population subjects:

• Excessive daytime sleepiness (EDS) was reported as having started during childhood or adolescence for 37.3% of the probands. For the majority of probands (62.7%), EDS first occurred during adulthood.
• Cataplexy was reported by 75% of probands. The first episode of cataplexy occurred during childhood in 6.3% of cases, during adolescence in 21.5% and in adulthood in 72.2% of cases. Episodes occurred on a daily basis in 25.3% of cases; 27.8% reported having cataplexy episodes several times a week. Another 16.5% reported about one episode a week and 30.4% said they had a cataplexy episode once a month or less. The most recent episode occurred during the week before the interview in 63.2% of cases. Another 16.7% reported the last episode occurred within the last month and the remaining (20.0%) said the last episode occurred more than one month ago.
• Episodes of sleep paralysis occurring at least once a week was reported by 9.7% of probands and episodes occurring once a month or less by 17.6% of probands. The first episode of sleep paralysis occurred in childhood in 5.5% of cases, in adolescence for 16.4% of cases and in adulthood in 78.2% of cases.
• Hypnagogic hallucinations occurring at least once a week was reported by 17% of probands and hallucinations occurring 3 times a month or less by 21% of probands. The first occurrence of hallucination was during childhood for 5.4% of cases, in adolescence for 16.2% of cases and in adulthood for 78.4% of cases.
• Automatic behaviors occurring at least once a week was reported by 14.2% of probands. Such behaviors occurred once a month or less in 10.8% of probands.

B) FIRST-DEGREE RELATIVES

- DAYTIME SLEEPINESS

As can be seen in Table 4, the frequency of feeling sleepy during the day was comparable between male and female first-degree relatives and the general population. At the symptom level, the only significant difference was for cataplexy, which was 22 times higher in the male first-degree relatives and 30 times higher in the female first-degree relatives compared to the general population. The frequencies of hypnagogic hallucinations, sleep paralysis and automatic behaviors also were comparable between first-degree relatives and the general population. When male first-degree relatives were compared to females of the same category, daytime sleepiness, as measured by the Epworth scale (Fisher exact test, p<.05), and the frequency of hypnagogic hallucinations (chi square, p<.05) were higher among female first-degree relatives.

- NEW CASES OF NARCOLEPSY

During the course of the study, 19 cases of narcolepsy were identified for the first time. These cases were distributed among the family members of 17 probands (10.8%). One male proband had two of his sisters and one brother also with narcolepsy. Six cases involved a male proband and his mother. In three cases, it was a male proband and a brother. Two cases were a male proband and his daughter; two other cases were a female proband and his brother. The other cases were one male proband and his son, one female proband and her mother and one female proband and her daughter. In 11 cases (64.7%), it involved a vertical mode of transmission, i.e., from one parent to an offspring.

Compared to the entire Italian representative sample (n=3970), the relative risk of narcolepsy among female first-degree relatives was 54.4 and 105.1 among male first-degree relatives. Even when limiting to narcolepsy-cataplexy, the relative risk remained high: 16.9 among female first-degree relatives and 13.5 among male first-degree relatives.

3) OTHER SLEEP DISORDER DIAGNOSES

Tables 5 and 6 shows the most frequent diagnoses observed in first-degree family members compared with the matched samples from the general population. As seen, several diagnoses were significantly more frequent among first-degree family members than in the general population.

1. For idiopatic hypersomnia disorder, obstructive sleep apnea syndrome, adjustment sleep disorder and insufficient sleep syndrome, higher risks were found in first-degree relatives compared to the general population subjects (Table 5).



2. Regardless of gender, circadian rhythm disorders were higher among first-degree relatives (3.7%) than in the matched sample of the general population (1.5%; odds ratio (OR): 2.5 (1.1-5.5)).



3. Among parasomnias, male first-degree relatives were found at higher risk for sleep talking (OR:2.5) and REM behavior disorder (OR:14.0). Female first-degree relatives had higher risk for sleep talking (OR: 1.9) and the presence of at least one parasomnia (OR: 1.7). When compared to each other, male and female first-degree relatives significantly differed only for restless legs syndrome (Fisher exact test, p<.01).

