100 Practical Neurology
NEUROLOGICAL SIGN
Dermatomes and dogma
V Apok,1 N T Gurusinghe,2 J D Mitchell,3 H C A Emsley3
1
Registrar in Neurosurgery, Royal
Manchester Children’s Hospital,
Manchester, UK
2
Consultant Neurosurgeon,
Department of Neurosurgery,
Royal Preston Hospital, Fulwood,
Preston, UK
3
Consultant Neurologist,
Department of Neurology, Royal
Preston Hospital, Fulwood,
Preston, UK
Correspondence to
Dr H C A Emsley, Department of
Neurology, Royal Preston Hospital,
Sharoe Green Lane, Fulwood,
Preston PR2 9HT, UK;
h.emsley@liv.ac.uk
10.1136/jnnp.2011.242222
Pract Neurol 2011; 11: 100–105
The concept of dermatomes came from early attempts to correlate the
physiology of sensation with anatomy. There are various definitions of
dermatomes and several maps in common use. While useful, dermatomes
are subject to considerable variation between maps and, indeed, between
individuals. Anecdotally, precise dermatome distributions are generally regarded
by experienced neurologists with a degree of caution, being viewed as an
approximation. In this article, we consider the validity of the dermatome maps
and their background, as well as introducing a relatively recent ‘evidence based’
dermatome map.
DERMATOMES AND THEIR question. Ever since the first attempts at
SIGNIFICANCE mapping dermatomes in the late 19th cen-
Localisation of sensory symptoms and signs tury, neurologists have used dermatomes in
to specific parts of the central and periph- their clinical diagnosis of radiculopathy and
eral nervous systems is a significant part of in determining the level of spinal cord injury.
the neurological examination and diagnostic Neurosurgeons and neurophysiologists rely on
evaluation—the crucial ‘where is the lesion?’ dermatomes for intraoperative monitoring of

spinal cord function through somatosensory
evoked potentials. And reliance is also placed
on dermatomes in the practice of regional
anaesthesia.
The concept of dermatomes originated in
early attempts to correlate the physiology of
sensory experience with an anatomical sub-
strate. Today, the term ‘dermatome’ gener-
ally refers to an area of skin innervated by a
particular neural element, specifically nerve
root, dorsal root ganglion or spinal segment.
Dermatomes are of course distinct from the
areas of skin supplied by particular peripheral
nerves, these often referred to as the periph-
eral nerve fields (or cutaneous nerve distri-
butions). Despite the long tradition, and the
emphasis still placed on teaching dermatomes
to medical students, experienced neurolo-
gists, perhaps because they are well aware of
the ‘approximate nature’ of the various maps,
probably attach rather less significance to the
precise dermatome distribution than the cor-
responding myotomes in lesion localisation.
Alteration of sensation in a dermatome is a
sign about which neurologists are rightly cir-
cumspect. In this article, we will consider the
validity of dermatome maps, how they evolved
and draw attention to a recently devised ‘evi-
dence based’ dermatome map.
VALIDITY OF CURRENT
DERMATOME MAPS
Perhaps surprisingly, the dermatome maps in
current use were largely constructed in the
early half of the 20th century by Sir Henry
Head, Otfried Förster, Jay Keegan and Frederic
Garrett. Currently, there are 14 different maps
in 13 different major texts.1 Even individual
texts (eg, Gray’s Anatomy) have variations
between different editions. The overwhelm-
ing message seems to be that these maps are
inaccurate, with a surprising lack of consensus
about the size and location of dermatomes.
This has much to do with the methodology
that was used to draw these maps.
The physiological means by which most
of the maps were derived did not take into
account various points of ambiguity, such as
the definition and the nature of the neural ele-
ment being mapped. Compounding this ambi-
guity is the degree of variation and overlap
between adjacent dermatomes, between and
even sometimes within (eg, unilateral brachial
plexus variant) individuals (box 1 refers to
Apok, Gurusinghe, Mitchell, et al 101
the prefixed and postfixed brachial plexus as
an example of such variation2). There is some
agreement in the literature that dermatomes
in reality might be larger in area than those
shown in traditional texts and therefore have
a greater degree of overlap than originally
acknowledged. So a lesion of a nerve root
may produce a much smaller area of sensory
loss than a casual glance at a dermatome map
might suggest. The three most commonly ref-
erenced dermatome maps in contemporary
anatomy texts are those of Head and Campbell
(1900), Otfried Förster (1933) and Keegan and
Garrett (1947).
HOW DERMATOMES USED
TODAY CAME TO BE
Henry Head’s map
Henry Head (1861–1940), a physician at the
Royal London Hospital, published the first
widely accepted dermatome diagram in 1900.
