5
Original Article Sleep syncope: Important clinical associations with phobia and vagotonia L. Busweiler a,1 , D.L. Jardine b,, C.M. Frampton c,2 , W. Wieling a,1 a Department of Internal Medicine, Academic Medical Centre, Amsterdam, The Netherlands b Department of General Medicine, Christchurch Hospital, Christchurch, New Zealand c Department of Medicine, Christchurch School of Medicine, New Zealand article info Article history: Received 17 December 2009 Received in revised form 19 April 2010 Accepted 24 April 2010 Keywords: Syncope Bradycardia Autonomic regulation Specific phobia Blood injection injury phobia Vagotonia abstract Objectives: To compare demographic and clinical data from patients with sleep syncope to those of patients with ‘‘classical” vasovagal syncope [VVS] collected over the last 8 years. Design: Retrospective case-controlled study. Setting: Syncope unit. Patients and methods: Fifty-four patients with a history suggestive of one or more episodes of sleep syn- cope (group SS) were matched for age and gender to 108 patients with VVS (control group). Patients and methods: A syncope questionnaire was completed immediately before tilt-testing and included frequency, age-of-onset and severity of episodes; situations, postures and perceived triggers; lifetime prevalence of specific phobias; and symptoms during syncope. Results: Group SS were mainly women (65%), mean age of 46 ± 2.1 years, with a mean lifetime total of 5.4 ± 0.83 episodes of sleep syncope. Compared to controls, SS episodes were more likely to start in child- hood, 26.9% versus 50% (p = 0.005), and more severe, score 2.40 ± 0.11 versus 2.81 ± 0.15 (p = 0.03). In group SS: syncope onset whilst lying down was more frequent, 4.6% versus 32.7% (p = 0.001); the lifelong prevalence of any specific phobia was higher, 32.4% versus 74.5% (p = 0.001), in particular blood injection injury (BII) phobia, 19.4% versus 57.4% (p = 0.001); and during attacks, distressing vagal symptoms were more frequent, e.g., abdominal discomfort, 13.9% versus 72.2% (p = 0.001). Conclusion: Sleep syncope is not rare and is characterised by lifelong, intermittent but severe episodes of vasovagal syncope which may occur in the horizontal position, with distressing abdominal symptoms. BII phobia is strongly associated and may be a predisposing factor or a co-existent disorder in these patients. Ó 2010 Elsevier B.V. All rights reserved. 1. Introduction We have previously described a group of patients who com- plained of syncope interrupting their nocturnal sleep [1]. They woke up feeling faint, often with pronounced abdominal symp- toms, and lost consciousness either in bed or immediately upon standing. These patients also fainted in response to typical emo- tional and postural stimuli in the course of daily living, and during head-up tilt in the laboratory. More recently we showed that pa- tients with sleep syncope and ‘‘classical” vasovagal syncope (VVS) had very similar sympatho-vagal responses to both barore- flex and non-baroreflex stimuli [2]. Sleep syncope has recently been proposed to be a new form of vasovagal syncope [3]. We also noticed that the clinical features of VVS and sleep syn- cope were not alike in all respects: first, episodes of sleep syncope often started in childhood and many patients with sleep syncope seemed to have specific phobias, in particular blood injection in- jury (BII); second, sleep syncope often occurred when the patient was horizontal and was associated with intense vagal symptoms (e.g., nausea, abdominal discomfort, urge to defecate) [1,2]. How- ever, details of these clinical features have not yet been reported in a large group of patients. With this in mind, the aims of the pres- ent study were to assess these clinical features in a large sample of patients with sleep syncope and compare them to a matched group with VVS. Clarifying differences between sleep syncope and VVS may provide further insight into possible physiological mecha- nisms underlying sleep syncope. 2. Patients and methods We used a retrospective case-controlled study design. 2.1. Patients Clinical data of patients diagnosed with sleep syncope were collected over an 8-year period (2002–2009 inclusive) from the 1389-9457/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.sleep.2010.04.013 Corresponding author. Tel.: +64 33786005; fax: +64 33641025. E-mail addresses: [email protected] (D.L. Jardine), [email protected]. nz (C.M. Frampton), [email protected] (W. Wieling). 1 Tel.:+3120 5669111; fax:+3120 6919658. 2 Tel.:+64 36401847. Sleep Medicine 11 (2010) 929–933 Contents lists available at ScienceDirect Sleep Medicine journal homepage: www.elsevier.com/locate/sleep

Sleep syncope: Important clinical associations with phobia and vagotonia

Embed Size (px)

