JOURNALOF

www.elsevier.com/locate/jpsychires

Journal of Psychiatric Research 40 (2006) 273–279

PSYCHIATRIC

RESEARCH

Upper airway resistance syndrome: A long-term outcome study

Christian Guilleminault a,*, Ceyda Kirisoglu a, Dalva Poyares a, Luciana Palombini a,Damien Leger a, Mehran Farid-Moayer a, Maurice M. Ohayon b

a Stanford Sleep Disorders Clinic, 401 Quarry Road, Ste. 3301, Stanford, CA 94305, United Statesb Stanford Sleep Epidemiology Research Center, School of Medicine, Stanford University, CA, United States

Received 8 December 2004; received in revised form 17 March 2005; accepted 23 March 2005

Abstract

This prospective study aimed to assess symptomatic evolution of patients diagnosed with Upper Airway Resistance Syndrome(UARS) four and half years after the initial UARS diagnosis.

For this purpose, 138 UARS patients were contacted by mail between 43 and 69 months after the initial evaluation; 105responded to the letter and 94 patients accepted to undergo new clinical and polysomnographic evaluations. Initial and follow-up polysomnographic recordings were scored using the same criteria.Results: Of the 94 patients who completed the follow-up examination, none of them were using nasal CPAP. It was related to refusalby insurance providers to provide equipment based on initial apnea–hypopnea index (AHI) in 90/94 subjects. Percentage of patientswith sleep related-complaints significantly increased over the four and half year period: daytime fatigue, insomnia and depressivemood increased by 12 to 20 times. Reports of sleep maintenance sleep onset insomnia and depressive mood was significantlyincreased. Hypnotic, antidepressant and stimulant prescription increased from initial to follow-up visit (from 11.7% to 61.7%; from3.2% to 25.5% and from 0% to 9.6%, respectively) with antidepressant given as much for sleep disturbance as mood disorder. Thepolysomnography results at follow-up showed that 5 subjects had AHI compatible with Obstructive Sleep Apnea Syndrome (OSAS)but overall, respiratory disturbance index had no significant change. Total sleep time was significantly reduced compared to initialvisit.Conclusions: Many UARS patients remained untreated following initial evaluation. Worsening of symptoms of insomnia, fatigueand depressive mood were seen with absence of treatment of UARS.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Upper Airway Resistance Syndrome; Prospective survey; Sleep maintenance insomnia; Sleep onset insomnia; Hypnotics; Antidepres-sants; Stimulants

1. Introduction

Upper Airway Resistance Syndrome (UARS)(Guilleminault et al., 1993, 2000) is associated withabnormal respiratory effort, nasal airflow limitation,absence of obstructive sleep apnea, minimal pulse oxy-gen fluctuation with oxygen saturation (SaO2) equal or

0022-3956/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jpsychires.2005.03.007

* Corresponding author. Tel.: +1 650 723 6601; fax: +1 650 7258910.

E-mail address: cguil@stanford.edu (C. Guilleminault).

greater than 92%, and frequent nocturnal arousals orreflex brainstem activation (Morruzi, 1963). Patientswith UARS frequently complain of daytime tirednessand fatigue. It has been emphasized that these patientspresent more as patients with functional somatic syn-dromes than those presenting with Obstructive SleepApnea Syndrome (OSAS) (Gold et al., 2003; Guillemin-ault et al., 2001a,b,c). There have been questions aboutthe usefulness of identifying this syndrome indepen-dently from the OSAS, and questions on its long-termevolution (Douglas, 2000). The treatment of UARS is

274 C. Guilleminault et al. / Journal of Psychiatric Research 40 (2006) 273–279

not well clarified: nasal continuous positive airway pres-sure (nCPAP) was used in the initial report (Guillemin-ault et al., 1993) to demonstrate response to short termupper airway widening during sleep, but long term pre-scription has had variable success (Guilleminault et al.,2002). Finally not only the incidence of the syndromeis unknown but also is its natural history. To improveour knowledge on several of these points, we designeda prospective study on a retrospective cohort. We ques-tioned the outcome of the diagnostic evaluation and thetreatment recommendation performed on subjectslabeled UARS patients at least 4 years prior the startof the prospective study. We questioned if their illnesshad progressed, if their symptoms had changed and ifthe prescribed treatment (nCPAP) had been helpful.Due to health insurance policies and contractual agree-ments with health policy payers, regulations of the med-ical center, and views of the medical center by thesurrounding local medical community, our sleep disor-ders clinic functioned for many years as referral clinicperforming diagnosis and treatment recommendations,with a one time patient contact, and no further interac-tion if not initiated by the referring primary care physi-cian. Due to these specific conditions we built-up ourprospective protocol.

