You are on page 1of 11

Brain & Development 27 (2005) S43–S53

Original article
Assessment of the maturity-related brainstem functions
reveals the heterogeneous phenotypes and facilitates
clinical management of Rett syndrome
Peter O.O. Julua,b,*, Ingegerd Witt Engerströmb
Peripheral Nerves and Autonomic Unit, Department of Neurology, Central Middlesex Hospital, Park Royal, Acton Lane, London NW10 7NS, UK
Rett Center, Frösö Strand, Östersund Hospital, Sweden
Received 20 August 2004; received in revised form 1 October 2004; accepted 22 February 2005


We have investigated whether brainstem assessment using the NeuroScope could be used for objective and quantitative monitoring of
early development and later progress in Rett syndrome. Brainstem features can be seen on bedside examination of Rett patients and are
included in the International Scoring system. The following cardiovascular vital signs were recorded simultaneously in real-time: cardiac
vagal tone (CVT), cardiac sensitivity to baroreflex (CSB), heart rate (HR), and mean arterial blood pressure (MAP) and respiratory vital
signs: breathing rate and pattern, transcutaneous partial pressures of oxygen (pO2) and carbon dioxide (pCO2). We assessed the occipito-
frontal head circumference (OFHC), height and body mass index (BMI). Results are from 72 patients with classical Rett syndrome
studied at the Swedish National Rett Centre. Three cardiorespiratory phenotypes, designated Forceful, Feeble and Apneustic breathers
were present in similar proportions in the Rett population but early development measured by OFHC and BMI differed. Height was not
affected. Baseline levels of CVT and CSB also differed within the phenotypes indicating differences in parasympathetic activities.
However, parasympathetic activity in the whole population was similar to that previously reported in Rett. Baseline levels of MAP and
HR were similar across the phenotypes, consistent with previous reports of little effect of Rett disorder on baseline sympathetic tone.
Adverse responses to opiate analgesics and hypocapnoeic attacks were unique to specific phenotypes. We recommend early
characterisation of these phenotypes in the management of Rett syndrome. We conclude that classical Rett syndrome consists of
heterogeneous clinical phenotypes with distinct cardiorespiratory states. Brainstem functions can be used to identify these and to monitor
development and treatment, thereby facilitating clinical management.
q 2005 Elsevier B.V. All rights reserved.

Keywords: Rett syndrome; Brainstem; NeuroScope; Parasympathetic nervous system; Sympathetic nervous system; Respiration; Cardiorespiratory state

1. Introduction assessment is indispensable in the management of any

disease. In Rett, we are beginning to recognise important
Rett syndrome is a neurodevelopmental disease probably features of the disease requiring objective clinical evalu-
caused by mutation in the MECP2 gene-‘probably’ because ation in order to manage the disorder. Some of these are
so far only 85% of classic and 50% of atypical cases have presented.
mutations identified [1,2]. Among more than 60 mutations
in the MECP2 gene to date [2], there is no convincing
1.1. Bedside evaluation
evidence relating genotype and phenotype [3,4]. Clinical
Females are overwhelmingly more affected by Rett than
males. Of 72 patients evaluated in the Swedish National Rett
* Corresponding author. Address: Peripheral Nerves and Autonomic
Unit, Department of Neurology, Central Middlesex Hospital, Park Royal, Centre, there was one Rett-like male. The females are born
Acton Lane, London NW10 7NS, UK. Fax: C44 20 8453 2246. apparently normal, but at 1–3 years, they regress with loss of
E-mail address: (P.O.O. Julu). hand use, social withdrawal and reduction in speech.
0387-7604/$ - see front matter q 2005 Elsevier B.V. All rights reserved. Unexplained screaming and sleep disturbance is common.
doi:10.1016/j.braindev.2005.02.012 The girls develop distinctive hand stereotypies, severe
S44 P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53