Citation preview

Page 1: Sleep syncope: Important clinical associations with phobia and vagotonia

Sleep Medicine 11 (2010) 929–933

Contents lists available at ScienceDirect

Sleep Medicine

journal homepage: www.elsevier .com/locate /s leep

Original Article

Sleep syncope: Important clinical associations with phobia and vagotonia

L. Busweiler a,1, D.L. Jardine b,⇑, C.M. Frampton c,2, W. Wieling a,1

a Department of Internal Medicine, Academic Medical Centre, Amsterdam, The Netherlandsb Department of General Medicine, Christchurch Hospital, Christchurch, New Zealandc Department of Medicine, Christchurch School of Medicine, New Zealand

a r t i c l e i n f o a b s t r a c t

Article history:Received 17 December 2009Received in revised form 19 April 2010Accepted 24 April 2010

Keywords:SyncopeBradycardiaAutonomic regulationSpecific phobiaBlood injection injury phobiaVagotonia

1389-9457/$ - see front matter � 2010 Elsevier B.V. Adoi:10.1016/j.sleep.2010.04.013

⇑ Corresponding author. Tel.: +64 33786005; fax: +E-mail addresses: [email protected] (D.L

nz (C.M. Frampton), [email protected] (W. Wielin1 Tel.:+3120 5669111; fax:+3120 6919658.2 Tel.:+64 36401847.

Objectives: To compare demographic and clinical data from patients with sleep syncope to those ofpatients with ‘‘classical” vasovagal syncope [VVS] collected over the last 8 years.Design: Retrospective case-controlled study.Setting: Syncope unit.Patients and methods: Fifty-four patients with a history suggestive of one or more episodes of sleep syn-cope (group SS) were matched for age and gender to 108 patients with VVS (control group).

Patients and methods: A syncope questionnaire was completed immediately before tilt-testing andincluded frequency, age-of-onset and severity of episodes; situations, postures and perceived triggers;lifetime prevalence of specific phobias; and symptoms during syncope.Results: Group SS were mainly women (65%), mean age of 46 ± 2.1 years, with a mean lifetime total of5.4 ± 0.83 episodes of sleep syncope. Compared to controls, SS episodes were more likely to start in child-hood, 26.9% versus 50% (p = 0.005), and more severe, score 2.40 ± 0.11 versus 2.81 ± 0.15 (p = 0.03). Ingroup SS: syncope onset whilst lying down was more frequent, 4.6% versus 32.7% (p = 0.001); the lifelongprevalence of any specific phobia was higher, 32.4% versus 74.5% (p = 0.001), in particular blood injectioninjury (BII) phobia, 19.4% versus 57.4% (p = 0.001); and during attacks, distressing vagal symptoms weremore frequent, e.g., abdominal discomfort, 13.9% versus 72.2% (p = 0.001).Conclusion: Sleep syncope is not rare and is characterised by lifelong, intermittent but severe episodes ofvasovagal syncope which may occur in the horizontal position, with distressing abdominal symptoms. BIIphobia is strongly associated and may be a predisposing factor or a co-existent disorder in these patients.

� 2010 Elsevier B.V. All rights reserved.

1. Introduction

We have previously described a group of patients who com-plained of syncope interrupting their nocturnal sleep [1]. Theywoke up feeling faint, often with pronounced abdominal symp-toms, and lost consciousness either in bed or immediately uponstanding. These patients also fainted in response to typical emo-tional and postural stimuli in the course of daily living, and duringhead-up tilt in the laboratory. More recently we showed that pa-tients with sleep syncope and ‘‘classical” vasovagal syncope(VVS) had very similar sympatho-vagal responses to both barore-flex and non-baroreflex stimuli [2]. Sleep syncope has recentlybeen proposed to be a new form of vasovagal syncope [3].

We also noticed that the clinical features of VVS and sleep syn-cope were not alike in all respects: first, episodes of sleep syncope

ll rights reserved.

64 33641025.. Jardine), [email protected]).

often started in childhood and many patients with sleep syncopeseemed to have specific phobias, in particular blood injection in-jury (BII); second, sleep syncope often occurred when the patientwas horizontal and was associated with intense vagal symptoms(e.g., nausea, abdominal discomfort, urge to defecate) [1,2]. How-ever, details of these clinical features have not yet been reportedin a large group of patients. With this in mind, the aims of the pres-ent study were to assess these clinical features in a large sample ofpatients with sleep syncope and compare them to a matched groupwith VVS. Clarifying differences between sleep syncope and VVSmay provide further insight into possible physiological mecha-nisms underlying sleep syncope.