2. Materials and methods

2.1. Subjects

For this investigation, we attempted to contact allpatients (n = 138) who were diagnosed with UARS atStanford Sleep Disorders Clinic between 1995 and 1998.

2.2. Procedures

2.2.1. Initial evaluationDiagnosis of UARS was based on:

(1) Complaints of daytime fatigue, tiredness, or sleep-iness, report of nocturnal sleep disruption withvariable difficulty to go back to sleep or presenceof symptoms associated with sleep disorderedbreathing (Basiri and Guilleminault, 2000).

(2) Clinical evaluation: Patients completed validatedquestionnaires (Sleep Disorders Questionnaire(SDQ) (Douglass et al., 1994), Epworth SleepinessScale (ESS) (Johns, 1991)), Hamilton depressionscale, had a clinical evaluation that included deter-mination of body mass index (BMI), neck circum-ference, cranio-facial features with systematic faceand profile photos and usage of published clinicalscales (Friedman et al., 1999; Mallampati et al.,1985).

(3) 18 channels nocturnal polysomnography withusage of esophageal pressure transducer.

A diagnosis of UARS was made based on the associ-ation of complaints, clinical symptoms and polygraphicfindings. The scoring criteria for polysomnography werekept constant over time. The definition for presence ofan OSAS was based on an apnea–hypopnea index(AHI)P 5 events/hour, with definition of an apnea asan event with cessation of airflow for at least 10 s termi-nated with a drop of SaO2 > 3% or/and an EEG arousalP3 s. Hypopnea was defined as 30% or more reductionin nasal airflow channel compared to prior normalbreathing for duration of 10 s or more, terminated witha drop of SaO2 P 3% and/or an EEG arousalP3 s. Thedefinition of an UARS was based on an AHI (number ofapnea and hypopnea per hour of sleep) <5, and oxygensaturation >92% throughout the night (Guilleminaultet al., 2001a,b,c; Bao and Guilleminault, 2004; AASM,1999), other breathing abnormalities were seen basedon the analysis of the nasal cannula pressure transducersystem and of the esophageal pressure (Pes) curve (thescoring criteria are described below) (Bao and Guille-minault, 2004; AASM, 1999; Guilleminault et al.,1995; ASDA, 1992; Black et al., 2000). All patientsunderwent a second night of nocturnal polysomnogra-phy for determination of appropriate pressure necessaryfor control of the breathing disorder using nasal CPAP,subsequently they received a prescription for equipment,and were referred back to their primary care physician.

2.2.2. Follow-up

A mean of 4 and 1/2 years later, all 138 patients werecontacted by mail using the last address available in thehealth system directory. Up to three successive letterswere mailed out. The patients were asked to call us.The letter explained the goal of the phone call and theywere asked to come back for a research follow-up.

Patients who called were explained again about thegoals of the study, content of a consent form was readto them, and they gave an informed consent over thetelephone. Additionally, they were offered to come backfor a re-evaluation that included a new polysomnogram,and to discuss further treatment options. For subjectswho declined clinical follow-up, a 5-min questionnaireand the ESS (Johns, 1991) were administered on thephone. The brief questionnaire investigates changes ininitial symptoms and complaints, major changes inhealth status, and response to the nasal CPAPprescription.

Patients who accepted to come to clinic signedinformed consent, filled a new set of questionnairesincluding SDQ (Douglass et al., 1994), ESS (Johns,1991), Hamilton depression scale, and a 15-day sleepdiary. Additionally, they were evaluated for their cur-rent health status, medications, and underwent a physi-

C. Guilleminault et al. / Journal of Psychiatric Research 40 (2006) 273–279 275

cal exam with measurements of body mass index (BMI),neck circumference, and re-evaluation of oro-pharyn-geal anatomy (Friedman et al., 1999; Mallampatiet al., 1985). To eliminate the inter-scorer variability,the same clinician performed the clinical evaluations.

