Medical Hypotheses (2004) 63, 980–985

http://intl.elsevierhealth.com/journals/mehy

The role of altered impedance in the

pathophysiology of normal pressure
hydrocephalus, Alzheimer’s disease
and syringomyelia
Grant A. Bateman*

Department of Medical Imaging, John Hunter Hospital, Locked Bag 1, Newcastle Region Mail Center,
Newcastle NSW2310, Australia
University of Newcastle, Newcastle, Australia

Received 13 April 2004; accepted 20 April 2004

Summary Normal pressure hydrocephalus, Alzheimer’s disease and syringomyelia appear to be completely unrelated
diseases, however, they share a reduction in subarachnoid space compliance as part of their pathophysiology. This
paper discusses the physiology of pulsatile fluid flow and its relationship to compliance/impedance. Unlike continuous
or non-pulsatile flow where the vessel resistance and pressure gradient are the major determinants of the volume of
fluid flowing, when the fluid flow in a vessel pulsates then the vessel compliance/impedance becomes important. A
reduction in compliance in the craniospinal cavity in each of the three diseases discussed, leads to a limitation of the
outflow vessel compliance. Therefore, there is an increase in outflow vessel impedance. The venous blood, CSF and
interstitial brain/spinal cord fluid all have significantly pulsatile flow and an increase in the impedance of the fluid
outflow in each disease would limit the volume of these fluids as they attempt to cross the subarachnoid space. It is
hypothesised that a reduction in the efficiency of the outflow of venous blood, CSF and interstitial brain/spinal cord
fluid would lead to the accumulation of CSF in NPH, cord fluid in syringomyelia and delay the excretion of beta amyloid
via the interstitial drainage pathways in AD.

c 2004 Elsevier Ltd. All rights reserved.

Introduction cortical venous return [1]. The effect of this is a

The dementia of normal pressure hydrocephalus
(NPH) has been shown to be related to larger than
normal pulse pressure waves interacting within a
craniospinal cavity of significantly reduced com-
pliance, producing compression of the superficial
*
Tel.: +61-2-49213414; fax: +61-2-49213428.
E-mail address: grant.bateman@hunter.health.nsw.gov.au.
decrease in vein compliance or, conversely, an in-
crease in venous impedance. Traditionally, only
mean fluid pressure and resistance have been
considered in the physiology of blood flow and CSF
resorption in the brain but both blood and CSF flow
are pulsatile in vivo. Therefore, pulse pressure,
pulsatile flow and impedance may be important in
both the physiology of fluid flow in the brain and
pathophysiology of disease. There is poor resorp-