2. Patients and methods

We used a retrospective case-controlled study design.

2.1. Patients

Clinical data of patients diagnosed with sleep syncope werecollected over an 8-year period (2002–2009 inclusive) from the

Page 2: Sleep syncope: Important clinical associations with phobia and vagotonia

930 L. Busweiler et al. / Sleep Medicine 11 (2010) 929–933

syncope clinic at Christchurch Hospital. Referrals to clinic (approx-imately 120 per year) were made by general practitioners, cardiol-ogists and neurologists.

The referral group consisted mainly of patients with recurrentreflex syncope orthostatic hypotension and any of the following:diagnostic uncertainty; distress secondary to the attacks; or signif-icant disability, e.g., loss of job, driving licence. ECG’s were under-taken and examined for conduction abnormalities. Patients wereexcluded if there was any form of chronic autonomic dysfunction,psychogenic syncope, arrhythmia, epilepsy, hypoglycaemia or par-asomnia. Initially, patients who gave a history of one or more epi-sodes of syncope interrupting sleep at night were put into the sleepsyncope group (SS, n = 54) [1]. In some SS patients, loss of con-sciousness occurred whilst lying in bed, in others immediatelyafter leaving the bed in an effort to get to the toilet. All of the SSgroup reported fainting in the daytime as well, in response to com-mon triggers (posture change, heat and emotions). Then each pa-tient in group SS was individually matched by age and gender totwo control patients with VVS [4], who were assessed at the syn-cope clinic over the same period. The control group (C), therefore,consisted of 108 patients with VVS who had never had episodesinterrupting sleep.

2.2. Questionnaire

A special questionnaire was designed to collect detailed infor-mation regarding the patient’s history of episodes of syncope(appendix). Both groups C and SS received the questionnaire aweek before the tilt test. The answers were checked by one ofthe authors (D.L.J.) during a visit to the clinic just before undergo-ing the tilt test. Information was gathered under four categories: (i)Syncope load: frequency, onset and severity of episodes; (ii) syn-cope predisposition: situations, postures and perceived triggers;(iii) specific phobias and other medical or psychiatric conditions;and (iv) syncope symptoms during episodes. For all categorieswe only included data from episodes during which there was totalloss of consciousness (LOC). Thus frequency scores were estimatedfrom lifetime history: 1 = 1 episode per year, 2 = 1 per month, 3 = 1per week, 4 = >1 per week; and lifetime duration score:1 = <3 months, 2 = 3–6 months, 3 = 6–12 months, 4 = >12 months.For approximate age of onset we asked whether episodes startedduring childhood (before adolescence). For severity of attacks weasked patients to estimate how often their episodes resulted inLOC: 1 = never, 2 = some, 3 = most, 4 = all. If LOC occurred we askedthem to estimate its duration: 1 = seconds, 2 = 1 min, 3 = 5 min,4 = >5 min. For situations, postures and perceived triggers, the an-swer was recorded as ‘‘yes” if the patient had ever fainted in thatsituation in his or her lifetime. For lifetime prevalence of (co-exis-tent) specific phobias we asked whether patients had ever faintedin response to fear or an unpleasant situation. For example, forBII phobia, patients were asked: ‘‘are you afraid of, or dislike ofthe sight of blood?”; ‘‘if so, have you always been like this?”;‘‘can you look at your own blood being taken?”; ‘‘have you faintedduring blood tests, injections, injury or pain?” For claustrophobia:‘‘would you go down a cave?”; ‘‘would you be OK if you werelocked in a wardrobe?” For crowd phobia: ‘‘would you feel uncom-fortable in a large crowd, e.g., at a rugby game or rock concert?”;‘‘do you have to go outside sometimes when you are in a crowdedroom?” To qualify for the diagnosis of specific phobia, patients hadto (i) have a chronic inappropriate fear of the stimulus; (ii) describesome form of avoidance behaviour; and (iii) admit to anticipatoryanxiety if avoidance was impossible. These are essentially theDSM-IV criteria for specific phobia [5]. However, the vast majorityof our patients were not distressed about having the phobia anddid not think it interfered significantly with their normal routine,work function or social activities. Lifetime prevalence of medical

conditions including heart problems, hypertension, stroke, epi-lepsy and migraine was recorded. Other psychiatric conditions,including obsessive–compulsive disorder, post-traumatic stressdisorder, separation anxiety disorder, agoraphobia, social phobiaand panic attacks, were included in this category. For symptomsduring syncope attacks, answers were recorded as affirmative ifthe symptom had ever been present during an attack.