DISCUSSION

Large family studies on inheritance of narcolepsy are still scant and often limited to the narcolepsy tetrad setting aside important symptomatic information that could better help explain the other predisposing factors for this syndrome.

The study of narcoleptic probands and their first-degree relatives in comparison to spouses, matched subjects from the general population and a representative sample of the general population allowed us to investigate the particularities and specifics of the symptoms of narcolepsy. One of the strengths of this study was that all participants were interviewed using the same methodology; i.e., telephone interviews using the Sleep-EVAL System.

1) NARCOLEPSY CHARACTERISTICS OF PROBANDS GROUP

As our results showed, most narcolepsy symptoms are highly prevalent in the general population. Only cataplexy was less common (0.2%). However, compared to narcoleptic individuals, auxiliary symptoms (sleep paralysis, hypnagogic hallucinations and automatic behaviors) occurred less frequently among subjects of the general population who reported them. For example, among individuals who reported sleep paralysis: 35.5% of narcoleptic subjects had such episodes at least once a week while it was the case for about 20% of the individuals in the general population sample.

Another main difference between narcoleptic individuals and the general population sample lies in the number of symptoms experienced. More than half (50.6%) of subjects with narcolepsy reported two or more narcolepsy symptoms (cataplexy, sleep paralysis, hypnagogic hallucinations and automatic behaviors). In the general population sample, even if we limit to those subjects with at least one of these symptoms (27% of the sample), we found only 22.7% of them with two or more symptoms. The most frequent combination was hypnagogic hallucinations with automatic behaviors.

For many years, there was considerable debate about whether cataplexy should be essential for the diagnosis of narcolepsy (Honda, 1988, Moscovitch et al., 1993; Mayer et al., 1998). Several clinical studies reported that up to 30% of patients with narcolepsy were without cataplexy. This was the case with 25% of our narcolepsy group. In the general population, the identification of narcolepsy without cataplexy is much more perilous since the existing criteria are too poorly defined to be applied effectively in a general population sample (Ohayon et al., 2002).

We found that 17 out of 157 probands (10.8%) had at least one first-degree relative with narcolepsy; five of them had narcolepsy-cataplexy (3.2%). The risk of narcolepsy-cataplexy was 13.5 to 16.9 times higher among first-degree relatives than in the general population. Our prevalence of narcolepsy among first-degree relatives (7.3%) was comparable to previous findings in other countries. The mode of transmission was vertical for most of the cases (64.7%); whether from one parent to the probands (7 cases) or from the probands to one of his/her offspring (4 cases). This vertical mode of transmission has also been observed in other studies (Nevsimalova et al., 1997).

2) PARTICULARITIES OF SLEEP PROBLEMS IN FIRST-DEGREE RELATIVES

The sleep characteristics of first-degree relatives (Tables 2 and 3) were, generally speaking, similar to the matched general population sample. The only significant differences were observed for the sleep duration and a greater frequency of dreams in male first-degree relatives. Compared to the matched general population sample, both male and female first-degree relatives were at higher risk for idiopathic hypersomnia disorder, obstructive sleep apnea syndrome, adjustment sleep disorder, insufficient sleep syndrome and circadian rhythm disorders.

Idiopathic hypersomnia disorder is a relatively uncommon sleep disorder affecting less than 1% of the general population.

Complaints of excessive daytime sleepiness are far more common and have multiple causes other than narcolepsy or hypersomnia disorder. Although high, the prevalence of excessive daytime sleepiness among our first-degree relatives of narcoleptic individuals does not significantly exceed what is observed in the Italian general population.

Hypersomnia and other forms of excessive daytime sleepiness have been constantly reported to be high among first-degree relatives of individuals with narcolepsy. Familial narcolepsy studies that also assessed excessive daytime sleepiness often lacked corresponding information from the general population or other control groups. As for our study, the prevalence of excessive daytime sleepiness in first-degree relatives of narcoleptic individuals in those studies often did not exceed the one observed in the general population (Nevsimalova et al., 1997; Billiard et al., 1994; Mayer et al., 1998).