cognitive impairment, motor impairment especially incoor- 2. Methods

dinate movement and asymmetric alteration in muscle tone
from hypotonia to hypertonia. Many have seizure disorder. 2.1. The NeuroScope’s method of brainstem assessment
Virtually all have breathing abnormality and/or
dysautonomia. A novel neurophysiological technique has been recently
developed for continuous and real-time assessment of
brainstem function in which the vital signs originating
1.2. International scoring system (ISS)
from and regulated by the brainstem are quantified and
recorded synchronously for the study of both temporal and
An International consensus listed 20 key clinical features
causal relationships [6]. These signs can be classified into
(A–T) which may be used to score disease severity [5].
two categories: cardiovascular and respiratory.
These features fall into two categories: effects on general
The cardiovascular vital signs quantified and recorded by
growth and development, and other characteristic
the NeuroScope are: cardiac vagal tone (CVT) believed to
be regulated by nucleus ambiguous in the caudal part of the
Growth and development (A–E) includes OFHC, height,
ventrolateral medulla oblongata [7], Cardiac sensitivity to
weight, and ability to walk (I) [5], contributing a maximum
baroreflex (CSB) believed to be regulated by the
of 12 of the 40 ISS points, or 30% of clinical severity.
commissural part of the nucleus of tractus solitarius in the
The characteristic abnormalities include musculo-skel-
lower part of the dorso-medial medulla oblongata [8],
etal abnormalities (F–H) like altered muscle tone, scoliosis
arterial blood pressure, especially the mean arterial blood
and joint contractures. These contribute a maximum of six
pressure (MAP) known to be regulated in the rostral
ISS points [5]. Movement disorders (J–L) include hand
ventrolateral medulla oblongata [9] and the heart rate (HR)
stereotypy, involuntary movements and lack of functional
which is regulated by both the nucleus ambiguous in the
hand movements—a maximum of six ISS points. Cortical
lower part and the cardioaccelerators in the rostral part of
abnormalities (N–P) include intellectual disability, speech
ventrolateral medulla oblongata [10].
retardation, and seizure disorders contributing a maximum
The respiratory vital signs quantified and recorded by the
of six ISS points. Brainstem and autonomic abnormalities
NeuroScope are: breathing movements, both rate and
(M and Q–T) include oro-motor disturbances, awake
rhythm known to be regulated by the ventrolateral brainstem
breathing dysrhythmia, peripheral circulatory and mood
from as low down as the level of first vertebrate to as rostral
disturbance, agitation and sleep disturbance [5]. These
as the parabrachial nuclei in the pons and the dorsal
contribute a maximum of 10 ISS points (25% of the total).
respiratory group of neurones in the vicinity of the nucleus
of tractus solitarius [11]. Transcutaneous partial pressures of
1.3. The need for objective clinical evaluation oxygen (pO2) and carbon dioxide (pCO2) representing
of Rett syndrome blood gases are also recorded. Blood gases are believed to
be regulated in the medulla oblongata in conjunction with
It is clear that certain features contribute more than inputs from the arterial chemoreceptors [12]. The Neuro-
others to the disease severity and may demand complex Scope’s method therefore quantifies, monitors and records
management systems. The brainstem features become the cardiorespiratory functions of the brainstem in real-time
prominent at regression and remain so throughout life as thereby facilitating the study of both temporal and causal
breathing abnormalities and/or dysautonomia. Effects on relationships of brainstem malfunctions.
growth and development and the brainstem account for
up to 55% in the ISS, but it is the brainstem features 2.2. What is actually measured by the NeuroScope
which frequently require urgent attention. There are no
existing regimes for management of the brainstem The patient sits comfortably reclined with a carer holding
features, yet these are common reasons requiring medical one hand for blood pressure monitoring during at least 1 h of
attention. These brainstem features therefore demand brainstem assessment. All procedures are non-invasive and
objective clinical evaluation to provide a better under- painless. A stretch-sensitive belt (plethysmograph) around
standing and allow development of evidence based the patient’s chest at xiphisternal level records the
management. amplitude of thoracic and abdominal breathing. A TCM3
We have now assessed brainstem function in 216 Rett monitor (Radiometer, Copenhagen, Denmark) quantifies the
patients in the United States, UK, and Sweden. This report partial pressures of oxygen (pO2) and carbon dioxide
presents experience with 72 patients assessed at the Swedish (pCO2) in the blood transcutaneously, transmitting to the
National Rett Centre. The aim is to illustrate the facilitations NeuroScope. A finger cuff (photoplethysmograph)
of clinical management of Rett syndrome achieved through (Finaprese Ohmeda, USA) detects and records arterial
objective assessment of brainstem functions using a novel blood pressure (BP) waveforms, which are quantified and
neurophysiological technique incorporated in the calibrated by the Finaprese and transmitted to the
NeuroScope. NeuroScope to calculate systolic, mean and diastolic BP
P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53 S45

beat-by-beat. The NeuroScopee (MediFit Diagnostics Ltd, 35

London, UK) continuously monitors brainstem regulation of 30
cardioinhibitory activity—cardiac vagal tone (CVT). A

Frequency (%)
non-invasive index defined as ‘pulse-synchronised phase
shifts in consecutive R–R intervals’ is used to quantify
CVT. It is a form of R–R interval variability. Here, the 15

machine calculates the CVT from ECG R–R intervals as 10

previously described [13] and displays it on the screen in 5
arbitrary units of a linear vagal scale (LVS). This scale is 0











