CLINICAL REHORHEOLOGY. Vol. 2, pp. 131-136. 1982
0271-5198/82/010131-06$03.0010 Printed In the USA.
Copyright (c) 1982 Pergamon Press Ltd. All rights reserved.
REVIEW ARTICLE
ARTERIAL FLUID MECHANICS AND ATHEROGENESIS
C, G. CARO
Physiological Flow Studies Unit, Imperial College. London, SW? England
ABSTRACT: The blood velocity field and sites of occurrence of early atheroma art
considered in arterial bifurcations in man. Evidence is increasing that these lesions occur
preferentially where wall %hear stress is low. Mechanisms are examined which might
account for this correlation and also the influence of wall shear Stress on the permeability
of the arterial intima to macromolecules. Hypertension seemingly fa o , rs the development
of atheroma, but the focal occurrence of lesions is not readily explained in terms or the
arterial blood pressure. A widely held view is that atheroms results from darn2ge to the
endothebum and increased influx of Upoprotein, and interaction of platelets with the
damaged wall leading to smooth muscle cell replication and migration. Exposed Sub-
endothelium is, however, rarely seen, except over advanced lesions and a theory appc2rs
necessary to account for the observed findings in the absence or endothelial damage.
Recent studies suggest that the media may offer a barrier to the drainage of material from
the subenclothelial space to the adventitia and that the transport properties of the media
may be influenced by arterial blood pressure and media] smooth muscle tone.
Arterial fluid mechanics, atherogenesis, blood velocity field, early atheroma,
wall shear stress, absence of endothelial damage.
The view has been held for over a century that et a], 1980; Pedley. 1980). Arterial distcnsibUity
arterial haemodynamics can influence the devel- influences arterial blood flow. However. as a
opment of a the rorn a Thus, Virchow (1862) in meant of examining aspects of the mechanics in
propounding the still widely regarded insudation detail and because several theories for athero-
theory for atheroma, suggested that the arterial genesis have been proposed relating to either the
wall takes up material from the passing blood. blood flow or the blood pressure, we sliall mainly
Anitschkow (1933) on the basis of various consider the flow and flow-related theories and
experiments extended this concept, proposing the pressure and pressure-related theories sepa-
that material, having traversed the endothelium, rately.
passes through the thickness or the wall to be The complexity of blood flow in arteries
carried away by the lymphatics or veins of the arises in large part from the inertia of the blood
adventitia. Rindfleisch (1872), in recognition of (both the Reynolds number and WOMMICY
the patchy occurrenceof atheroma.hypothesized frequency parameter take relatively high values),
that the lesions develop at sites in arteries which the oscillatory nature of the flow and the compli-
experience the full stress and impact of the cated geometry of the vessels. Attention has been
blood. Progress in fields including the haemo- devoted to blood how both in branches and
dynamics, structure and mass transport of curved vessels. However, consideration will be
arteries, has permitted the more precise formula- confined here to flow in branches. because this
tion and assessment of these theories. It must be has been of particular interest in respect of
said, however, that many aspects of the total theories for atherogenesis (Caro et a], 1978).
subject remain poorly understood. Furthermore, Understanding of arterial flow has been retarded
given the extent of the field. no more than a because of the difficulties of measuring blood
brief examination of a few topics can be made in velocity in arteries, indeed, no satisfactory means
the lecture. yet exists for measuring the wall shear in blood
The fluid mechanics of the arteries have been vessels.
the subject of recent reviews (Cox, 1979, Patel We examine first the relatively simple case of
131
IC!10
BATCo document for Province of British Columbia 15 April 1999

132 FLUID HFCHANICS AND ATHFROGENESIS Vol.1, No I/ 2
a steady fully developed laminar flow of a
Newtonian fluid entering a symmetrical model
arterial bifurcation, with equal distribution of
the flow between the daughter tubes (it is neces-
sary to consider a three-dimensional junction,
because a two-dimensional structure rules out the
important secondary motions). The velocity
profile is not axisymmetric in the daughter tubes
as it is in the parent tube. In the plane of bifur-
cation, for example, velocity is high near to the
inner wall (flow divider) and low near to the
outer will. Since the fluid is Newtonian (a reason-
able approximation for blood flowing in arteries)
the velocity gradient (shear rate) anti streo
at the wall have the same spatial distribution.