0306-9877/$ - see front matter c 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mehy.2004.04.019
The role of altered impedance
tion of CSF in NPH [2], reduced transport of Beta
amyloid (b amyloid) in Alzheimer’s disease [3] and
accumulation of CSF in the cord in syringomyelia
[4]. A recent paper discussed the possible associ-
ation between compliance and dementia [5]. This
paper builds on this work and discusses the possible
role of impedance changes in the craniospinal
cavity and the expected changes that would occur
in the flow of fluids in and around the neuraxis.
Vascular impedance
Resistance and its close relation impedance both
measure the physical properties of vessels that
tend to inhibit the flow of fluid. In the vascular tree
total flow can be shown to be the sum of two
components: non-pulsatile flow and pulsatile flow
[6]. This is illustrated in Fig. 1 where a normal
blood flow wave form (Fig. 1(a)) can be broken
down into non-pulsatile flow (all the blood volume
which is represented below the diastolic minimum,
Fig. 1(b)) and a pulsatile flow component, repre-
senting the volume of flowing blood between the
diastolic minimum and systolic maximum
(Fig. 1(c)). The reason that this is important is that
it can be shown that a change in vascular resistance
alters non-pulsatile blood flow but does not change
the pulsatile component of flow [6]. Similarly, an
increase in impedance affects pulsatile flow but
not non-pulsatile flow [6]. Impedance for the sake
of simplicity is often lumped together with “total
resistance”, however; this maybe a grave error
in the brain. From Poiseuille’s law, the resistance
to flow depends only on the viscosity of the fluid,
the length of the vessel (both are usually constants
in vivo) and the fourth power of the radius of the
vessel, whilst the impedance also depends funda-
Figure 1 (a) A normal arterial wave form. (b) The non-pulsatile flow from 1(a) representing all flow below the di-
astolic minimum. (c) The pulsatile flow from 1(a) representing the flow between the diastolic minimum and the systolic
maximum.
981
mentally on the compliance or the viscoelastic
properties of the vessel wall that allow it to expand
and contract [7]. A change in wall compliance will
not affect the mean cross-sectional area of the
vessel, therefore, it will not affect the resistance
but can still significantly affect total flow [7]. Fig. 2
schematically illustrates this point. The arterial
and venous blood flow pulsate in the brain as does
the CSF and an understanding of impedance to flow
will shed light on many of the apparent paradoxes
associated with brain hydrodynamics to be dis-
cussed.
Measuring impedance in the craniospinal
cavity
To accurately measure impedance in the cranio-
spinal cavity and brain would require the simulta-
neous measurement of the pulse pressure and flow
pulsatility. Flow pulsatility can be measured non-
invasively using MR or Doppler ultrasound but pulse
pressure requires invasive catheterisation. Relative
impedance could be measured provided a vessel
could be found which changed its pulsatility ap-
propriately with changes in craniospinal compli-
ance and an internal reference existed which
reflected only the input pulse pressure and was
relatively resistant to changes in compliance. It is
known that when an increase in the intracranial
pressure occurs there is an increase in the imped-
ance on the venous side of the capillary bed [8].
Thus, the cortical veins, which are surrounded by
the subarachnoid space and are thin walled, are
most likely to represent an accurate manometer of
craniospinal compliance [1]. Shunting patients with
(NPH) represents a unique opportunity to study
changes in compliance and therefore impedance. It
982 Bateman

Figure 2 (a) The same pulsatile total flow previously seen in (Fig. 1). (b) The change in 2(a) following a selective
increase in resistance but with impedance held constant, note the non-pulsatile flow is reduced but the pulsatile flow is
unchanged. The mean or average (incorporating pulsatile and non-pulsatile) blood flow decreases from 5.3 to 3.3 ml/s
and if a Doppler ultrasound were preformed the resistive index (peak systolic flow minus the diastolic flow divided by
the systolic flow) would have increased from 0.53 to 0.73 correctly indicating an increase in resistance. (c) The change
in 2(a) following a selective increase in impedance but with resistance held constant, note the pulsatile flow is reduced
but the non-pulsatile flow is not. The mean blood flow decreases from 5.3 to 4.1 ml/s, increased impedance therefore
decreases total flow, however paradoxically the resistive index decreases from 0.53 to 0.28. Thus, a Doppler exami-
nation would misinterpret an elevation in impedance as a reduction in resistance.

is known that there is a higher than normal CSF
pulse pressure in NPH [9] and that the CSF pulse
pressure correlates inversely with craniospinal
compliance [10]. Therefore, NPH represents a very
low compliance and high impedance state. Pro-
viding an alternative route for the dissipation of
the CSF pulse pressure wave and reducing the mean
CSF pressure by using CSF diversion would be pre-
dicted to increase the compliance and lower im-
pedance. A comparison of the vascular pulsatility
before and after ventriculo-peritoneal shunting in
NPH supports this concept [11]. In NPH, cortical
veins have been shown to have a low mean pulsa-
tility (0.28), whilst the internal reference from the
deep venous system (the straight sinus) shows a
high pulsatility (0.68) [11]. Following ventriculo-
peritoneal shunting cortical vein pulsatility in-
creased (0.80), whilst the straight sinus pulsatility
remained unchanged (0.68) [11]. An example of the
change in cortical venous waveform following
shunting in an actual patient with NPH is shown in
Fig. 3. The difference between the pre-shunt wave
form (Fig. 3(a)) and the post shunt wave form
(Fig. 3(b)) is very similar to the changes between
Figs. 2(a) and (c) , only in reverse, i.e. the shunt
tube has not changed the non-pulsatile flow to any
great degree but the pulsatile flow has increased
greatly. Thus, the change is purely one of de-