fully validated for clinical measurement of CVT [14].
The VaguSoft software (MediFit Diagnostics Ltd,
London, UK) in the NeuroScope system synchronously
records electrocardiogram (ECG), heart rate (HR), CVT, Age groups (years)
systolic, mean and diastolic BP, levels of blood gases and
Fig. 1. Age distribution of the Rett population studied at the Swedish National
calculates the cardiac sensitivity to spontaneous arterial Rett Centre in Frösön. The distribution was skewed towards the younger age
baroreflex (CSB). CSB is defined as the change in pulse group where the modal age-group was between 9.6 and 13.5 years.
interval per unit change in systolic blood pressure calculated
beat-by-beat [15]. It is a measure of the negative effect of 3.2. Cardiorespiratory status of the Rett population
blood pressure on HR and its measurement is made possible
by the synchronous records of both ECG and BP Only one of 72 patients had no breathing abnormality
waveforms. within the 1 h of brainstem function monitoring. Of the 71
The electroencephalogram (EEG) is recorded by a 16- patients who had episodes of both normal and the awake
channel PL-EEG (Nervus, Copenhagen, Denmark). A head- breathing abnormalities, all had multiple-rhythm breathing
cap similar to a swimming cap is used to obtain EEG signals dysrhythmias, each having between 3 and 9 (meanZ5)
from the patient. This is quick and well tolerated by Rett abnormal breathing rhythms, changing in between these
patients. A continuous video record time-locked with the abnormal rhythms at random. A total of 13 abnormal awake
physiological measurements records behaviour. breathing rhythms were confirmed among these 71 patients, as
previously reported [6]. The efficiency of ventilation by these
abnormal breathing rhythms was very variable requiring
2.3. Data analysis and statistics further analysis below, but the overall baseline transcutaneous
pO2 in the population was 76.7G2.1 (range 42–115) mmHg
Baseline functions are defined as cardiorespiratory and pCO2 was 33.9G1.3 (range 13.5–67) mmHg.
activity during normal breathing rhythm without agitation, Baseline brainstem autonomic tones within the whole
normal blood gases and no epileptiform activity on EEG. study population are shown in Fig. 2. The functions of the
Baseline brainstem functions were measured under this caudal part of medulla oblongata are represented by: CVT
standard condition. Data are presented as meansGSEM and (Fig. 2A, range was between 1 and 16.6 units in the LVS) and
analysis of variance in the F-test is used to compare data CSB (Fig. 2A, range was between 0.1 and 18.5 ms/mmHg).
between clinical phenotypes, where the probability values These are all parasympathetic functions of the autonomic
P!0.05 were considered statistically significant. nervous system. The functions of the rostral part of medulla
oblongata are represented by: MAP (Fig. 2B, range was
between 44 and 110 mmHg) and HR (Fig. 2B, range was
between 67 and 131 beats/min). These are largely sympath-
3. Results etic functions of the autonomic nervous system. The HR is
largely a sympathetic index here because of the lack of
3.1. The Rett population studied parasympathetic restraint in Rett syndrome. The cardior-
espiratory statuses varied extensively from nearly normal to
Patients studied were 71 females and one male, aged severely abnormal within the Rett population. Therefore, a
1–45 years (mean 17.3), 67 classical Rett, three Form Fruste further analysis of brainstem functions was carried out.
and two variants. Age distribution is shown in Fig. 1.
Although the maximum body mass index in the population 3.3. Heterogeneous cardiorespiratory phenotypes
was overweight, the average was underweight. Body in the Rett population
mass index (BMI) was (meansGSEM) 17G0.5 kg/m2
(range 9.2–32.7 kg/m2). The recommended normal range Further detailed analysis of 65 patients with awake
for BMI is 19–25 kg/m2. OFHC was 50.5G0.5 cm (range breathing abnormalities revealed three major cardiorespira-
32.7–57 cm). tory phenotypes within the Rett population which we call;
S46 P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53

A 12 3.4. Cardiorespiratory phenotypes affected

clinical presentations
Baseline level

8 3.4.1. Age
Of the pure groups without Valsalva’s manoeuvre, the
6 mean age among Feeble breathers was 18.6 (range 3.9–37.3)
* *
4 years, among Apneustic breathers was 18.6 (range 7.2–39.7)
years and among Forceful breathers was 16.3 (range 1–32)
2 years. There was no significant age difference between these
0 three cardiorespiratory phenotypes (PO0.4). Of those
Rett Normal Rett Normal performing Valsalva’s manoeuvre, mean age among Feeble
CVT CSB breathers was 16.5 (range 10–32.4) years and among
B 120 Forceful breathers was 16.7 (range 6–35.4), again no
significant age difference (PO0.8). Among those who
100 performed continuous Valsalva’s manoeuvre, the mean age
was 23.2 (range 10.4–45), significantly higher than the
Baseline level

average ages among the pure cardiorespiratory phenotypes
60 (P!0.01).
3.4.2. Indices of physical growth
20 OFHC was significantly smaller among the Forceful
breathers measuring only 49.8G1.5 cm compared with
0 51.2G0.5 cm within Apneustic and 51.9G1.8 within
Rett Normal Rett Normal
Feeble breathers (P!0.05). There was no significant
difference in height among the basic cardiorespiratory
Fig. 2. Baseline levels of cardiovascular parasympathetic and sympathetic phenotypes. Heights were 1.35G0.04 m within Apneustic,
activity in the Rett population at the Swedish National Rett Centre in 1.34G0.1 m within Forceful and 1.39G0.07 within Feeble
Frösön compared with published normal values obtained from 11 healthy breathers (PO0.1). However, the Rett girls who performed
control girls aged between 5 and 28 years [6] where *ZP!0.05.
Parasympathetic activity (in A) represented by levels of cardiac vagal
Valsalva’s manoeuvre were a little taller than those who did
tone (CVT) was measured in arbitrary units of an atropine derived linear not, especially among the Feeble breathers—1.47G0.01 m
vagal scale (LVS) [14] and cardiac sensitivity to baroreflex (CSB) was (P!0.001).
measured in ms/mmHg. The sympathetic activity (in B) represented by the
mean arterial blood pressure (MAP) was measured in mmHg and heart rate 3.4.3. Index of nutrition and growth
(HR) in beats/min. The CVT was about 50% and CSB was 40% lower in the
Rett population compared to control (A), which confirms the previously
Of the pure categories without Valsalva’s manoeuvre,
reported significant reduction of parasympathetic activity in Rett patients. Forceful breathers had significantly lower BMI of 16.6G
There was no significant difference in baseline sympathetic activity 1.0 kg/m2 compared to Feeble breathers at 18G2.4 kg/m2
between Rett and control populations (B). and Apneustic breathers at 19.9G1.7 kg/m2 (P!0.03).
Feeble breathers performing Valsalva’s manoeuvre had
Apneustic, Forceful and Feeble breathers. These are significantly lower BMI of 15.7G0.8 kg/m2 compared with
equivalent to Apneustic, Forceful and Inadequate breathing those who did not (P!0.007). BMI of Forceful breathers
styles previously described [16]. In their pure forms performing Valsalva’s manoeuvre was 16.3G0.9 kg/m2,
without Valsalva’s manoeuvre, these phenotypes have not significantly different from the already low BMI in those
approximately equal preponderance. Apneustic breathers who did not (PO0.5). The minority of patients with
form 12% (8/65), Forceful breathers 14% (9/65) and undetermined cardiorespiratory phenotypes who had con-
Feeble breathers 14% (9/65). In this Swedish Rett tinuous episodes of Valsalva’s manoeuvre had normal BMI
population, only the Forceful and Feeble breathers of 20.2G2.9 kg/m2.
performed Valsalva’s manoeuvre in varying degrees and
they were the majority—52% (34/65) of the population 3.4.4. Indices of brainstem functions
studied in detail. 29% (19/65) of Forceful breathers and The effects of cardiorespiratory phenotypes on baseline
23% (15/65) of Feeble breathers performed Valsalva’s brainstem functions are shown in Figs. 3 and 4. The effects
manoeuvre. No Apneustic breather performed Valsalva’s on autonomic cardiovascular regulation are represented in
manoeuvre in this population. A minority 6% (4/65) of the Fig. 3 by CVT and CSB for parasympathetic tone, then HR
Rett population performed Valsalva’s manoeuvre continu- and MAP for sympathetic tone. The effects of cardior-
ously during their episodes of abnormal breathing, thereby espiratory phenotypes on baseline respiratory regulation are
preventing the determination of the basic cardiorespiratory represented in Fig. 4 by the breathing rate during episodes of
phenotypes. normal respiratory chest movements and the effectiveness
P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53 S47