2.3. Polysomnography

The same montage was used in initial evaluation andfollow-up recording. Each time patients underwent aone-night polysomnography; the following variableswere monitored: electroencephalogram (EEG) (C3/A2,C4/A1, Fz/A1–A2, O1/A2), electrooculogram (EOG)of both eyes, chin and leg electromyograms (EMGs),electrocardiogram (ECG) with modified V2 lead, andrespiration by using nasal cannula pressure transducer,mouth thermistor, thoracic and abdominal impedancebelts, esophageal pressure manometer (Pes), neck micro-phone, and pulse oximetry (SaO2). Recordings were per-formed on the same computerized polygraphic sleepsystems (Sandman�, Kanata, Ont., Canada). Polysom-nographic recordings from the initial and follow-upevaluations were all scored by an individual blind todiagnostic and date of recording. The same interna-tional criteria were used for recording (AASM, 1999;ASDA, 1992; Rechtschaffen and Kales, 1968). Allrecords were re-scored following the same criteriaincluding the initial records of subjects that did notcome back from a second follow-up recording. Sleepand wake was scored following the international criteria(Rechtschaffen and Kales, 1968) and short arousals weredetermined following the American Sleep DisordersAssociation recommendations (ASDA, 1992). The crite-ria to score breathing during sleep were set before anyscoring and followed the international criteria (AASM,1999): Apnea and hypopnea were defined as indicatedabove (AASM, 1999). As indicated above, patientswho had an AHI < 5, and oxygen saturation >92%throughout the night were considered with UARS andapnea and hypopnea were defined using internationaldefinition. Other respiratory variables were scored toaffirm pathology in UARS patients. They were: (a) pres-ence of abnormal nasal cannula curve for a duration ofat least 10 s with disappearance of the normal roundshape of a breath and presence of a ‘‘flattening’’ of thecurve as previously defined by Hosselet et al. (1998),with a reduction of in nasal airflow between 3% and30% compared to prior normal breaths; (b) presenceof a change in the Pes curve lasting at least 4 breaths;these changes, as previously defined (Guilleminaultet al., 2001a,b,c), could be of 2 types: either with a morenegative peak end inspiratory pressure with each succes-sive breath called a Pes crescendo, or with a more nega-tive peak end inspiratory Pes reached in one or 2 breathsbut with maintenance of an elevated but leveled peak-end inspiratory value, called a ‘‘continuous sustained

effort’’ (Guilleminault et al., 2001a,b,c). These two indi-cators of increase respiratory effort may or may not havebeen associated with a ‘‘flattening’’ of the nasal cannula.The Pes events terminated with an EEG arousal, andfollowing the American Academy of Sleep Medicine rec-ommendations (AASM, 1999), were called ‘‘Respira-tory-Event-Related-Arousal’’ (RERA), or terminatedonly with an abrupt return to a lesser respiratory effortwith sudden decrease in negative peak end inspiratoryPes called a ‘‘Pes reversal’’(Guilleminault et al., 1995,2001a,b,c) but without an EEG arousal 3 s or longeror a SaO2 drop of 3% or more. These abnormal breath-ing events were tabulated together with the apnea andhypopnea; and a ‘‘respiratory disturbance index’’(RDI) was calculated (number of abnormal respiratoryevents/hour of sleep).

2.4. Statistical analysis

Kruskal–Wallis analysis of variance test with posthoc test with Bonferroni correction was used to compareresults on initial and follow-up evaluations. Mann Whit-ney U-test was used to compare two groups and nor-mally distributed pair t-test was used for repeatedmeasures. Percentages were compared using v2 statistics.P-value was set at <0.05. All statistical analysis was car-ried out using SPSS statistical package, version 11.5.