Figure 3 (a) Measurement of the flow in a cortical vein in a patient with NPH. (b) Measurement of the flow in the
same cortical vein as in 3(a) five days latter following shunt insertion.
The role of altered impedance
creased impedance. Note that this is associated
with a significant increase in mean (or average)
flow for this vein and also that shunting pa-
tients with NPH increases sagittal sinus flow by
28% [12]. Therefore, an increase in vascular im-
pedance by limiting the ability of the cortical veins
to allow pulse waves to pass, also limits total blood
flow.
Impedance and blood flow in dementia
A large pulsation in the venous outflow is inherently
inefficient. It can be shown that blood flow is op-
timised i.e. the greatest flow for the lowest pres-
sure gradient across the vascular tree occurs when
the CSF impedance is at its lowest and the cranio-
spinal compliance its highest [13]. This situation
maximises the dampening of pulsations. Egnor
et al. [13] have noted, using modelling, that in-
creasing the venous pulsations in the venous sinuses
should give an increase in the sinus pressure purely
due to the impedance effect with the increased
pressure required to maintain blood flow until auto
regulation fails. Czosnyka et al. [14] found that
with aging there is an “increase in the elastic co-
efficient of the brain and a reduction in the cere-
brospinal compensation reserve indicating the
brain becomes stiffer and the CSF space less com-
pliant with age”. Given that with ageing there is an
increase in venous pulsations, a prediction of an
increase in the sagittal sinus pressure with aging
would be assumed purely due to the impedance
effect. An elevation in sinus pressure has been di-
rectly measured in normal ageing [15]. The data in
NPH suggest increased venous sinus pulsation over
and above the ageing effect discussed [12], so
there should be an elevation in the sagittal sinus
pressure in NPH. In clinical and experimental hy-
drocephalus the sagittal venous pressures have
been shown to rise [16]. If one were to ignore the
impedance effect, then a prediction of a reduction
in venous sinus pressure would be expected in NPH.
This is because the resistance of the sinuses is
relatively fixed and the sagittal sinus mean blood
flow in NPH is reduced by one third [12]. The
pressure gradient in a vessel equals the resistance
times the flow. A one third reduction in sinus flow in
NPH would be expected to give a low sinus pres-
sure, not the increased pressure actually found. As
impedance and pulsations increase there should be
a stepwise reduction in blood flow efficiency
through the entire vascular tree from normal age-
ing to NPH. Ischemia and metabolic derangement is
a major component of NPH, thus implicating im-
pedance in its pathophysiology.
983
Impedance and CSF flow in NPH
A reduction in the flow of CSF out of the cranio-
spinal cavity in NPH has been implicated in the
pathophysiology of this disease [2]. CSF is resorbed
across the arachnoid granulations and this resorp-
tion is dependant on the “resistance to flow”
across the granulations and the sagittal sinus
pressure. It has already been noted that the sinus
pressure rises in NPH. From the figures published by
Portnoy et al. [17] using dogs with a naturally oc-
curring form of hydrocephalus as a model, there
was an elevated CSF pressure of 5.31 mm Hg in the
hydrocephalic compared to the normal animals. Of
this rise, 2.63 mm Hg (45%) was due to a rise in the
pressure gradient across the arachnoid granulations
and 3.28 mm Hg (55%) was due to the elevation in
sagittal sinus pressure. Thus, the majority of the
CSF circulatory failure in NPH may be due to venous
pressure and not the arachnoid granulations per se.
The apparent change in the “resistance” to CSF
reabsorption across the arachnoid granulations has
remained an enigma. The resistance to CSF out
flow increases by 5 mm Hg/ml/min in a healthy 80