8 120


Heart rate (beats/min)


* 60


0 0

9 120
CSB (ms/mmHg)

6 80

MAP (mmHg)
a * 60
3 40
Apneustic Forceful Forceful+Val Feeble Feeble+Val Valsalva Apneustic Forceful Forceful+Val Feeble Feeble+Val Valsalva

Fig. 3. Baseline levels of cardiovascular parasympathetic and sympathetic activity in the various clinical groups of Rett patients studied at the Swedish National
Rett Centre in Frösön. In A, parasympathetic activity is represented by levels of cardiac vagal tone (CVT) measured in arbitrary units of an atropine derived
linear vagal scale (LVS) [14] and cardiac sensitivity to baroreflex (CSB) measured in ms/mmHg. In B, the level of mean arterial blood pressure (MAP) and the
heart rate represent sympathetic activity (see text for details). Comparing the baseline levels of autonomic tones among the clinical phenotypes: Apneustic,
Forceful and Feeble breathers, *ZP!0.05 and comparing autonomic tones within similar clinical phenotypes with and without Valsalva’s manoeuvre (Val),

of pulmonary ventilation measured in real-time by recovered well. We have not encountered opiate intolerance
transcutaneous pO2 and pCO2. The baseline parasympa- in Forceful or Apneustic breathers.
thetic tone was significantly lower in Feeble breathers
compared with Forceful and Apneustic breathers (P!0.05). 3.5.2. Hypocapnoeic attacks have clinical presentations
Apneustic breathers had the highest level of baseline pCO2 that can be mistaken for epileptic seizures
among the Rett patients (Fig. 4). Valsalva’s manoeuvre Some Forceful breathers performing Valsalva’s
significantly lowered the normal baseline breathing rate and manoeuvre reduced their pCO2 below 20 mmHg (Fig. 5A)
increased pCO2 among Feeble breathers (P!0.01 and P! precipitating jerky limb and trunk movements resembling
0.05, respectively), while it significantly lowered only the clinical seizures, which were associated with the eyeballs
CSB in Forceful breathers (P!0.02). Baseline sympathetic rolling upwards. There was evidence of tetany like carpal
tone was not significantly different among the cardior- spasm and sometimes opisthotonus. These attacks were
espiratory phenotypes (Fig. 3B). reversed by re-breathing in a paper bag to elevate the pCO2
above 20 mmHg when the attacks stopped (Fig. 5B),
confirming their hypocapnoeic nature. We have not
3.5. The cardiorespiratory phenotypes presented unique
observed such attacks in Apneustic breathers or in Feeble
clinical challenges
breathers including those with Valsalva’s manoeuvre.

3.5.1. Opiate intolerance 3.5.3. Brainstem shutdown

Some girls, all of whom were Feeble breathers, were Some Forceful and Feeble breathers who performed
given morphine as part of medications for various surgical intensive Valsalva’s manoeuvre, at the end of prolonged
procedures in different hospitals and failed to recover episodes of continuous manoeuvres, had progressive drops
properly from anaesthesia. All these girls were transferred in CVT and CSB to levels very close to zero, MAP fell
to intensive care units where it became very difficult to wean steadily and HR rose to near 110 beats/min, the natural rate
them off ventilation until on our advice opiates and of the pacemaker in the sino-atrial node without any neural
respiratory depressants were withdrawn, whereupon all regulation (Fig. 6). These are evidence of brainstem
S48 P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53

A 30 objective evaluation of the maturity-related brainstem

functions. It is worthwhile to re-emphasise here that the
25 Rett girls are often passed as normal at birth, abnormalities
a only becoming obvious during early childhood development

when the brainstem features become prominent. We

15 previously evaluated brainstem function in a small sample
of Rett girls and concluded then that the brainstem was
10 immature, basing our conclusion on brainstem cardiore-
spiratory regulation only [17]. This was the first functional
evidence of brain immaturity in Rett syndrome supported by
neuropathological evidence [18,19] and Single Proton
Apneustic Forceful Forceful+Val Feeble Feeble+Val Valsalva Emission Computed Tomography [20] making a strong
argument that the Rett disorder was a developmental disease
B 90
and not a degenerative disease as was widely believed at
80 that time [17]. This study confirms the low cardiovascular
70 parasympathetic tone previously described in Rett [6]. This
60 was reported to be associated with apparently normal
pO2 (mmHg)

sympathetic tone and therefore responsible for the type of
sympatho-vagal imbalance that gives unique autonomic
features described in Rett syndrome, for example;
30 wide pupils, chronic constipation, cold and blue lower
20 extremities [6].
10 We confirm the 13 types of awake breathing dysrhythmia
previously described in a smaller international population
[6]. We have demonstrated that 1-h monitoring of brainstem
function in a clinical setting can establish the cardior-
45 *
espiratory status of the Rett patient. This 1-h monitoring
40 a with the NeuroScope has shown that virtually all Rett girls
35 have multiple-rhythm awake breathing abnormalities, each
pCO2 (mmHg)