3. Results

3.1. Initial evaluation

About 138 subjects including 88 women were identi-fied from the computerized database. Men were signifi-cantly younger than women. (P-value < 0.05) (Table 1)No other significant differences were found betweenmen and women. Symptoms at entry are presented inTables 2A,2B(A and B). All patients complained of dis-rupted nocturnal sleep. Five percent of all subjects hadcomplaints of sleep onset insomnia. There were 7patients with a Hamilton D scale with a score between15 and 18 (4 of them were in the follow-up study). Allsubjects had a AHI < 5 events/hour, but presented ahigher RDI. All subjects had had a second polysomnog-raphy for nasal CPAP titration at time of initial visit andthe prescribed nasal CPAP pressure ranged between 5and 8 cm H2O.

3.2. Follow-up

One hundred and five patients (76% of retrospectivegroup) called in response to the mailings. They included72 women and 33 men (respectively 87.5% and 66% ofinitial group). Eleven of them (7 men) declined to have

Table 1Demographics of the patients on the first and current clinical evaluation

Total group Initial visit n = 138

Female (n = 88) Male (n = 50)

Age (yrs) 30.8 ± 4.1 (18–43) 26.3 ± 3.2 (22–31)BMI (kg m�2) 23.8 ± 1.3 (19–26) 24.1 ± 2.1 (19–27)AHI (events/h) 2.3 ± 0.5 (0.3–4) 2.4 ± 0.6 (0.4–4)RDI (events/h) 8.4 ± 3.2 (5.1–12.8) 9.5 ± 3.1 (5.2–14.2)ESS 9.1 ± 1.2 (6-12) 8.8 ± 1.9 (5–12)

Prospective cohort Initial visit n = 94 Follow-up visit n = 94

Female (n = 68) Male (n = 26) Female (n = 68) Male (n = 26)

Age (yrs) 30.0 ± 4.8 26.3 ± 3.5 35.1 ± 4.8 30.4 ± 3.5BMI (kg m�2) 23.4 ± 1.9 (19–26) 23.7 ± 2.3 (19–27) 23.6 ± 1.8 (19–26.5) 24.0 ± 3.0 (19–28.5)AHI (events/h) 2.2 ± 0.6 (0.3–4) 2.4 ± 0.6 (0.4–4) 2.4 ± 1.1 (0.4–4) 2.5 ± 1.2 (0.4–9)RDI (events/h) 8.6 ± 3.4 (5.3–12.8) 9.5 ± 3.4 (5.5–14.2) 8.3 ± 3.8 (5.1–13) 9.9 ± 4.1 (5.5–15)ESS 9.3 ± 1.3 (7–12) 8.6 ± 1.6 (5–11) 9.4 ± 1.4 (6–12) 9.1 ± 2.0 (6–12)

Values are depicted as mean, standard deviation, and range. BMI: body mass index, AHI: Apnea hypopnea index, RDI: Respiratory disturbanceindex, ESS: Epworth sleepiness scale. Top table presents group at first visit (n = 138), bottom table presents the prospective study group (n = 94) withresults when first seen and during current study. There is no significant difference between the total group and the prospective group at first visit forany of the considered variables.

Table 2AClinical complaints and symptoms at initial visit and follow-up visit

Complaints/symptoms Initial visit (n = 138) n (%) Follow-up visit with assumption X (n = 138) n (%) P < v2

(A) Subjects at initial visit

Snoring 101 (73.2) 90 (65.2) 0.02 (–)Daytime fatigue 109 (79) 94 (68) 0.02 (–)Disrupted nocturnal sleep 138 (100) 94 (68) 0.02Sleep maintenance insomnia 47 (34) 80 (58) 0.0001Difficulty concentrating 79 (57) 78 (56.5) 0.04Daytime irritability 51 (37) 69 (50) NsMorning headaches or heaviness of the head 22 (16) 36 (26) NsSleep onset insomnia 7 (5) 50 (36) 0.0001Daytime sleepiness 11 (8) 20 (15) NsDepressed mood 8 (5.8) 47 (34) 0.0001

If we assume that all subjects that did not responded to the offer of a new evaluation had appropriate treatment (an assumption – called hereassumption X – that we already know is inaccurate from the 11 subjects phone interview) and were controlled for their sleep disordered breathing, wecan see that the untreated subjects would still significantly positively impact the total results and a larger number of subjects would complained ofclinical symptoms except for snoring and daytime fatigue.