yr old compared with young patients, adding to the
delay in CSF flow noted over the convexities in
aging [18]. This suggests there may be a correlation
between the age related increases in impedance
and CSF outflow “resistance”. It has been found
that the effect of the skull and dura on the CSF
hydrodynamics is such that the “resistance” to CSF
outflow after craniectomy decreases two fold and
that brain compliance (the pressure volume index)
increases. Therefore, it has been said that the
“resistance” to CSF outflow multiplied by the
compliance of the cerebrospinal space tends to
remain constant [19]. This last statement is by
definition a contradiction in terms. If craniospinal
compliance is the greatest determinant of the CSF
out flow then the arachnoid granulations should
have high impedance and not high resistance. It
would be expected that they should consist of rigid
walled fluid channels and that any reduction in
craniospinal compliance by increasing the pulse
pressure of the CSF outflow should reduce the ef-
ficiency of this drainage and therefore increase the
apparent granulation “resistance”. The ultrastruc-
ture of the arachnoid granulation consists of bun-
dles of thick collagen defining spaces, which form
channels in direct communication with the
periphery of the granulation [20] i.e. rigid tubes
defining the CSF flow pathways. Thus, changes in
compliance by affecting the CSF pulse pressure can
directly alter the efficiency of CSF resorption in
NPH.
984
Impedance and interstitial fluid flow in
Alzheimer’s disease
The third, and perhaps the most important fluid
leaving the brain is the interstitial brain fluid. Ten
percent of CSF is made up of interstitial brain
fluid with the remainder draining via the perivas-
cular spaces [21]. Interstitial cerebral fluid
drainage occurs as a consequence of the pulsation
of the arteries with regular dilatation and recoil
driving the interstitial fluid and proteins in the
opposite direction to the flow of blood. Rennels
et al. [22] have found that the periarterial and per-
iarteriolar spaces were rapidly penetrated with
horseradish peroxidase injected into the sub-
arachnoid space, in preference to the perivenular
spaces and that the passage of the tracer was
dependent on the presence of arterial pulsations.
Most theories of the progression of dementia im-
plicate beta amyloid (AbÞ deposition as the likely
cause. Weller et al. [3] have suggested that the
defect lies in the transport of Ab out of the brain.
The beta amyloid appears to accumulate in the
perivascular interstitial fluid drainage pathways of
the brain [3]. The drainage pathway for intersti-
tial fluid extends along the leptomeningeal arter-
ies bounded by the very thin and compliant pia
matter and their passage through the subarach-
noid space means that they will come under the
same craniospinal compliance conditions as the
cortical veins experience. The mean flow rates in
the perivascular spaces are undoubtedly low and
because of the close proximity to the arterial
walls the pulsatile components would be the ma-
jor contributor to mean flow. As the arterial pul-
sations with aging are larger than normal, the flow
in these spaces would be almost all pulsation and
therefore exquisitely dependant on impedance but
not resistance. From the findings in the cortical
veins in NPH it would be expected that high cra-
niospinal impedance, as occurs in aging, should
lead to a significant reduction in perivascular
systolic fluid flow causing a delay in Ab excretion
[5]. The maximal Ab accumulation would be ex-
pected to be at the site of maximal impedance
i.e. in the walls of the small leptomeningeal ar-
teries as they pass through the subarachnoid
space.
Impedance and syringomyelia secondary to
Chiari I malformation
In syringomyelia secondary to Chiari I malforma-
tions, Oldfield et al. [23] found there was a rapid
Bateman
inferior motion of the cerebellar tonsils which