30 displaying a unique set of five or so abnormal breathing

25 rhythms, changed at random. Detailed characterisation of
20 the breathing abnormalities in Rett syndrome is available in
15 the literature [6,16].
10 It is clear from the discussion above and from bedside
5 examinations and the ISS point scoring system that
0 brainstem abnormalities are very important in the Rett
Apneustic Forceful Forceful+Val Feeble Feeble+Val Valsalva
disorder and there is a strong need for objective clinical
Fig. 4. Baseline levels of respiratory control in the various clinical groups of evaluation of these brainstem abnormalities if we are to
Rett patients studied at the Swedish National Rett Centre in Frösön. achieve rational clinical management in Rett syndrome.
Breathing rates are shown in A and concentrations of blood gases measured
as partial pressures of oxygen (pO2) and carbon dioxide (pCO2) in B.
4.1. Basic clinical phenotypes in Rett syndrome
Comparing breathing rates and concentrations of blood gases among the
clinical phenotypes: Apneustic, Forceful and Feeble breathers, *ZP!0.05
and comparing these among groups of similar clinical phenotypes with and 4.1.1. Definition of clinical phenotype
without Valsalva’s manoeuvre (Val), aZP!0.05. Phenotype is defined as ‘the entire physical, biochemical,
and physiological make-up of an individual determined both
shutdown as previously described in Rett syndrome [6], environmentally and genetically’ [21]. Detailed analysis of
when brainstem cardiovascular regulation temporarily the cardiorespiratory status in the Rett population here
stops. The longest duration of brainstem shutdown observed showed that there are three basic physiological types in the
was 3 min. We have not observed brainstem shutdown in population, which we have named Apneustic, Forceful and
Rett girls who do not perform Valsalva’s manoeuvre. Feeble breathers. These three basic physiological types have
distinct levels of baseline brainstem autonomic tones unique
to each, so we have called each of these categories a
4. Discussion cardiorespiratory phenotype. Only brainstem cardiorespira-
tory function was considered during the determination of the
The major purpose of this paper is to share our new physiological types, a potential weakness that we acknowl-
clinical information on Rett syndrome, obtained by edge, justified by the fact that all the 13 different types
P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53 S49

350 80
MAP (mmHg)


MAP (mmHg)
200 60

150 50
0 30
180 140
170 130
HR (beats/min)


HR (beats/min)
140 100
6 7
5 6

4 CVT (LVS) 5
3 4

2 3
120 pO2 100
Blood gases (mmHg)

Blood gases (mmHg)

40 pCO2
20 pCO2
0 2138 2198 2258
1700 1760
Time (s) Time (s)

Fig. 5. A 50-s segment in A and a 2-min segment in B of the same continuous record of cardiovascular variables and concentrations of blood gases acquired in
real-time from a Rett girl who had the clinical phenotype characterised by forceful breathing. The segments were recorded during a hypocapnoeic attack in A
and when the patient was re-breathing air from a 5-l paper bag in B. In A, the vertical dotted line marks the beginning of jerky limb movements similar to tonic
clonic seizure. These were associated with rolling up of the eyeballs and large, simultaneous and repetitive spikes in the mean arterial blood pressure (MAP),
heart rate (HR) and cardiac vagal tone (CVT) measured in arbitrary units of an atropine derived linear vagal scale (LVS) [14]. The HR increased beyond the
limit set for the machine repeatedly and the CVT habituated towards lower values with each repetitive spike and there was also severe carpal spasm indicating
tetany. About 6 min later in B, we encouraged the patient to re-breath air in a 5-l paper bag and the partial pressure of carbon dioxide (pCO2) rose steadily above
20 mmHg then the jerky limb movements stopped. The partial pressure of oxygen (pO2) fell steadily below 100 mmHg during this process. The vertical dotted
line in B marks one clear episode of simultaneous increases in both CVT and MAP where the large CVT brought down the HR abruptly. Heart rate spikes
accompanied later episodes of simultaneous spikes in MAP and CVT due to the moderate increases in the CVT allowing increases in HR. These are signs of
abnormal spontaneous brainstem activation, which are very common in Rett patients. Note the moderate levels of MAP, HR and CVT in B during the steady
increase in pCO2 compared to levels in A. Re-breathing air in the bag ameliorated the carpal spasm and abolished the hypocapnoeic attack. Time was measured
from the beginning of the record.

of abnormal breathing rhythms previously described in Rett following hyperventilation. We also now know that
[6,16] are consistent with the suggested three cardiorespira- Valsalva’s manoeuvre is performed across the cardior-
tory phenotypes. It has become clear in our analysis here espiratory phenotypes within the Rett population and not
that Forceful breathers are those that tend to have the restricted to any one phenotype. The effects of Valsalva’s
periodic style of breathing on certain occasions, especially manoeuvre on baseline cardiorespiratory regulation differ
S50 P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53

among these unique cardiorespiratory phenotypes and this

130 will be discussed later.
110 4.1.2. Difference in physical and biochemical make-ups
MAP (mmHg)