The now in an actual arterial bifurcation is
more complicated. Even if vessel distensibility is
neglected, there will still be, due to the inertia of
the blood and the oscillatory nature of the flow,
local variation during the pulse cycle of the magni-
tude and direction of the wall shear stress
(Langille and Adamson, 1981). Quite close
correspondence has been 'found between the
spatial distribution of the mean, the maximum
and the amplitude of the shear at the wall in a
cast of the human aortic bifurcation (Friedman
et al, 1981). Moreover, the distribution of the
wall shear corresponded quite closely with that
described above for the model bifurcation. Never-
theless, additional complexity can be expected
in vivo. The total flow into the parent tube, the
fraction of the flow leaving it and entering
branches upstream and the distribution of the
flow between the daughter tubes can be expected
to vary physiologically. As a result there may be
variation on a time-scale longer than the pulse
cycle, of the overall shear stress and of the local
velocity profile both in the parent tube and in
the bifurcation (Roach, 1977). Furthermore,
flow separation can occur if, for example, there
is a sharp comer at the outer wall of the bifur-
cation (Pedley, 1980). Large fluctuations of the
shear stress will develop in the region of reattach-
ment of in oscillating separated flow (Matunobu,
1978). A further complication that could arise in
vivo is that, at least in the larger proximal arteries,
flow instability and turbulence can occur (Nerem
and Seed, 1972; Pedley, 1980).
Atheromatous lesions appear to develop
particularly near to sites of branching and
curvature. Two findings which have aroused
interest in this distribution are that (i) the arterial
endothelium can be damaged by experimental
elevation of the hydrodynamic shearing stress
permitting increased entry of macromolecules to
the wall (Fry, 1968) and (H) elevation of the
shearing stress to an extent insufficient to
produce endothelial damage can stJI enhance
macromolecule uptake by the wall (Fry. 1968;
Caro, 1973;). in view of the complexity of art trial
blood flow and the temporal and spatial variation
of the magnitude, and direction of the wall shear-
ing stress, it is not possible to state with certainty
that a causal relationship exists between the level
of the shear stress and the development of
atheromatous lesions. It appears, however, that
the lesions tend to occur preferentially where the
wall shear stress is expected to be low and that
high shear stress regions tend to be spared. The
evidence underlying that statement derives from
Caro et al (I 969. 197 1) relating to 'early' lesions
in man, Friedman et al (1981) relating to intimal
thickening and sudanophUia in man and Kjaernes
et al (198 1) relating to 'early' lesions in younger
persons. A similar distribution is also reported
for the spontaneous lesions which develop in the
arteries of White Cameau pigeons (Comhdl et a[,
1980). In contrast, the arterial lesions that
develop at least initially after placing animals on
lipid enriched diets, appear to be located mainly
in regions expected to experience high wall shear
stress (Caro et al, I 969, 1971; CornhOl and
Roach, 1976; Kjaernes et al. 198 1 ).
These findings would not seem to support the
view proposed that high wall shear stress plays a
causative role in the development of spontaneous
atheroma. They would seem consistent. however,
with a causative role for low wall shear. A theory
involving shear - dependent egress -controlled
efflux of accumulating material from the arterial
wall has been advanced to account for the distri-
bution of the lesions both in spontaneous and
experimentally induced atheroma (Caro et al,
197 1 ). The dependence of the mass transport on
shear relates to the shear stress rather than to the
shear rate (Caro and Nerem, 1973). If findings
concerning the uptake and metabolism of lipo-
protein and cholesterol by cells in culture are
applicable to the arterial wall, then it would be
expected, as is observed, that the fatty acid
L.P4
L.n
.Z:w
BATCo document for Province of British Columbia 15 April 1999

Vol.2, No-1/2 FLUID MECHANICS AND ATHEROGENESIS 133
composition of some of the acLarriulating
cholesterol ester would differ from that of the
cholesterol ester in plasma lipoprotein (Brown
and Goldstein, 1976). If the concentration of the
accumulating material came to exceed that in the
blood and it was able to diffuse from the wall
(which is not established) it would be expected
to pass both towards the blood in the lurnen and
towards the vasa vasorum and advenLitial
lymphatics. Another mechanism proposed to
account for the above findings involves concent-
ration polarisation in low shear regions in fluid at
the blood: wall interface (Keller, 1974).