2.3. Tilt-table testing

After written consent, patients underwent tilt-table testingaccording to a protocol approved by the Canterbury Ethics Com-mittee. Patients were positioned in the horizontal position on thetilt table and a Finapres finger cuff was wrapped around the middlefinger of the right hand. Blood pressure and heart rate were re-corded continuously for 15 min prior to head-up tilt. Patients weretilted to the head-up 70-degree position over a 30-s period. Pa-tients were tilted back to the horizontal position in the event ofpresyncope symptoms and a fall in systolic BP to 80 mm Hg or less.If the patient remained well at 20 min, GTN spray � 1 was givenand tilt continued for 10 min further unless presyncope developed.

2.4. Statistics

Comparisons among data obtained from the questionnaires andtilt tests were undertaken using SPSS. Standard 2-tailed t-tests andPearson Chi-square analyses were used for comparisons of abso-lute and percentage values, respectively. Fisher’s exact test wasused for low percentage values. The Mann–Witney test was usedfor comparing ordinal rating scales.

3. Results

Over the eight-year period we collected 54 patients with sleepsyncope, approximately 6% of patients assessed in clinic. They weremainly women (65%), mean age of 46 ± 2.1 years, with a mean life-time total of 5.4 ± 0.83 episodes of sleep syncope. Descriptive fea-tures of groups C and SS are listed in Tables 1–4.

3.1. Syncope load (frequency, onset, and severity of episodes)

Between groups, the lifetime frequency of attacks and numberof attacks during the previous year was similar (Table 1). Com-pared to group C, group SS had a higher rating scale for lifetimeduration of episodes (3.44 ± 0.09 versus 3.85 ± 0.09 [p = 0.001]),and the attacks were more likely to start in childhood (26.9% ver-sus 50% [p = 0.005]). Also group SS was more likely to becomeunconscious during an episode: rating scale for relative frequencyof LOC 2.40 ± 0.11 in group C versus 2.81 ± 0.15 (p = 0.03) in groupSS. Tilt test results were similar between groups with regard tobaseline haemodynamics, sensitivity, tilt time, and requirementfor GTN.

3.2. Syncope predisposition (situations, postures and perceived triggersfor syncope)

Compared to group C, group SS reported fainting morecommonly triggered by an unpleasant experience (e.g., blood tests,pain), 30.6% versus 55.6% (p = 0.002) or going to the toilet 11.1%versus 40.7% (p = 0.001; Table 2). Syncope occurring in the horizon-tal position (lying down) was far more frequent in group SS: 4.6%versus 32.7% (p = 0.001). However, in group SS, episodes were lesslikely to be triggered by standing up (88.9% versus 77.8% [p = 0.06])and exercise (34.3% versus 13% [p = 0.004]).

Page 3: Sleep syncope: Important clinical associations with phobia and vagotonia

Table 1Syncope load: frequency, onset, and severity of episodes; tilt results.

Group C[n = 108]

Group SS[n = 54]

p

Frequency [arbitrary units] 1.66 ± 0.09 1.49 ± 0.08 0.76Episodes last year 3.10 ± 0.32 3.70 ± 0.24 0.23Lifetime duration [arbitrary units] 3.44 ± 0.09 3.85 ± 0.09 0.001Childhood syncope [Y/N] 27 [26.9%] 27 [50%] 0.005How often LOC [arbitrary units] 2.40 ± 0.11 2.81 ± 0.15 0.03Duration of LOC [arbitrary units] 1.35 ± 0.12 1.37 ± 0.14 0.52Tilt time [min] 22.78 ± 0.55 21.80 ± 0.89 0.32Systolic BP supine [mm Hg] 131.2 ± 2.4 139.0 ± 3.5 0.07Systolic BP after 1 min tilt [mm Hg] 139.3 ± 2.4 144.0 ± 3.1 0.16Heart rate supine [bpm] 70.4 ± 1.2 67.4 ± 1.6 0.15Heart rate after 1 min tilt [bpm] 81.7 ± 1.4 77.8 ± 2.0 0.11GTN [Y/N] 96 [88.9%] 45 [83.6%] 0.32Tilt positive [Y/N] 93 [86.1%] 47 [87.0%] 0.46