Having defined the basic clinical phenotypes physiologi-
cally, we looked for evidence of physical and biochemical
differences among the phenotypes. The indices adopted
80 were: height and the OFHC as measures of physical growth
70 [22] and BMI as an index of the nutritional status [23]
60 loosely linked to the nitrogenous balance and therefore
150 representing both physical and biochemical state. Since
140 these indices depend on the age of the patient it was
130 essential to discover any age difference between the basic
HR (beats/min)

120 cardiorespiratory phenotypes, but there was none. This

110 allowed comparison of the developmental indices among
100 the three cardiorespiratory phenotypes. The concentrations
90 of respiratory gases pO2 and pCO2 in the blood, was taken as
80 an additional biochemical marker for biochemical status.
Of the developmental markers for physical growth that
we studied, only OFHC differed significantly between the
three phenotypes. Heights were similar, suggesting that the
14 post-natal growth of the long bones did not differ. Since
12 OFHC depends on brain size [24], our results suggest that
10 post-natal brain growth differs in the three phenotypes and it

8 is least in the Forceful breathers. However, our observation

6 requires confirmation by direct measurement of the actual
4 brain size post-mortem. Nevertheless, the results confirm
2 that the physical status related to brain growth differs among
0 the Rett phenotypes.
120 BMI was minimal in Forceful breathers, perhaps
because Forceful breathers use more energy due to active
Blood gases (mmHg)

and forcefully expulsion of air. Water is also lost.
Increased energy consumption during forceful breathing
60 is a more plausible explanation of the reduction in BMI in
40 these Rett patients because even the Feeble breathers had
20 pCO2 reduction of body mass index when they were doing
Valsalva’s manoeuvres that require extra energy. It is
3513 3573 3633
therefore evident from our results that the basic index of
Time (s)
nutrition, BMI, is different among the three clinical
phenotypes. This is important in the clinical management
Fig. 6. A 2-min segment of a continuous record of cardiovascular of Rett syndrome because it clearly shows that the energy
variables and partial pressures of blood gases acquired in real-time from requirement is more than average for Forceful breathers
a Rett patient with the clinical phenotype characterised by forceful
and also in Feeble breathers who perform Valsalva’s
breathing. She was agitated with episodes of screaming. The vertical
dotted line marks the start of an episode of brainstem shutdown during manoeuvres. Adequate re-hydration must also be con-
which there were steady drops in both mean arterial blood pressure sidered in these Rett patients.
(MAP) and cardiac vagal tone (CVT) measured in arbitrary units of an It is expected that the cardiorespiratory phenotypes
atropine derived linear vagal scale (LVS) [14]. The heart rate (HR) would affect the concentrations of blood gases. However,
increased steadily towards the rate of the sino-atrial node. About 30 s
we compared the levels of blood gases in the three clinical
later, the partial pressure of oxygen (pO2) in the blood began to drop
steadily because of very shallow breathing during brainstem shutdown. phenotypes during controlled baseline state with patients
However, there was only a very slow and gradual rise in the partial breathing normally. Even at the baseline state, the
pressure of carbon dioxide (pCO2). This episode of brainstem shutdown equilibrium levels of blood gases differed in the three
lasted for about 1 min ending with simultaneous spikes in MAP, HR phenotypes. Apneustic breathers had the highest baseline
and CVT without any clinical intervention. Simultaneous spikes in
levels of pCO2, suggesting that this phenotype accumulates
MAP, HR and CVT, which are signs of abnormal spontaneous
brainstem activations, were very frequent in this patient, particularly carbon dioxide even during normal breathing. This is
preceding brainstem shutdowns. Time was measured from the beginning important in clinical management, especially if availability
of the record. of oxygen is reduced, for example, during chest infection or
P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53 S51