The hydrodynamic shearing stress influences
the shape and alignment of the endothelial cells
(Flaherty et al, 1972; Stelibens, 1979; Nerem et
al, 1980; Dewey et al, 1980; Longille and Adam-
son, 198 1). Indeed, it may emerge that this
effect will serve to indicate the magnitude and
direction of the wall shear stress (Nerem et al,
1980). The mechanism whereby the shear stress
influences intimal permeability to macromole-
cules at levels below those required to cause
damage is not established. Consideration has
been given to this problem by Fry (1969);
Gerrity and Schwartz (1977); @Veinbaum and
Caro (1976); Oka (1979); Copley (1979; 1981).
It may be of significance that steady shear stress
does not appear to influence the permeability of
the intima of arteries to albumin in experiments
of short (10 min) durafion (Auvert and Lever,
1979).
We make no attempt to review the extensive
literature on the gross mechanics of the arterial
wall, or the propagation of the pressure pulse in
arteries (see, for example. Caro et al, 1978; Cox,
1979; Pedley, 1980; Patel et al, 1980). We con-
fine consideration of the arterial blood pressure
to its possible role in the causation ofatheroma,
either by producing wall damage or by influenc-
ing wall mass transport. Hypertension is a risk
factor for myocardW infarction and is believed
to predispose to the development of atheforna.
Mechanisms can be envisaged for an influence of
the blood pressure on the averaLU extent of
atheromia; the blood pressure influences, for
example, the transport of macromolecules across
the intima (Duncan et a[, 1962; Fry, 1974;
Auvert et al, 1980). It is. however, less easy to
envisage blood pressure being implicated in the
focal development of lesions since, unlike wall
shear stress, blood pressure does not appear to
vary spatially to a major extent. However, it must
be added, that the details of the mechanical
properties of arteries are not known on a fine
scale, say in relation to sites of branching. More-
over, substantial elevation of the blood pressure
can alter the morphology of the endothelial cells
(Baldwin, 1980) and cause wall damage (Fry,
1974).
It is appropriate in the light of these various
observations to examine a widely held theory
for atherogenesis. This involves (i) damage to
the endothelium, possibly by the arterial blood
pressure or Dow fli) increased influx of Upo.
protein resulting from the darhage and (ifi)
interaction of the platelets with the damaged
wall, leading to changes including the replication
and migration of smooth muscle cells. Since early
lesions appear to occur in regions in arteries
where the endothelium is not expected to suffer
mechanical damaet. we also consider evidence
concerning the integrity of the arterial endo.
thelium. It is reported that areas of exposed sub-
endothelium are not demonstrable in arteries
under normal or hyperlipaemic conditions. even
in regions where there is a high rate of cell
replication (Schwartz, 1980): endothelial cell
denudation may be observed over advanced
lesions, or after prolonged hyperlipaernia. These
various results would seem to call for (i) further
study, including critical assessment of hypotheses
for the development of spontaneous early
atheroma, which require endothelial damage and
(H) consideration of hypotheses which could
account for the findings in the absence of
endothelial damage.
Most emphasis in the study of atherogenesis
has attached to the intima despite (i) long-stand-
ing evidence consistent with a continuous
transport of material occurring across the normal
arterial wall (ii) the blood p1esSUTe having been
shown to influence convective transport across
the wall in certain circumstances in a manner
consistent with the wall becoming compacted
(Yamartino et al, 1974; Harrison and Massaro,
1976) and (iii) inadequate drainage across the
media, or to the adventitial lymphatics having
been suggested as mechanisms favouring
atherogenesis (Walton, 1975, Ad2ms, 1973;
C=)
t__J
BATCo document for Province of British Columbia 1 5 April 1999

134 FLUID MECHANICS AND XTHEROGENESIS Vol-1, No.1/2
Jellinck at al. 1970. Caro et al, 1980a).