276 C. Guilleminault et al. / Journal of Psychiatric Research 40 (2006) 273–279

follow-up investigation at the clinic. Out of those 11patients, 2 subgroups could be identified:

(1) 3 women that had received nasal CPAP as pre-scribed, had seen clear improvement of their initialsymptoms, were having regular contact with theirphysicians and medical equipment providers andwere non-complainer;

(2) 8 patients who still had regular nocturnal snoring,complaints similar as at entry (n = 4) or reportedworsening of nocturnal sleep disruption (n = 4).ESS scores were not significantly changes from ini-tial visit. All these 8 patients were currentlyuntreated: phone short interviews revealed that 5patients never had their CPAP prescription filled,2 tried CPAP for one month but due to insuranceissues had their CPAP discontinue, and one

patient stopped his CPAP after 6 months of inter-mittent usage and nasal intolerance of thetreatment.

Ninety-four patients (68% of retrospective group),(68 women) agreed to come back for a new evaluationand new polysomnography and are called the ‘‘prospec-tive cohort’’. The follow-up investigation occurredbetween 43 and 69 months after initial evaluation (meanof 54.2 months).

3.2.1. Outcome of nasal CPAP prescription

None of the subjects were treated with nasal CPAP.About 90 out of 94 had had their request for nasalCPAP treatment rejected by health insurance at timeof initial request, and in 4 cases after first month ofdeliverance of equipment by medical equipment suppli-

Table 2B

Complaints/symptoms Initial visit (n = 94) n (%) Follow-up visit (n = 94) n (%) OR (95%CI) P< v2

(B) Complaints of patients with UARS in the prospective study at initial visit and at follow-up

Snoring 75 (79.8) 90 (95.7) 5.7 (1.9–17.5) 0.001Daytime fatigue 78 (83) 94 (100) 19.1 (2.5–>40) 0.0001Disrupted nocturnal sleep 94(100) 94 (100) 1.0 NsSleep maintenance insomnia 30 (32) 80 (85) 12.2 (6.0–24.9) 0.0001Difficulty concentrating 47 (50) 78 (83) 4.9 (2.5–9.6) 0.0001Daytime irritability 37 (39) 69 (63) 4.3 (2.3–7.9) 0.0001Morning headaches or heaviness of the head 18 (19) 36 (38) 2.6 (1.4–5.1) 0.004Sleep onset insomnia 5 (5.3) 50 (53.2) 20.2 (7.5–54.3) 0.0001Daytime sleepiness 9 (9.5) 20 (21.3) 2.6 (1.1–6.0) 0.03Depressed mood 7 (7.4) 47 (50) 12.4 (5.2–29.7) 0.0001

When subjects of the follow-up were investigated, a much larger number of subjects complained during the current study that at time of initial visit.The most important changes were on the number of subjects reporting sleep maintenance insomnia, sleep onset insomnia and depressed mood.Subjects complained of disturbed nocturnal sleep at first visit but did not mention ‘‘insomnia’’, but their views of their problem lead them toemphasize much more the sleep disturbance.

C. Guilleminault et al. / Journal of Psychiatric Research 40 (2006) 273–279 277

ers. The rational for the rejection of the nasal CPAP pre-scription was always the same: ‘‘patients did not meetcriteria for prescription of nasal CPAP’’ set by insurancecompany experts that also referred to criteria set by pro-fessional societies. General practitioner made no appeal.No referral back to sleep clinic was made based on lowAHI and belief by practitioners that sleep disorderedbreathing had been ruled-out due to low AHI.

3.2.2. Clinical complaints

All subjects presented sleep related complaints at thefollow-up interview. None of the subjects had spontane-ous improvement. All mentioned persistence of sleeprelated problems. Tables 2A,2B(A and B) present com-plaints at initial and follow-up visits. As seen, frequencyof complaints had significantly increased during theelapsed time between the 2 evaluations. The highestincreases were seen for fatigue, insomnia, and scores atthe Hamilton depression scale with 47 subjects with ascore of 15 or above, (with 5 subjects with scoresbetween 19 and 22).