Syncope load variables refer to episodes of total loss of consciousness (LOC). Fre-quency was scored from lifetime history: 1 = 1 episode per year, 2 = 1 per month,3 = 1 per week, 4 = >1 per week; Lifetime duration score: 1 = <3 months, 2 = 3–6 months, 3 = 6–12 months, 4 = >12 months; childhood syncope was recordedaffirmative if episodes started before adolescence; how often LOC refers to whenLOC occurred during episodes: 1 = never, 2 = some, 3 = most, 4 = all; duration of LOCscore (approximate): 1 = s, 2 = 1 min, 3 = 5 min, 4 = >5 min.Tilt time refers to time from tilt-up to tilt back (at onset of presyncope, or 30 min);GTN refers to glyceryl trinitrate spray routinely administered sublingually at 20 mintilt; tilt positive refers to presyncope (symptomatic fall in systolic BP < 80 mm Hg)during the tilt test. Absolute values are expressed as means ± standard error. Othervalues and percentages refer to affirmative answer.

Table 2Syncope situations, postures and perceived triggers.

Group C[n = 108]

Group SS[n = 54]

p

Unpleasant emotion [Y/N] 33 [30.6%] 30 [55.6%] 0.002Going to toilet [Y/N] 12 [11.1%] 22 [40.7%] 0.001Ambient heat [Y/N] 36 [33.3%] 20 [37.0%] 0.64Change to upright posture [Y/N] 96 [88.9%] 42 [77.8%] 0.06Prolonged standing [Y/N] 56 [51.9%] 20 [37.0%] 0.08Head movement [Y/N] 5 [4.6%] 7 [13.0%] 0.11Sitting [Y/N] 21 [19.4%] 8 [14.8%] 0.47Lying [Y/N] 5 [4.6%] 17 [32.7%] 0.001Exercising [Y/N]a 37 [34.3%] 7 [13.0%] 0.004

All variables refer to lifelong prevalence. Each variable was recorded affirmative ifpatients recalled it to be associated with one or more episodes of syncope (total lossof consciousness) in their lifetime (lifetime prevalence).

a For exercise, presyncope episodes were included. Values and percentages referto affirmative answers.

Table 3Specific phobias and other medical/psychiatric conditions (lifelong prevalence).

Group C[n = 108]

Group SS[n = 54]

p

Any specific phobia [Y/N] 35 [32.4%] 40 [74.5%] 0.001Blood injection injury phobia [Y/N] 21 [19.4%] 31 [57.4%] 0.001Syncope [Y/N] 26 [24.1%] 32 [59.3%] 0.001Claustrophobia [Y/N] 15 [13.9%] 16 [29.6%] 0.02Syncope [Y/N] 7 [6.5%] 6 [11.1%] 0.36Crowd phobia [Y/N] 6 [5.6%] 7 [13.0%] 0.13Syncope [Y/N] 29 [26.9%] 14 [25.9%] 0.9Other medical/psychiatric [Y/N] 26 [24.1%] 8 [14.8%] 0.17Anti-depressant medication [Y/N] 12 [11.3%] 4 [7.5%] 0.46

All variables refer to lifelong prevalence. Any specific phobia refers to bloodinjection injury-, claustro- or crowd-phobia. Syncope refers to total loss of con-sciousness occurring in association with the phobic stimulus. Other medical/psy-chiactric refers to co-existing medical and psychiatric conditions (e.g., migraine,diabetes, depression, and anxiety). Anti-depressant medication refers to long-termdrug therapy with tricyclics or serotonin uptake inhibitors. Values and percentagesrefer to affirmative answers.

Table 4Syncope symptoms during episodes.

Group C[n = 108]

Group SS[n = 54]

p

Light headed [Y/N] 85 [78.7%] 45 [83.3%] 0.49Hot [Y/N] 40 [37.0%] 33 [61.1%] 0.004Palpitations [Y/N] 7 [6.5%] 11 [20.4%] 0.008Sweating [Y/N] 43 [39.8%] 35 [64.8%] 0.003Hearing or vision change [Y/N] 54 [50.0%] 15 [27.8%] 0.007Tiredness [Y/N] 31 [28.7%] 13 [24.1%] 0.53Nausea [Y/N] 50 [46.3%] 33 [61.1%] 0.08Abdominal discomfort [Y/N] 15 [13.9%] 39 [72.2%] 0.001Urge to defecate [Y/N] 2 [1.9%] 27 [50.0%] 0.001Urge to micturate [Y/N] 0 [0%] 6 [11.1%] 0.001Yawning [Y/N] 3 [2.8%] 6 [11.1%] 0.06

All variables refer to lifelong prevalence. Each variable was recorded affirmative if itwas associated with one or more episodes of syncope. Values and percentages referto affirmative answers.