In a recent investigation involving rabbit
arteries perfused in situ and the essentially inert
extracellular material. radioactively labelled
albumin, we have obtained evidence which
provides strong supphirt for the existence of a net
transport across the normal arterial wall (Caro
et al, 1980a). The findings imply an important
role for the adventitial circulation and lympha-
tics. They imply, moreover, that when a steady
transmural transport has been established the
media offers a substantial transport resistance.
This may relate to the observed marked
exclusion of the tracer from the tissue.
In more recent studies (Caro et al, 1980b), we
have found that (i) elevation of arterial blood
pressure, when PTCSSUre-driven convection of
fluid across the wall Is restricted by interfacial
forces, reduces the distribution volume or labelled
albumin, seemingly by slightly compacting the
medial interstitium and (ii) vasoactive materi"
under similar experimental conditions influence
the distribution volume for the tracer in the
media. The control value in arteries pressurized
to 70mmHg with saturated air was 0.035 SEM
0.003(n=9). At the same air pressure with
10-'Mnoradrenaline in the external incubating
fluid it was 0.027SEM 0.003 (n=6) and with the
smooth muscle relaxant sodium nitrite at a
concentration of 1(r4M it was 0.049 SEM 0.007
(n=7). The influence of lurninal pressure and
smooth muscle tone are still under investigation
when vessels are pressurized with plasmas Under
these conditions endothelial permeability (which
can be affected by vasoactive substances (Shima.
moto, 1971; Robertson and Kh2jrallah, 1972))
could also play a role, since in such arteries
endothelial damage has been found to slightly
enhance the sucrose space in the wall and to
substantially increase the space for labelled
albumin (Lever and Tedgui, 1981).
These various findings are consistent with the
media providing a barrier to the drainage of
materials from the subendothelial space to the
adventitia and the magnitude of the barrier being
susceptible to factors, including the level of the
arterial blood pressure and the tone of the
smooth muscle of the media. Extensive investiga-
tion is needed to test these possibilities, but it
may be noted that the lesions of atheroma
develop in the larger musculu-elasiic artarie-. znd
between the endothelium and inner media.
References
1 .Adams, C. W. M. (1967). Vascular Histochernistry
in Relation to the Chemical and Structural Patho-
logy of Cardiovascular Disease, Lloyd Luke,
London.
2.Anitschkow. N. (1933). Experimental arteriosclc-
ratis in animals. In: Arteriosclerosis Jed E. V.
Cowdry). Macmillan. New York.
steady shear stress on tht uptake ofalburnin by the
rabbit thoracic aorta. 1. PhysioL. 292. 67p.
4.Auvert, B., Lever, M. J, and Caro, C. G. (I 980).
Influence de la presuon aztenelle et du cisaillement
parietall sur le transport de I'altiamine , "' I dans
I'aorte thoraique de lapin. J. Biophys. et Med.
Nucl. 4. 273.
5.Baldwin, A. (1980). Personal communication.
6.Brown.M. S. and Goldstein, J. L. (I 976). Receptor-
mediated control of cholesterol metabolism.
Science 191, 150.
7.Caro, C. G.. Fitz-Gerald. J. M. and Schroter. R. C.
(1969). Arterial wall shear and distribution of early
atheroma in man. Narure, 223. 1159.
B.Caro, C. C., Film-Gerald. J. M. and Schroter, R. C,
(I 97 1). Atherarria and arltenal wall shear. Observa-
tion, correlation and proposal of a shear depencien t
mass transfer mechanism for atherogericsis. Proc.
R oy. Soc. Land. B. 177,109-159.
9.Caro, C. G. and Nercm. R. hi. (1973). Transport or
"C cholesterol bet,%een serum and wall in perfused
dog common carotid artery. Circulation Res. 32,
187.
10.Caro. C. G.. Pedley, T. J.. Schroter. R. C. and Seed,
W. A. (1978). The Mechanics of the Circulation.
Oxford University Press.
11.Caro, C. G.. Lever, M. L. Laver-Rudich. Z., Meyer,
F., Liron,N., Ebel. W., Parker, K. H. and Winlove,
C. P. (198%). Net albumin tramport across the
wall of the rabbit common carotid artery perfuscd
in situ. Atherosclerosis. 37.497. ,
12.Caro, C. G.. Lever, M. J. and Tedgui, A. (19ROb).