in anaesthesia. A state of asphyxia may be reached faster 4.2. Sustained effects of Valsalva’s manoeuvre
than expected with unpredictable outcome. It is quite the
reverse situation in Forceful breathers who have chronic low We have previously reported the immediate effects of
levels of pCO2 and may tolerate low levels of oxygen for a Valsalva’s manoeuvre on brainstem cardiovascular regu-
much longer period. It cannot be over emphasised that the lation in Rett syndrome in general [28]. In this present
clinical phenotype must be known in order to manage the study, we investigated if there were sequalae of Valsalva’s
respiratory needs of Rett patients appropriately. manoeuvre on baseline brainstem function and found that
there were in deed sequalae and they were different in the
clinical phenotypes. Feeble breathers who performed
4.1.3. Brainstem autonomic regulation in the phenotypes Valsalva’s manoeuvre benefited from it because it
Cardiovascular parasympathetic activity was affected significantly lowered the baseline breathing rate and
most in Feeble breathers, where cardiac vagal tone was allowed the pCO2 to rise to normal values when the
significantly lower than in other Rett phenotypes. Forceful manoeuvre stopped. However, the Forceful breathers who
breathers with the smallest OFHC had baseline cardiac performed Valsalva’s manoeuvre were worse off because
vagal tone higher than the Feeble breathers with larger they had a significant reduction of only the baroreflex
OFHC, suggesting that the low cardiovascular parasympa- regulation of their cardiovascular system and were there-
thetic activity in Rett syndrome is probably not related to the fore at a higher risk of cardiovascular instability even when
overall size of the brain. We have previously suggested that the manoeuvre had stopped. Our results therefore suggest
the low level of cardiovascular parasympathetic activity in that the sequalae of even the commonest ailment in Rett
Rett indicates brainstem immaturity [17]. This supposition syndrome, the Valsalva’s manoeuvre, are different in the
remains valid because the size of the whole brain is not clinical phenotypes with virtually opposite clinical impli-
necessarily an index of brainstem maturity. Neuropatholo- cations in Feeble compared to Forceful breathers.
gical studies confirm brainstem immaturity in Rett [25]. A Valsalva’s manoeuvre is the commonest ailment in Rett
recent study has also shown that the expression of the syndrome because all the clinical phenotypes perform it.
MeCP2 protein is severely deficient in the brainstem of Rett Although in this Swedish group of Rett patients no
patients indicating immaturity of the brainstem as Apneustic breather was observed performing Valsalva’s
expression of MeCP2 protein normally increases as monoeuvre, the majority of 52% of the patients did
neurones mature [26]. The maturity of brainstem in Rett perform it. It is also evident here that the intensity of
indicated by tritiated lysergic acid diethylamide-binding Valsalva’s manoeuvre increased with age so that the
studies [27] is very similar to the maturity age given by the patients who performed the manoeuvre were taller in
level of cardiovascular parasympathetic activity [17]. Our height and those who performed it continuously were
results here show that the baseline cardiovascular para- significantly older than the rest. This is consistent with
sympathetic activity is different in the three phenotypes in previous reports [6]. These older patients appear to tolerate
Rett. It would therefore be useful in future to investigate the extra energy required for the Valsalva’s manoeuvre
better because they had normal body mass indices. It is not
whether the degrees of brainstem maturity are different in
clear from our results at what age the Rett patient would
each phenotype.
begin to tolerate Valsalva’s manoeuvre better.
Sympathetic activity, which is also regulated by the
brainstem, was not affected in the different Rett phenotypes.
4.3. The unique clinical challenges
This selective lack of effect on sympathetic activity would
cause sympatho-parasympathetic imbalance in many target-
The diverse clinical challenges provide the argument for
organs due to the mandatory reciprocal innervation required
early characterisation of the clinical phenotypes in the
for normal function. The sympatho-parasympathetic imbal- management of Rett syndrome; correct management
ance that ensues gives the characteristic autonomic picture depending on the clinical phenotype. As reported here,
of wide pupils and cold, blue extremities. It is clear that Rett even the responses to common medications such as opiate
syndrome affects largely the integrative inhibitions in analgesics differ within the phenotypes. We suggested
brainstem functions—modulation of awake breathing previously that the serotonin agonist Buspirone may be
rhythms, baroreflex regulation of the cardiovascular system useful in apneustic breathing [29]. A further pilot study has
and cardiac vagal tone—all of which rely on neuronal shown that this is only useful in the Apneustic phenotype
networks of complex and multiple inhibitions for their (personal communication). We have had some success
proper execution. The brainstem functions that are treating feeble breathing with respiratory stimulant drugs
controlled by simple pacemaker activity like sympathetic but such treatment would be inappropriate for Forceful
tone are not much affected in Rett. This suggests that the breathers although they may have prolonged episodes of
target cells affected by mutations in the MECP2 gene central apnoeas because their apnoeas are due to
include integrative inhibitory brainstem neurones or their hypocapnoea, while Feeble breathers have neuronal
neurotransmitters or both. deficiency in maintaining the amplitude of breathing
S52 P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53