3.2.3. Medication intake

As disrupted nocturnal sleep, poor sleep and sleeponset insomnia were frequent complaints, patients weresymptomatically treated and a significant increase inhypnotic medication prescription was seen at currentvisit: At time of first visit 19/138 (14%) patients weretaking hypnotic medication for their nocturnal sleep dis-turbances. These 19 included 11 subjects that were in theprospective cohort (11.7%). At the follow-up evaluation58 subjects (61.7%) were using hypnotic medicationsmore than twice per week prescribe by private practi-tioner (P = 0.0001; v2 statistics).

At initial visit 6/138 (4%) subjects (3 of them – 3.2% –in the follow-up cohort) had had prescription of specificserotonin re-uptake inhibitors (SSRI). At the follow-upevaluation, 24 (25.5%) received a SSRI, Trazodone or

Venlafaxine. Mostly the medication was prescribed bya general practitioner for insomnia more than depressivemood; 8 subjects only had been referred to apsychiatrist.

In summary, at the time of first visit 15/138 subjects(16%) had received hypnotic or antidepressant medica-tions due to poor sleep mostly and sometime associateddepressive mood. At the follow-up evaluation, 82/94(87.2%) had sought medical help, and were prescribedmedications in relation with their insomnia and in somecases due to the depressive mood.

Stimulant medication: No patient received stimulantat initial visit, but 9/94 subjects (9.6%) were receivingmodafinil (200 mg/AM) due to daytime complaint; 6/9used both a medication at bedtime due to poor sleepand modafinil in the morning due to impaired alertnessand difficulty at work. Clinical interviews and systemevaluations showed no significant change in blood pres-sure, cardio-vascular or neurological status.

3.3. Polysomnography

The overall AHI and RDI at follow-up, when the 94subjects were tabulated, were not significantly differentfrom those measured at the initial visit (see Table 1).Individual results indicated however, that 5 out of 94subjects (3 men) were now diagnosed with OSAS. Theyhad an AHI P 5 events/hour (mean 7 ± 2.3, range 5–11). The changes were associated with increase inweight, particularly in the subject with the highestincrease in AHI (AHI = 11) had a 8 kg weight gain.However, the overall BMI did not significantly changefor the total group (Table 1). Lowest SaO2 was verymoderately affected in these 5 subjects with OSAS, asthe overall lowest reading seen in association with oneevent was 88%. Total recording time was kept constantbetween the 2 sleep studies, with 8.00 h of dark time.The mean total sleep time (TST) was 355 ± 31 min at

278 C. Guilleminault et al. / Journal of Psychiatric Research 40 (2006) 273–279

the initial evaluation but it was only 302 ± 28 min(P = 0.001) at the follow-up evaluation, with longertime awake from light out supporting claims of insom-nia. Light out was based on patient preference andwas not significantly change from first visit. Since respi-ratory events are calculated based on sleep time, thisreduction in TST did not impacted calculation.

In summary, the 94 untreated UARS patients, for thelarge majority (n = 89) did not evolve toward OSAS in 4and half years; the very few that did, increase their bodymass index. But the complaints involved a much largergroup of subjects and were much more numerous (Table2BB). Practitioners responded to these complaints withsymptomatic treatments that had limited success. Evenif we assume that all subjects that did not respond toour letters and/or to the offer to have a new investiga-tion were without UARS (assumption called ‘‘assump-tion X’’, and assumption that we already know asflawed due to response to short questionnaire for 11 sub-jects), there would still be a significantly higher percent-age of subjects with complaints and with morecomplaints at current study (see Table 2AA). Theincrease in complaint is associated with a shorter TSTand more sleep disruption after light out, at currentrecording.

4. Discussion

This is the first study that examines the natural evo-lution of patients diagnosed with UARS. The resultsclearly showed that sleep and daytime symptoms inuntreated UARS increased over time and there was nocase of spontaneous remission of symptoms. More spe-cifically, the likelihood of having daytime fatigue,insomnia or depressive mood had a 12 to 20 timesincrease over a 4-year period. One of the direct conse-quences was the increase in hypnotic and antidepressantprescription: 11.7% and 3.2% and at initial visit com-pared with 61.7% and 25.5% at the follow-up.