Effect of active and passive stresses on distribution
volume in the arterial wall in Hernodynamics and
the Arterial Wall. ed. R. M. Nerem and J. R.
Guyton, Nat. Science Foundation, Sponsored
Specialists' Mtj. Houston, Texas.
13.Copley. A. L. (1979). Fibrinogen. platelets and a
new theory of atherogenesis. Thrombosis Res. 14,
249.
14.C opley. A. L. (I 98 I). Hem orheclogical aspects of
the endothelial fibrin lining and of fibrinogen pi
clotting. Their importance in physiology and
pathological conditions. Clinical Hcrnorheology, 1,
15.Conihill, J. F. and Roach, M. R. (1976). A quanti-
BATCO document for Province of BritiSh Columbia 15 April 1999

Vol-2, No-1/2 FLUID MECHMICS AND ATHMOGENESIS 135
tairve study of the localisittion of achcromwWolic
lesions in the rabbit aorta. Atherosclerosis 24 499.
16.Cornhill. 1. F.. Levesque. M. 1. and Nerem. R. M.
(1990). Quantitative study of the localisation of
suclanophilic coeliac lesions in the while cameau
pigeon. Atherosclerosis 33, 103.
17.Cox. R. H. (1979). Physiology and hemodynamics
of the macrocirculation in: Hemodynamics and the
Blood Vessel Wall, Ed. W. E. Stehbens, Charles C.
Thomas, Springfield.
18.Dewey. C. F. Jr., Gimbrove, M. A. It., Bussolani.
S- R. and Cruies, S- A. (I 980). The dynamic
response Of VI=IAF endothelial cells to fluid shear
stress in: Hernodynarnics and the Arterial WalL ad.
R. M. Nercm and J. R. Guytom National Science
Foundation Sponsored Specialists' Mtg, Houston,
Texas
19.Duncan, L. E, Cornfield, J. and Buck. K. (1962).
The effect of blood pressure on the passage of
labelled plasma albumin into canine aortic wall. 1.
Clin. Invest. 41. 1537.
20)Flaherty. J. T., Ferran, V. J.. Pierce, J. E., Carew,
T. E. and Fry,D. L. (1972). Localising factors in
experimentalatherosclerosis in Likoff. W., Seg&L
B. L, Insult.W. and Mayer. L H. (ads.) Atheyo-
sclerosis andCoronary Heart Disease. Gnine and
Stratton, New York.
21.Friedman, M. H.. Hutchins, M. C, Bargercin, C. B.,
Deters. 0. J. and Mark, F. F. (1981). Correlation
between :ntimal thickness and fluid shear in human
arteries. Atherosclerosis 39. 425,
22.Fry, D. L. (1968). Certain histologicaland chemical
responses of the vascular interface to acutely
induced mechanical stress in the aorta of the dog.
Circulation Res. 24, 93.
23Fry, D. L. (1%9). Certain chemorheologic con-
siderations regarding the blood vascular interface
with particular reference to coronary artery disease.
Circulation 39, Suppl. 4, 39.
24.Fry, D. L. (1974). Some aitenal changes associated
with hemodynamLc events in: Fluid Dynamic
Aspects of Arterial Disease: National Science
Foundation Sponsored Specialists' Meeting, ad. R.
M.Natern, Columbus, Ohio.
25.Gerrity, R. G. and Schwartz. C. 1. (I 977). Progress
in BiochLm. and Phannaccil. 14, 134.
26.Harrison, R. G. and Massaro. T. A. (1976). Water
flux through porcine aortic tissue due to a hydro-
static pressure gradient Atherosclerosis 24. 363.
27.Jellinek, H., Veress, B, Balint, A. and Nagy, Z.
(1970). Lymph vessels of rat aorta and their
changes in experimental atherosclerosis an electron
microscopic study. Exp. M ol. Path. 13, 3 70.
28.Keller, K. H. (1974). The influence of shear
dependent diffusion in blood on atherogenesis and
thrornbogentsis in Fluid Dyntimic Aspects of
Arterial Disease: National Science Foundation
Sponsored Specialists' Meeting, ad. R.M. Nerem,
Columbus, Ohio.