movements. These examples illustrate clearly the clinical References

and physiological bases for our view that it is inappropriate
to regard a mixed population of patients with classical Rett [1] Dragich J, Houwink-Manville I, Schanen C. Rett syndrome: a
surprising result of mutation in MECP2. Hum Mol Genet 2000;9:
syndrome as homogeneous either for research purposes or
for clinical management. [2] Shahbazian MD, Zoghbi HY. Molecular genetics of Rett syndrome
The hypocapnoeic attack is unique to Forceful breathers and clinical spectrum of MECP2 mutations. Curr Opin Neurol 2001;
and may mimic epileptic seizure. Brainstem shutdown 14:171–6.
seems unique to Rett patients with Valsalva’s manoeuvres, [3] Ravn K, Nielsen JB, Uldall P, Hansen FJ, Schwartz M. No correlation
between phenotype and genotype in boys with a truncating MECP2
the apparent precipitating factors being intense agitation
mutation. J Med Genet 2003;40:e5.
with equally intense Valsalva’ manoeuvres. Brainstem [4] Christodoulou J, Weaving LS. MECP2 and beyond: phenotype–
shutdown worries clinicians and carers because the patient genotype correlations in Rett syndrome. J Child Neurol 2003;18:
looks pale and lifeless, blood pressure may be unmeasurable 669–74.
and breathing movements very shallow and easily missed. [5] Kerr AM, Nomura Y, Armstrong D, Anvret M, Belichenko PV,
Budden S, et al. Guidelines for reporting clinical features in cases with
The episode may last for two or more minutes. MECP2 mutations. Brain Dev 2001;23(Suppl. 1):S208–S11.
The real importance of this discussion is to draw the [6] Julu PO, Kerr AM, Apartopoulos F, Al Rawas S, Engerstrom IW,
attention of health care workers dealing with Rett patients Engerstrom L, et al. Characterisation of breathing and associated
to such events. We cannot recommend special treatment central autonomic dysfunction in the Rett disorder. Arch Dis Child
for brainstem shutdown. Shorter episodes may require no 2001;85:29–37.
[7] Guyenet PG, Koshiya N, Huangfu D, Baraban SC, Stornetta RL,
clinical action, but longer episodes may require treatment Li YW. Role of medulla oblongata in generation of sympathetic and
for shock due to the very low blood pressure during the vagal outflows. Prog Brain Res 1996;107:127–44.
attack. Recovery is just as sudden as the start of the [8] Kasparov S, Paton JF. Differential effects of angiotensin II in the
attack. nucleus tractus solitarii of the rat—plausible neuronal mechanism.
J Physiol 1999;521(Pt 1):227–38.
[9] Guyenet PG. Role of the ventral medulla oblongata in blood pressure
regulation. In: Loewy AD, Spyer KM, editors. Central regulation of
autonomic functions. Oxford: Oxford University Press; 1990. p. 145–67.
5. Conclusion [10] Jordan D. Central nervous integration of cardiovascular regulation. In:
Jordan D, Marshall J, editors. Cardiovascular regulation. London:
We have identified three cardiorespiratory phenotypes in Portland Press; 1995. p. 1–14.
[11] Blessing WW. Breathing. In: Blessing WW, editor. The lower
Rett—Feeble, Forceful and Apneustic, which affect early brainstem and bodily homeostasis. Oxford: Oxford University Press;
childhood development differently as evidenced by OFHC, 1997. p. 101–64.
BMI and cardiovascular parasympathetic activity. Height [12] Spyer KM, Dale N, Gourine AV. ATP is a key mediator of central and
did not vary with clinical phenotype. The results confirm the peripheral chemosensory transduction. Exp Physiol 2004;89:53–9.
[13] Little CJ, Julu PO, Hansen S, Reid SW. Real-time measurement of
clinical importance of the brainstem features in Rett
cardiac vagal tone in conscious dogs. Am J Physiol 1999;276:
syndrome in deciding the appropriate management. We H758–H65.
therefore recommend early characterisation of the clinical [14] Julu POO. A linear scale for measuring vagal tone in man. J Auton
phenotypes in the management of Rett syndrome. The Pharmacol 1992;12:109–15.
brainstem features required for the characterisation of [15] Julu POO, Hansen S, Barnes A, Jamal GA. Continuous measurement
of the cardiac component of arterial baroreflex (ccbr) in real-time
clinical phenotypes in Rett syndrome can be measured
during isometric-exercise in human volunteers. J Physiol—London
and monitored in a routine clinical setting using the 1996;497P:P7–P8.
NeuroScope. [16] Julu POO. The central autonomic disturbance in Rett syndrome. In:
Kerr A, Witt-Engerstrom I, editors. Rett disorder and the developing
brain. Oxford: Oxford University Press; 2001. p. 131–81.
[17] Julu PO, Kerr AM, Hansen S, Apartopoulos F, Jamal GA. Functional
evidence of brain stem immaturity in Rett syndrome. Eur Child
Acknowledgements Adolesc Psychiatry 1997;5(Suppl. 1):47–54.
[18] Armstrong D. Recent developments in neuropathology—electron
This was the result of teamwork, although we have microscopy—brain pathology. Eur Child Adolesc Psychiatry 1997;
written it here in only two names. We are particularly 6(Suppl. 1):69–70.
[19] Armstrong D, Dunn JK, Antalffy B, Trivedi R. Selective dendritic
indebted to our colleagues and members of my team Stig alterations in the cortex of Rett syndrome. J Neuropathol Exp Neurol
Hansen and Flora Apartopoulos. We acknowledge with 1995;54:195–201.
thanks the grant from Östersund Hospital Foundation for [20] Nielsen JB, Friberg L, Lou H, Lassen NA, Sam ILK. Immature pattern
Medical Research and the valuable support from Dr Bengt of brain activity in Rett syndrome. Arch Neurol 1990;47:982–6.
Engerström. We acknowledge the supports from the [21] Anon. Dorland’s illustrated medical dictionary. Philadelphia, PA:
W.B. Saunders Company; 1994.
Department of Physiology in Sundsvall Hospital, the [22] Engerstrom IW. Rett syndrome in Sweden. Neurodevelopment—
Department of Paediatric in Ostersund Hospital and disability—pathophysiology. Acta Paediatr Scand Suppl 1990;369:
the collaboration of the families and young volunteers. 1–60.
P.O.O. Julu, I. Witt Engerström / Brain & Development 27 (2005) S43–S53 S53

[23] Motil KJ, Schultz R, Brown B, Glaze DG, Percy AK. Altered energy [27] Armstrong DD, Kinney HC. The neuropathology of the Rett disorder.
balance may account for growth failure in Rett syndrome. J Child In: Kerr AM, Witt-Engerstrom I, editors. Rett disorder and the
Neurol 1994;9:315–9. developing brain. Oxford: Oxford University Press; 2001. p. 57–84.
[24] Armstrong DD, Dunn JK, Schultz RJ, Herbert DA, Glaze DG, [28] Engerstrom L, Witt-Engerstrom I, Julu POO, Apartopoulos F,
Motil KJ. Organ growth in Rett syndrome: a postmortem examination Hansen S. Cardiovascular response to Valsalva’s manoeuvre (VM)
analysis. Pediatr Neurol 1999;20:125–9. in conscious girls with Rett syndrome (RS). J Physiol 2001;536:
[25] Armstrong DD. The neuropathology of Rett syndrome—overview 149P–150.
1994. Neuropediatrics 1995;26:100–4. [29] Kerr A, Julu P, Hansen S, Apartopoulos F. Serotonin and breathing
[26] Armstrong DD, Deguchi K, Antallfy B. Survey of MeCP2 in the Rett dysrhythmia in Rett syndrome. In: Perat MV, editor. New
syndrome and the non-Rett syndrome brain. J Child Neurol 2003;18: development in child neurology. Bologna: Monduzzi Editore; 1998.
683–7. p. 191–5.