pluged the subarachnoid space posteriorly and
caused a partial isolation of the spinal subarach-
noid space. There was also minimal movement of
CSF across the foramen magnum during systole or
diastole [23]. Heiss et al. [24] found that, although
there was an increase in the CSF velocity at the
foramen magnum, there was a reduction in flow
volume across this space. These findings caused
the spinal CSF compliance to be reduced with an
elevation in cervical CSF pressure and pulse pres-
sure [24]. The findings of a reduction in compli-
ance and an elevation in CSF pulse pressure are
reminiscent of the changes previously discussed
for NPH. Could development of a syrinx cavity be
caused by an alteration in the venous and inter-
stitial fluid outflow impedance analogous to NPH?
In a Kaolin-induced model of syringomyelia, mi-
croangiograms showed venous engorgement [25]
and laser Doppler flowmetry has shown a signifi-
cant increase in regional spinal cord blood flow
after decompression of a syrinx [26]. Both of these
findings suggest a selective venous outflow ob-
struction. A cat model of subarachnoid space
scarring caused significant dilatation of the peri-
vascular spaces in the central grey matter and the
central white matter in the lower cervical cord,
consistent with interstitial oedema [27]. This ap-
pears similar to the presyrinx state; this is a re-
versible increase in interstitial water content of
the cord [28]. Both of these latter findings suggest
obstruction of the interstitial fluid outflow of the
cord. As discussed above, the optimum drainage of
blood and interstitial fluid from the cord would
require optimum compliance of the walls of the
draining vessels because the flow in these tubes is
likely to be pulsitile. Partial isolation of and/or
obstruction to the flow of CSF in the thecal sac
decreases the compliance of the canal and of the
walls of the vessels, which pass through it. In a dog
model of hydrocephalus, the highest pressures
developed within the cortical veins passing
through the subarachnoid space [17] and in a cat
model of arachnoid scarring, the intramedullary
pressure was consistently higher than the adjacent
subarachnoid space pressure. Therefore, the
pathophysiology of syringomyelia appears analo-
gous to NPH. The primary abnormality in syringo-
myelia must be a block to the fluid drainage
pathways at the level of the subarachnoid space. If
the cat model, as discussed above, is correct and
medullary pressure is higher than subarachnoid
pressure in syringomyelia, then an increase in
fluid inflow from the subarachnoid space into the
cord due to subarachnoid pressure alone is not
sustainable.
The role of altered impedance
Conclusion
There is limited dampening in the craniospinal
cavity and because of this, there is compression of
the structures giving pulsatile fluid outflow. A re-
duction in compliance and elevation in impedance
of the fluid outflow pathways of the brain and
spinal cord means that there will be increasingly
reduced efficiency in fluid transfer through the
veins, arachnoid granulations and perivascular
spaces. The consequence of this are elevated fluid
pressures in the brain or spinal cord depending on
the site of the compliance reduction.
Acknowledgements
Australian Brain Foundation and John Hunter Hos-
pital Research Committee for granting funding for
research incorporated in this paper.
References
[1] Bateman GA. Pulse-wave encephalopathy: a comparative
study of the hydrodynamics of leukoaraiosis and normal
pressure hydrocephalus. Neuroradiology 2002;44:740–8.
[2] Borgesen SE, Gerris F. The predictive value of conductance
to outflow of CSF in normal pressure hydrocephalus. Brain
1982;105:65–86.
[3] Weller RO, Massey A, Kuo Y, Roher AE. Cerebral amyloid
angiopathy: accumulation of Ab in interstitial fluid drainage
pathways in Alzheimer’s disease. Ann NY Acad Sci 2000;903:
110–7.