Nearly half (43.9%) of narcoleptic individuals had at least one parasomnia. This was three times higher than what was observed in the general population. Four parasomnias were higher than what can be found in the general population: sleep talking, nightmares, REM behavior disorders and sleep bruxism. High rates of parasomnias in individuals with narcolepsy were also reported in another study (Mayer et al., 1998). The prevalence of at least one parasomnia was also higher in first-degree relatives of narcoleptic individuals than in the matched general population sample. As for the narcoleptic patients, we found higher rates of sleep talking and REM behavior disorder in first-degree relatives than in the matched general population sample.

As already mentioned, we found a higher frequency of several sleep disorders in the family. However, it should be underlined that some sleep disorders were also frequent among spouses of narcoleptics. This might further illustrate the role of environmental factors in the development of narcolepsy but it can also be the result of a greater awareness of sleep disorders in this specific group (as for the family members too). Another possibility could be related to psychological aspects and specific personality traits of narcoleptics' spouses.

3) SLEEP CHARACTERISTICS OF PROBANDS AND THEIR FIRST-DEGREE RELATIVES

The question of sleep specificities across the group of probands and their family members could be raised. It is obvious that narcoleptic patients are sleepier. However, they are not different from the general population subjects in their sleep duration, sleep latency or extra sleep on days off and weekends.

For the first-degree relatives (without narcolepsy diagnosis), males were significantly different from the general population for sleep duration and extra sleep time on weekends. However, it was observed only in working or retired males when examining with the occupation. This was not the case for the females, who were only different on sleep duration. On the other hand, first-degree relatives were more likely than the general population to have various sleep disorders that may have influenced the sleep duration. Consequently, the observed difference could be the result of these disorders.

4) Limitations

The methodology used for this study is not without limitations. The general population sample included participants 15 years of age or older while the first-degree relatives were aged 12 years and older. However, the effect on the comparability of prevalences should be minimal since we had only 6 first-degree relatives younger than 15 years old. Moreover, the Sleep-EVAL system shown very good ability to detect narcolepsy even when used by non-physician interviewers. The use of a standardized tool to assess al the five groups must be underlined: all subjects were submitted to the same questionnaire and all the probands were seen by a sleep specialist, polysomnographically recorded and blood tested in a similar way.

CONCLUSIONS

Because of the low prevalence of narcolepsy in the general population, it is very difficult to verify if a spectrum of narcolepsy exists in the first-degree relatives of probands with narcolepsy in comparison to what could be expected for the general population subjects. Our methodology allowed to conduct a study describing this spectrum in comparison with group of spouses (having the same environment than the probands), with a group of subjects from the general population matched with the first-degree relatives of the probands and finally with general population subjects. Our results showed that first-degree relatives of patients with narcolepsy have several unique features that distinguished them from the general population. First-degree relatives have a risk factor for narcolepsy culminating at 105.1 while for the females, it was 54.4. The vulnerability to hypersomnia and its different forms of expressivity and severity in first-degree relatives can be confirmed based upon an odds ratio of 23.0 (idiopatic hypersomnia) when they are compared to general population individuals.