L. Busweiler et al. / Sleep Medicine 11 (2010) 929–933 931

3.3. Specific phobias and other medical/psychiatric conditions

Compared to group C, group SS more often reported a history ofspecific phobia(s): 32.4% versus 74.5% (p = 0.001), in particular BII,19.4% versus 57.4% (p = 0.001) and claustrophobia, 13.9% versus29.6% (p = 0.02). Group SS was more likely to develop syncope inresponse to BII: 24.1% versus 59.3% (p = 0.001) but not in responseto claustrophobic stimuli: 6.5% versus 11.1% (p = 0.36). Althoughcrowd phobia was less common, both groups reported relativelyhigh rates of syncope when standing in crowds (26.9% versus25.9%, p = 0.9). The lifelong prevalence of other medical conditions(including chronic psychiatric conditions) and anti-depressantmedication use were similar in groups C and SS: 24.1% versus14.8% (p = 0.17) and 11.3% versus 7.5% (p = 0.46).

3.4. Syncope symptoms during episodes

Compared to group C, group SS reported the following symptomsmore often: feeling hot, 37% versus 61.1% (p = 0.004); sweating,39.8% versus 64.8% (p = 0.003); palpitations, 6.5% versus 20.4%(p = 0.008); abdominal discomfort, 13.9% versus 72.2% (p = 0.001);urge to defecate, 1.9% versus 50.0% (p = 0.001); and urge to mictu-rate, 0% versus 11.1% (p = 0.001; Table 4). Hearing or vision changewas less likely, 50% versus 27.8% (p = 0.007).

4. Discussion

The main results of the present investigation can be summa-rised as follows: in hospital practice, sleep syncope is not rareand may occur in up to 6% of VVS referrals. Although it presentsmainly in middle-aged women, the attacks typically start in child-hood and the majority occur in patients with BII phobia. Comparedto classical VVS, sleep syncope episodes are more severe and morelikely to occur in the horizontal position. They are characterised bysymptoms of vasodilation (sweating, feeling hot and palpitations)and specifically, extreme vagotonia (abdominal discomfort, urgeto defecate and micturate).

The mean age distribution of our sleep syncope patient groupwas similar to that in other hospital-based studies of VVS [6,7].The predominance of women is consistent with what is reportedfor VVS in the general population though not in all hospital-basedstudies [6,8,9]. The onset of classical VVS is usually in adolescence,whereas we found that sleep syncope is more likely to start in child-hood [7,8]. We suspect this relates to the very strong association wefound between sleep syncope and BII phobia, a well-describedcondition which also usually starts in childhood [10,11].

Studies on psychiatric conditions in the general population,undertaken by lay-interviewers using our criteria, have estimated

Page 4: Sleep syncope: Important clinical associations with phobia and vagotonia

932 L. Busweiler et al. / Sleep Medicine 11 (2010) 929–933

the lifetime prevalence of any specific phobia (including BII, claus-trophobia and crowd phobia) to be 10–20% [12,13]. Even allowingfor age, the predominance of women in our study, and differentinterviewing techniques, our results suggest a fivefold increase inthe lifetime prevalence of specific phobias in patients with sleepsyncope. More specifically, the majority of our patients with sleepsyncope had BII phobia, which represents an order of magnitudemore than the prevalence in the normal population (3–5%) [14].In patients with VVS, there are no data on specific phobias, butthe prevalence we found for BII-syncope in the control groupwas similar to what has been reported in VVS patients and the gen-eral ‘‘young” population [6,8]. BII syncope data are likely to overes-timate the true prevalence of BII phobia because, as demonstratedby our findings, a minority of BII-syncope patients do not have thestrict criteria for phobias [5,11].

Why should patients with vasovagal syncope, in particular sleepsyncope, have such a high prevalence of BII phobia? Part of the rea-son is because BII phobia patients have a clear predisposition toVVS even in the absence of the phobic stimulus [15]. However,the extreme prevalence of BII phobia in sleep syncope patients re-quires further explanation. One possibility would be that theacquisition of a specific phobia during childhood somehow predis-poses the individual to sleep syncope thereafter. Propagation of BIIattacks into adulthood is thought to occur by conditioning but it isdifficult to understand how this could affect sleep [3,16]. Only aminority of sleep syncope patients consistently report frighteningdreams preceding their attacks [1], but whatever the trigger forthe attack, the severe autonomic symptoms we have describedmay condition the individual, in a similar way to BII phobia.