The first conclusion is that despite claims that UARSis recognized, integrated for some in the OSAS (Doug-las, 2000), and treated; there was a clear refusal to sup-port treatment by medical insurances in the late 90sbased on definitions provided by professional medicalsocieties, and secondary negation of presence of the syn-drome in the medical community. The difference in clin-ical presentation and polygraphic findings betweenUARS and the better known OSAS includes complaintsusually associated with functional somatic syndromesand AHI < 5 events/hour. The lack of education onUARS in the medical community, and the controversiesrelated to its position within sleep disordered breathing,lead practitioners to accept the decisions from medicalinsurance plans, and lead them to treat patients symp-tomatically and without challenging the refusal of more

specific treatment. One clear motivation for many sub-jects to participate to the study was assurance to addressthe initially diagnosed health problem.

Our study also shows that UARS patients withweight (and BMI) increase may evolve toward a definedOSAS, but this concerned a limited number of patients;the vast majority of patients did not increase theirweight significantly and were still UARS four and halfyears later. On the other hand, the number of patientswith clinical complaints related to their poor sleep andits daytime consequences greatly increased over timedespite absence of progression toward OSAS, andabsence of polysomnographic evidence of worsening ofsleep disordered breathing indices. One of the interestingoutcomes was that the worsening of nocturnal sleep dis-turbance not only involved a greater number of com-plainers but was also associated with a greater sleepdisruption as demonstrated by polysomnography. Thisworsening was associated with a greater amount ofwakefulness after light out; and the absence of signifi-cant increase in the number of abnormal breathingevents between initial and follow-up polysomnographymay be related to this increase in amount of awake timeduring the night. Interestingly, subjects developed sleeponset insomnia and not only maintenance insomnia.These findings have to be placed together and may givesome insights on the development of chronic insomnia:We still have little knowledge on the factors contributingto the development of chronic insomnia. Many chronicillnesses may lead to insomnia complaint, but the spe-cific role of a sleep disorder as a predisposing factorfor development of sleep onset insomnia is not wellunderstood. In a previous study on post-menopausalwomen with chronic insomnia, we have shown that sleepdisordered breathing is not only common, but also thattreating the underlying breathing disorder can improvesome of the complaints (Guilleminault et al., 2002).Moreover, this improvement was observed using nasalCPAP that is not the most effective treatment forUARS. Our study showed that untreated UARS is asso-ciated with chronic sleep onset and sleep maintenanceinsomnia and depressive mood. And the chronic com-plaint led to chronic prescription of hypnotics and anti-depressant medications in about 60% of the prospectivecohort, without addressing the underlying cause of thesleep disruption.

Another outcome is that UARS patients present verycommonly with functional somatic complaints (Goldet al., 2003) including insomnia, headache, myalagia,daytime fatigue. These symptoms do not orient immedi-ately toward a SDB in general practice, as UARS is notnecessarily on the differential diagnosis list of theserelated complaints. If subjects are referred to sleep spe-cialists it is important not only to perform appropriateinvestigation but also to follow the patient and providethe referring physician with help in securing the appro-

C. Guilleminault et al. / Journal of Psychiatric Research 40 (2006) 273–279 279

priate treatment for the patient, as non specialists maydecide, in absence of clear support and help from thespecialist, to treat symptomatically using hypnotics,antidepressant, and stimulant drugs, without valid treat-ment response.

Overall the outcomes found with this study are disap-pointing: Despite efforts to emphasize the existence ofthe UAR syndrome and to emphasize need to treat thesleep disordered breathing very few subjects were in facttreated at the beginning of the new millennium; anddespite polysomnography, failure to convince the refer-ring practitioner at the initial visit did not bring thepatient back to specialist. But absence of treatment leadto development of more complaints that lead to an inad-equate treatment response; the UARS and its non-rec-ognition is a challenge for the field of Sleep Medicine.

References

American Academy of Sleep Medicine: Sleep-related breathing disor-ders in adults: recommendations for syndrome definition andmeasurement techniques in clinical research. The Report of anAmerican Academy of Sleep Medicine Task Force. Sleep1999;22:667–89.

American Sleep Disorders Association. EEG arousals: scoring rulesand examples: a preliminary report from the Sleep Disorders AtlasTask Force of the American Sleep Disorders Association. Sleep1992;15:173–84.