REFERENCES

1. American Academy of Sleep Medicine (AASM). The international Classification of Sleep Disorders - Revised, Diagnostic and Coding Manual. Rochester, AASM, 1997.
2. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. Washington, DC: American Psychiatric Association; 1994.
3. Baraitser M, Parkes JD. Genetic study of narcoleptic syndrome. J Med Genet 1978;15:254-259.
4. Billiard M, Pasquie-Magnetto V, Heckman M, Carlander B, Besset A, Zachariev Z, Eliaou JF, Malafosse A. Family studies in narcolepsy. Sleep 1994;17(8 Suppl):S54-S59.
5. Black J, Ohayon MM, Okun M, Guilleminault C, Mignot E, Zarcone V. The narcolepsy diagnosis: comparison between the Sleep-EVAL system and clinicians. Sleep 2001; 24(Abst. Suppl.):A328
6. Dement W, Zarcone W, Varner V et al. The prevalence of narcolepsy. Sleep Res 1972; 1:148.
7. Franceschi M, Zamproni P, Crippa D, Smirne S. Excessive daytime sleepiness: a 1-year study in an unselected inpatient population. Sleep 1982;5:239-247.
8. Gelineau J. De la narcolepsie. Gaz des Hop (Paris) 1880; 53:626-628, 635-637.
9. Guilleminault C, Mignot E, Grumet FC. Familial patterns of narcolepsy. Lancet 1989; 2(8676):1376-1379.
10. Hayduk R, Flodman P, Spence MA, Erman MK, Mitler MM. HLA haplotypes, polysomnography, and pedigrees in a case series of patients with narcolepsy. Sleep 1997;20:850-857.
11. Honda M, Honda Y, Uchida S, Miyazaki S, Tokunaga K. Monozygotic twins incompletely concordant for narcolepsy. Biol Psychiatry 2001;49:943-947.
12. Honda Y. Census of narcolepsy, cataplexy and sleep life among teenagers in Fujisawa city. Sleep Res 1979; 8:191.
13. Honda Y. Clinical features of narcolepsy: Japanese experiences. In: Honda Y, Juji T, eds. HLA in narcolepsy. Berlin: Springer-Verlag, 1988: 24-57.
14. Hosn R, Shapiro CM, Ohayon MM. Diagnostic concordance between sleep specialists and the sleep-EVAL system in routine clinical evaluations. J Sleep Res 2000; 9:86.
15. Hublin C, Kaprio J, Partinen M, Koskenvuo M, Heikkila K, Koskimies S, Guilleminault C. The prevalence of narcolepsy: an epidemiological study of the Finnish Twin Cohort. Ann Neurol 1994;35:709-716.
16. Krabbe E, Magnussen G. Familial aspects of narcolepsy. Tran Am Neurol Ass 1942; 17:149-173.
17. Mayer G, Lattermann A, Mueller-Eckhardt G, Svanborg E, Meier-Ewert K. Segregation of HLA genes in multicase narcolepsy families. J Sleep Res 1998 Jun;7:127-133.
18. Mignot E. Genetics of narcolepsy and other sleep disorders. Am J Hum Genet 1997;60:1289-1302.
19. Moscovitch A, Partinen M, Guilleminault C. The positive diagnosis of narcolepsy and narcolepsy's borderland. Neurology 1993;43:55-60.
20. Nevsimalova S, Mignot E, Sonka K, Arrigoni JL. Familial aspects of narcolepsy-cataplexy in the Czech Republic. Sleep 1997;20:1021-1026.
21. Ohayon M. Improving decision-making processes with the fuzzy logic approach in the epidemiology of sleep disorders. J Psychosom Res. 1999;47:297-311.
22. Ohayon M. Knowledge-Based System Sleep-EVAL: Decisional Trees and Questionnaires. Ottawa: National Library of Canada, 1995.
23. Ohayon MM, Guilleminault C, Zulley J, et al. Validation of the Sleep-EVAL system against clinical assessments of sleep disorders and polysomnographic data. Sleep 1999; 22:925-930.
24. Ohayon MM, Priest RG, Zulley J, Smirne S, Paiva T. Prevalence of narcolepsy symptomatology and diagnosis in the European general population. Neurology. 2002; 58:1826-1833.
25. Ohayon MM, Smirne S. Prevalence and consequences of insomnia disorders in the general population of Italy. Sleep Medicine 2002; 3:115-120.
26. Okun ML, Ohayon MM, Mignot E. The comparability of the Stanford Sleep Inventory and the Sleep-EVAL System in narcolepsy diagnosis. Sleep 2001; 24(Abst. Suppl.):A328-A329.
27. Partinen M, Hublin C, Kaprio J, Koskenvuo M, Guilleminault C. Twin studies in narcolepsy. Sleep 1994;17(8 Suppl):S13-S16.
28. Pollmacher T, Schulz H, Geisler P, Kiss E, Albert ED, Schwarzfischer F. DR2-positive monozygotic twins discordant for narcolepsy. Sleep 1990;13:336-343.
29. Roth B. Narcolepsy and hypersomnia. Basel: Karger, 1980 301 pp.
30. Tashiro T, Kanbayashi T, Iijima S, Hishikawa Y. An epidemiological study on prevalence of narcolepsy in Japanese. J Sleep Res 1992; 1(suppl):228.