A more likely explanation is that BII phobia and sleep syncopeare both manifestations of the same underlying condition. Thismay explain their similarities (for example, onset in childhood,severity of the attacks, and LOC when lying down) [1,17]. Bothare fascinating examples of the power of non-baroreflex pathwaysover autonomic mechanisms controlling blood pressure and heartrate. The diphasic response seen preceding BII syncope is thoughtto be a ‘‘human example” of a mammalian fear response [18]. Fearresponses have evolved as survival mechanisms and are expres-sions of a mixture of sympathetic (fight and flight) and vagal (con-servation and withdrawal) activity [16]. It may be that anexaggerated fear response is the underlying condition predisposingto both BII and sleep syncope. The autonomic substrate for this re-sponse has been studied in animals, but more recently in conscioushumans using functional MRI [19,20]. The neural pathwaysresponsible for the sympathetic and vagal symptoms descend fromthe cortex to the medulla, whilst feedback pathways for condition-ing ascend to the insula and amygdala via the thalamus [21]. Thereis now evidence that these pathways may be functionally and evenanatomically different in patients with BII phobias and vasovagalsyncope [22,23]. In addition, BII phobia patients have demon-strated longer inhibition of MSNA bursts following arousal stimuli[24]. Furthermore, we recently showed that sleep syncope patientshave attenuated MSNA responses to cold stress, implying thatstimulation of sympathetic activity by central command may beimpaired [2]. Presumably these cortico-medullary pathways maybe even more susceptible to sympatho-vagal imbalance duringsleep when cortico-medullary inhibition is less intense [25,26].

A less-likely explanation for the association between sleep syn-cope and specific phobia is that these patients have a chronic anx-iety disorder which includes both conditions [5,27]. Psychiatristsgenerally do not include specific phobias under generalised anxietydisorder and we found no evidence that any of our patients had anundiagnosed anxiety condition in addition to their specific pho-bia(s) [10]. Agoraphobia is the most common phobic conditionassociated with generalised anxiety and, unlike BII phobia, it isusually acquired in adulthood. Patients may have specific phobias

as part of agoraphobia, for example, claustrophobia with panic at-tacks, but fainting is a rare association [28–30]. We suspect thatpatients with claustrophobia are somehow predisposed to sleepsyncope but the physiological nature of this is uncertain.

4.1. Study limitations

As previously stated this is a retrospective study, totally relianton expert history-taking and patients’ recollection of symptoms,both of which may be selective. When considering phobia epidemi-ology, it is important to realise that specific phobias are graded intomild, moderate, or severe, and that the milder forms are most pre-valent [14]. We did not grade our patients, but suspect that nearlyall would fit into the mild or moderate categories. This needs to beconsidered when comparing lifelong prevalence to other studies,but does not detract from our important finding that BII phobiaof any severity is much more common in patients with sleep syn-cope than classical VVS.

We suspect that sleep syncope is not a rare condition, but is notyet generally accepted or understood by doctors. Our experience isthat many patients are undiagnosed, resulting in inappropriateinvestigation and treatment. Episodes are difficult to capture;therefore the diagnosis is dependent upon an accurate history.The association between BII phobia and sleep syncope impliesthere may be an exaggerated underlying fear response predispos-ing these patients to both conditions.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.sleep.2010.04.013.

References

[1] Krediet CT, Jardine DL, Cortelli P, Visman AG, Wieling W. Vasovagal syncopeinterrupting sleep? Heart 2004;90(5):e25.

[2] Jardine DL, Krediet CTP, Cortelli P, Frampton CM, Wieling W. Sympatho-vagalresponses in patients with sleep and typical vasovagal syncope. Clin Sci2009;117:345–53.

[3] Jardine DL, Krediet CT, Cortelli P, Wieling W. Fainting in your sleep? Clin AutonRes 2006;16(1):76–8. Feb.

[4] Brignole M, Alboni P, Benditt D, Bergfeldt L, Blanc JJ, Bloch Thomsen PE, et al.Guidelines on management [diagnosis and treatment] of syncope. Eur Heart J2001;22:1256–306.

[5] American Psychiatric Association. Diagnostic and statistical manual of mentaldisorders, 4th ed. Washington, DC: American Psychiatric Association; 1994.

[6] Romme JJ, van Dijk N, Boer KR, Dekker LR, Stam J, Reitsma JB, et al. Influence ofage and gender on the occurrence and presentation of reflex syncope. ClinAuton Res 2008;18:127–33.