Bao G, Guilleminault C. Upper Airway Resistance Syndrome, 10 yearslater. Current Opinion in Pulmonary Medicine 2004;10:461–7.

Basiri A, Guilleminault C. Clinical features and evaluation ofObstructive Sleep Apnea–Hypopnea Syndrome. In: Kryger MH,Roth T, Dement WC, editors. Principle and practice of sleepmedicine. Philadelphia: Saunders; 2000. p. 869–78.

Black J, Guilleminault C, Colrain I, Carillo O. Upper airway resistancesyndrome: Central EEG power and changes in breathing effort.American Journal of Respiratory and Critical Care Medicine2000;162:406–11.

Douglas NJ. Upper Airway Resistance Syndrome is not a distinctsyndrome. American Journal of Respiratory and Critical CareMedicine 2000;161:1413–6.

Douglass AB, Bornstein R, Nino-Murcia G, Keenan S, Miles L,Zarcone Jr VP, et al.. The Sleep Disorders Questionnaire. I.Creation and multivariate structure of SDQ. Sleep 1994;17:160–7.

Friedman M, Tanyeri H, La Rosa M, Landsberg R, Vaidyanathan K,Pieri S, et al.. Clinical predictors of obstructive sleep apnea.Laryngoscope 1999;109:1901–7.

Gold AR, Dipalo F, Gold MS, O’Hearn D. The symptoms and signsof upper airway resistance syndrome: a link to the functionalsomatic syndromes. Chest 2003;123:87–95.

Guilleminault C, Black JE, Palombini L, Ohayon M. A clinicalinvestigation of obstructive sleep apnea syndrome (OSAS) andupper airway resistance syndrome (UARS) patients. Sleep Medi-cine 2000;1:51–6.

Guilleminault C, Faul JL, Stoohs R. Sleep-disordered breathing andhypotension. American Journal of Respiratory and Critical CareMedicine 2001a;164:1242–7.

Guilleminault C, Kim YD, Stoohs R. Upper airway resistancesyndrome. Oral and Maxillofacial Surgical Clinics of North-America 1995;7:243–56.

Guilleminault C, Palombini L, Poyares D, Chowdhuri S. Chronicinsomnia, premenopausal women and sleep disordered breathing:part 2. Comparison of non-drug treatment trials in normal breathingand UARS postmenopausal women complaining of chronic insom-nia. Journal of Psychosomatic Research 2002;53:617–23.

Guilleminault C, Poyares D, Palombini L, Koester U, Pelin Z, Black J.Variability of respiratory effort in relation to sleep stages in normalcontrols and upper airway resistance syndrome patients. SleepMedicine 2001b;2:397–405.

Guilleminault C, Poyares D, Palombini L, Koester U, Pelin Z, Black J.Variability of respiratory effort in relationship with sleep stages innormal controls and upper airway resistance syndrome patients.Sleep Medicine 2001c;2:397–406.

Guilleminault C, Stoohs R, Clerk A, Cetel M, Maistros P. A cause ofexcessive daytime sleepiness. The upper airway resistance syn-drome. Chest 1993;104:781–7.

Hosselet JJ, Norman RG, Ayappa I, Rapoport DM. Detection of flowlimitation with a nasal cannula/pressure transducer system. Amer-ican Journal of Respiratory and Critical Care Medicine 1998;157(5Pt 1):1461–7.

Johns MW. A new method for measuring daytime sleepiness: theEpworth sleepiness scale. Sleep 1991;14:540–5.

Mallampati SR, Gatt SP, Gugino LD, Desai SP, Waraksa B,Freiberger D, et al.. A clinical sign to predict difficult trachealintubation: a prospective study. Canadian Anaesthestists’ SocietyJournal 1985;32:429–34.

Morruzi G. Active process in the brain-stem during sleep. The HarveyLectures series, vol. 58. New York: Academic Press; 1963. p.233–97.

Rechtschaffen A, Kales A. A manual of standardized technologytechniques and scoring systems for sleep stages of humansubjects. Los Angeles: UCLA Brain Information Service, BrainResearch Institute; 1968.