[4] Ball BJ, Dayan AD. Pathogenesis of syringomyelia. Lancet
1972;2:799–801.
[5] Bateman GA. Pulse wave encephalopathy: a spectrum
hypothesis incorporating Alzheimer’s disease, vascular
dementia and normal pressure hydrocephalus. Med Hypoth-
eses 2004;62:182–7.
[6] Saunders HM, Burns PN, Needleman L, et al. Hemodynamic
factors affecting uterine artery Doppler waveform pulsitil-
ity in sheep. J Ultrasound Med 1998;17:357–68.
[7] West JB, editor. Best and Taylor’s physiological basis of
medical practice. 12th ed. Baltimore: Williams and Wilkins;
1991.
[8] Takizawa H, Gabra-Sanders T, Miller JD. Changes in the
cerebrospinal fluid pulse wave spectrum associated with
raised intracranial pressure. Neurosurgery 1987;20:355–61.
[9] Di Rocco C, Maira G, Rossi GF, Vignati A. Cerebrospinal fluid
pressure studies in normal pressure hydrocephalus and
cerebral atrophy. Eur Neurol 1976;14:119–28.
[10] Barcena A, Mestre C, Canizal JM, Rivero B, Lobato RD.
Idiopathic normal pressure hydrocephalus: analysis of
factors related to cerebrospinal fluid dynamics determining
985
functional prognosis. Acta Neurochir (Wein) 1997;139:
933–41.
[11] Bateman GA. The reversibility of reduced cortical vein
compliance in normal-pressure hydrocephalus following
shunt insertion. Neuroradiology 2003;45:65–70.
[12] Bateman GA. Vascular compliance in normal pressure
hydrocephalus. AJNR 2000;21:1586–90.
[13] Egnor M, Zheng L, Rosiello A, Gutman F, Davis R. A model of
pulsations in communicating hydrocephalus. Pediatr Neu-
rosurg 2002;36:281–303.
[14] Czosnyka M, Czosnyka ZH, Whitfield PC, Donovan T, Picard
JD. Age dependence of cerebrospinal pressure–volume
compensation in patients with hydrocephalus. J Neurosurg
2001;94:482–6.
[15] Rubenstein E. Relationship of senescence of cerebrospinal
fluid circulatory system to dementias of the aged. Lancet
1998;351:283–5.
[16] Shulman K, Yarnell P, Ransohoff J. Dural sinus pressure in
normal and hydrocephalic dogs. Arch Neurol 1964;10:
575–80.
[17] Portnoy HD, Branch C, Castro ME. The relationship of
intracranial venous pressure to hydrocephalus. Child’s Nerv
Syst 1994;10:29–35.
[18] Ableck MJ, Skak C, Neilsen PR, Olsen KS, Borgesen SE,
Gjerris F. Age dependence to cerebrospinal fluid outflow. J
Neurosurg 1998;89:275–8.
[19] Shapiro K, Fried A, Takei F, Kohn I. Effect of the skull and
dura on neural axis pressure–volume relationships and CSF
hydrodynamics. J Neurosurg 1985;63:76–81.
[20] Conegero CI, Chopard RP. Tridimensional architecture of
the collagen element in the arachnoid granulations in
humans. Arq Neuropsiquiatr 2003;61:561–5.
[21] Spector R, Johanson CE. The mammalian choroid plexus.
Sci Am 1989;261:68–74.
[22] Rennels ML, Gregory TF, Blamanis OR, Fujimoto K, Grady
PA. Evidence for a ‘paravascular’ fluid circulation in the
mammalian central nervous system provided by the rapid
distribution of tracer protein throughout the brain from the
subarachnoid space. Brain Res 1985;326:47–63.
[23] Oldfield EH, Muraszko K, Shawker TH, Patronas NJ. Path-
ophysiology of syringomyelia associated with Chiari I
malformation of the cerebellar tonsils. J Neurosurg 1994;
80:3–15.
[24] Heiss JD, Patronas N, De Vroom HL, et al. Elucidating the
pathophysiology of syringomyelia. J Neurosurg 1999;91:
553–62.
[25] Yamada H, Yokota A, Haratake J, Horie A. Morphological
study of experimental syringomyelia with kaolin-induced
hydrocephalus in a canine model. J Neurosurg 1996;84:
999–1005.
[26] Young WF, Tuma R, O’Grady T. Intraoperative measure-
ment of spinal cord blood flow in syringomyelia. Clin Neurol
Neurosurg 2000;102:119–23.
[27] Klekamp J, Vokel K, Bartels CJ, Samii M. Disturbances of
cerebrospinal fluid flow attributable to arachnoid scarring
cause interstitial edema of the cat spinal cord. Neurosur-
gery 2001;48:174–85.
[28] Fischbein NJ, Dillon WP, Cobbs C, Weinstein PR. The
“presyrinx” state: a reversible myelopathic condition that
may precede syringomyelia. AJNR 1999;20:7–20.