[7] Sheldon RS, Sheldon AG, Connolly SJ, Morrillo CA, Klingenheben T, Krahan AD,et al. Age of first faint in patients with vasovagal syncope. J CardiovascElectrophysiol 2006;17:49–54.

[8] Ganzeboom KS, Mairuhu G, Reitsma JB, Linzer M, Wieling W, van Dijk N.Lifetime cumulative incidence of syncope in the general population: a study of459 Dutch subjects aged 35–60 years. J Cardiovasc Electrophysiol2006;17:1172–6.

[9] Alboni P, Brignole M, Menozzi C, Raviele A, Del Rosso A, Dinelli M, et al.Diagnostic value of history in patients with syncope with or without heartdisease. J Am Coll Cardiol 2001;37:1921–8.

[10] Marks I. Blood-injury phobia: a review. Am J Psychiatry 1988;145:1207–13.[11] Öst L-G, Sterner U, Lindahl I-L. Physiological responses in blood phobics. Behav

Res Ther 1984;22:109–17.[12] Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Eshleman S, et al.

Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in theUnited States. Arch Gen Psychiatry 1994;31:8–19.

[13] Fredrikson M, Annas P, Fischer H, Wik G. Gender and age differences in theprevalence of specific fears and phobias. Behav Res Ther 1996;34:33–9.

[14] Costello CG. Fears and phobias in women: a community study. J AbnormPsychol 1982;91:280–6.

[15] Accurso V, Winnicki M, Shamsuzzaman ASM, Wenzel A, Johnson AK, SomersVK. Predispostion to vasovagal syncope subjects with blood/injury phobia.Circulation 2001;104:903–7.

[16] Marks IM. Fears phobias and rituals. Oxford University Press; 1987.[17] Page AC. Blood-injury phobia. Clin Psychol Rev 1994;14:443–61.

Page 5: Sleep syncope: Important clinical associations with phobia and vagotonia

L. Busweiler et al. / Sleep Medicine 11 (2010) 929–933 933

[18] Graham DT, Kabler JD, Lunsford L. Vasovagal fainting: a diphasic response.Psychosom Med 1976;23:493–507.

[19] Lofving B. Cardiovascular adjustments induced from the rostral cingulategyrus with special reference to sympatho-inhibitory mechanisms. Acta PhysiolScand 1961;184(Suppl):1–82.

[20] Henderson LA, Keay KA, Bandler R. Caudal midline medulla mediatesbehaviourally-coupled but not baroreceptor-mediated vasodepression.Neuroscience 2000;98:779–92.

[21] Critchley HD, Mathias CJ, Dolan RJ. Fear conditioning in humans: the influenceof awareness and autonomic arousal on functional neuroanatomy. Neuron2002;33:653–63.

[22] Beacher FD, Gray MA, Mathias CJ, Critchley HD. Vulnerability to simple faints ispredicted by regional differences in brain anatomy. Neuroimage2009;47:937–45.

[23] Caseras X, Giampietro V, Lammas A, Brammer M, Vilarroya O, Rovira M, et al.The functional neuroanatomy of blood-injection-injury phobia: a comparisonwith spider phobics and healthy controls. Psychol Med 2009;13:1–10.

[24] Donadio V, Liguori R, Elam M, Karlson T, Montagna P, Cortelli P, et al. Arousalelicits exaggerated inhibition of sympathetic nerve activity in phobic patients.Brain 2007;130:1653–62.

[25] Dickerson LW, Huang AH, Nearing BD, Verrier RL. Primary coronaryvasodilation associated with pauses in heart rhythm during sleep. Am JPhysiol Regul Integr Comp Physiol 1993;33:R186–96.

[26] Somers VK, Dyken ME, Mark AL, Abboud FM. Sympathetic nerve activityduring sleep in normal subjects. N Engl J Med 1993;328:303–7.

[27] Kuakam C, Lacroix D, Klug D, Baux P, Marquie C, Kacet S. Prevalence andsignificance of psychiatric disorders in patients evaluated for recurrentunexplained syncope. Am J Cardiol 2002;89:530–5.

[28] Green SM, Antony MM, McCabe RE, Watling MA. Frequency of fainting,vomiting and incontinence in panic disorder: a descriptive study. Clin PsycholPsychother 2007;14:189–97.

[29] Katon WJ. Panic disorder. N Engl J Med 2006;354:2360–7.[30] Kapoor WN, Fortunato M, Hanusa BH, Schulberg HC. Psychiatric illnesses in

patients with syncope. Am J Med 1995;99:505–12.