29.Kajaernes. M.. Svindland, A.. Wall6e, L. and Witte,
S. 0. (1981). Localisation of early atherosdaot3c
lesions in a arterial bifurcation in human& Acta
Path et MicTobiol. Scand. A. 89.35.
30.Langille. B L. and Adamson. S. L. (1981). Reli-
tionships between blood flow direction and endo-
thelial call orientation at arterial branch site% in
rabbits and rnice. Circulation Res. 48, 481.
31.Lever. M. 1. and Tedgui, A. (I 981). Oedema of the
artery wall following intimal damage. Proc. Physiol.
Soc. Cambridge Meeting, June I S.
32.Matunobu. Y, (1978). The wall-shear-stress intens-
ification due to pulsatile blood flow. Proc. 3rd Intl.
Congr. Biorheol. 204, La Jolla. California.
33.Nerem, R. M. and Seed, W. A. (1972). An in vivo
study of the nature of aortic flow disturbances.
Cardiovasc. Res. 6. 1-14.
34.Nercin. R. M., Levesque. &I. 1. and Comhdl, 1. F.
(1980). Arterial fluid mechanics and the enclo.
thelium in Hemodynimics and the Arterial Wall.
ad. R. M. Nerem and J. R. Guyton, National
Science Foundation Sponsored Specialists'Meating,
Houston. Texas.
35.Oka, S. (1979). Physical theory of permeability of
vascular walls in relation to atherojenests. Biarhe-
ology 16. 203.
36.Patel. D. J. and Vaishnav. R. N (1980). Basic
Hernodynarnics and its Role in Disease Process.
University Park Press. Baltimore.
37.Pedley, T. J. (1980). The Fluid Mechanics of Large
Blood Vessels, Cambridge University Press.
38.Rindfleisch, E. (1872). A manual of Pathological
Histology, Vol. 1. London: New Sydenham
Socie ty.
39.Roach, M. R, (1977). The effects of bifurcations
and stenoses on arterial disease in Cardiovascular
Flow Dynamics and Measurements. ad. N. H. C.
Hwang. and N. A. Normann, University Park Press.
Baltimore
40.Robertson, A. L. and KhaLrallah. P. A (1972).
Effects of angtolenston 11 and some analogues on
tlie vascular permeability in the rabbit. Circulation
Res. 31. 923.
41.Schwartz, S. M. (1980). Role of enclothehal inte.
grity in atherosclerosis in Hydrod;,namics and the
Arterial Wait, ad. R. M. Ncrem and J. R. Guyton,
National Science Foundation Sponsored Specialists'
Meeting, Houston, Texas
42.Stchbens, W. E. (I 979). Anatomy and structure of
blood vessels in Hemodynamics and the Blood
Vessel Wall, ad. W.E. Stchbens. Charles C. Thomas.
Springfield.
43.Shimamoto, T.. Yamashita, Y., Numino, F. and
Sunaga. T. (1971). Scanning and electron micro-
scopic observation or enciDthel3al cells in the
normal condition and in initial stages of athero-
sclerosis. Acta Path. Japan 21, 93.
44.Virchow. R. von (I 86 2). Phlogrise und Thromhose
in Gcfassystems Gesammelte Abhandlungen zum
BATCo document for Province of British Columbia 15 April 1999

136 FLUID MECHANICS AND ATHUOGENESIS Voi.2, No.1/2
Wissenschaftlichim Medum. M3x Hinch, Berlin. sateen observed in election microscopic studies.
4S. Walton. K. W. (1975). Pathopnetic mechanism in J. Fluid Mech. 74. 61 1.
atherosclerosis. Amer. 7. Cardiol. 35,542. 47, Yamartino. L, Bratzlef, R., Coltori, C.. Smith. K.
46. Weinbassm, S. and Aro, C. G. (I 976). A macro- and Lzes, R. (1974). Hydraulic permeability of
molecule transport model for the arterial wall and arterial tissue. Circulation 49 and 50 (Supplement
cridothelium based on the ultrastructurt speciali- 111).273.
ON
BATCo document for Province of British Columbia 15 April 1999