A voracious researcher into sensory localisa-
tion and pain in visceral disease, he sectioned
his own superficial radial nerve and diligently
documented the developing sensory distur-
bance.3 His dermatome mapping work was
coauthored with Alfred Walter Campbell A lesion of a nerve
(1868–1937) and was largely based on draw-
ings and photographs of 450 patients with root may produce a
herpes zoster eruptions.4 The final product was much smaller area
the result of this large study of herpes zoster
patients as well as observations of patients
of sensory loss than
with spinal cord injuries and those with pain a casual glance at
due to visceral non-neurological disorders, in
whom Head described, for example, “positions
a dermatome map
over which the patient experienced pain in might suggest
Box 1 The ‘prefixed’ and ‘postfixed’ brachial plexus2
Anatomical sources of variation leading to deviation from the expected
dermatome distribution, as well as differences in motor supply of the upper
limb, include the prefixed and postfixed brachial plexus. Most standard
medical textbooks describe the brachial plexus as arising from the lower
four cervical nerves and the first thoracic nerve with an occasional
contribution from the fourth cervical and second thoracic nerve. A prefixed
brachial plexus has been described as one with a large contribution from
the fourth cervical nerve with or without a small contribution from the first
thoracic nerve. A postfixed brachial plexus is one with a large contribution
from the second thoracic nerve and little or no communication with the
fifth cervical nerve. It is worth being aware that brachial plexus variations
are more the rule than the exception—not only in terms of unexpected
dermatome distribution but also because of the potential predisposition to
certain conditions, such as thoracic outlet syndrome.
www.practical-neurology.com
102 Practical Neurology
gastric disturbances”—that is, areas of referred
cutaneous tenderness or allodynia.5
Analysis of Head’s work reveals limitations
in methodology and interpretation. His earlier
work on the L1 dermatome for instance was
based on a patient with first and second lum-
bar vertebral fractures who had bilateral L1
nerve root involvement at surgery with bilat-
eral T12 nerve root sparing.6 The upper border
of sensory loss in this case was taken to be the
upper border of the L1 dermatome, although
the possibility of dermatomal overlap was
apparently not considered. L5 was determined
by analysing cutaneous tenderness in a patient
with an inflamed right lobe of the prostate
gland; having already mapped S1–5 and L1 in
other patients, Head considered that the only
remaining area within this region of tender-
ness, the lateral aspect of the leg, on account
of its adjacency to the sacral skin segments,
must represent the L5 dermatome. This con-
cept of adjacent skin segments having adja-
cent root and spinal segments was not true
however for the L4 and S2 segments of his
map. The L4 dermatome was loosely derived
from the observation of a single patient whom
Head claimed had a spinal cord injury although
no further details are available on the nature
of the injury. Following this, he established the
L3 dermatome by elimination in a patient with
herpes zoster in whom he deemed there to be
L3, L4 and L5 involvement. He conjectured that
the L3 dermatome must represent the area not
included in his previously determined L4 and
L5 dermatomes.6
A substantial drawback of Head’s study of
herpes zoster cases was that histological con-
firmation of single dorsal root ganglion inflam-
mation was apparently obtained in only 16 of
the 450 patients. Furthermore, among these
16, not all dermatomes were represented, there
were no recorded examples of C5 through C8
or any root level below L1. Assumptions were
often made about the precise dorsal root
involved in the production of the resulting
map, the first to document the thoracic der-
matomes. It is also now known that herpetic
eruptions can affect several adjacent dorsal
root ganglia simultaneously. Furthermore, not
all the cutaneous fibres within a dorsal root
ganglion may be affected. Consideration of
these points highlights various flaws in the
proposed map. Indeed Head himself acknowl-
edged the ambiguity of his findings by noting
10.1136/jnnp.2011.242222
that“….my areas correspond to the supply not
of roots, but of segments of the spinal cord
from which the roots in part arise”.
Otfried Förster’s map
Otfried Förster (1873–1941) was a German
neurologist who turned to neurosurgical prac-
tice at the age of 40 years. Posterior rhizoto-
mies were being undertaken at the end of the
19th and start of the 20th centuries for the
treatment of spasticity, but fell out of favour
because of unacceptable adverse effects.
He developed his map by surgically isolating
single dorsal nerve roots after sectioning the
dorsal nerve roots above and below the root
under investigation. These experiments were
modelled on those of Sir Charles Sherrington
(Nobel Prize winner for physiology and medi-
cine in 1932) who researched dermatomes in
monkeys with this same method of section-
ing nerve roots to isolate a single nerve root.
Förster extended these experiments to human
subjects stating, “….I need not discuss circum-
stances under which such a selected proce-
dure may be undertaken”.5
A significant flaw in Förster’s methodol-
ogy was his lack of consistent documentation.
There is little information provided on his
method of assessing and reporting on der-
matomes. Moreover, he also failed to note the
time lapse between sectioning and dermato-
mal testing. This, in particular, may have had
significant implications as a result of physi-
ological processes such as Wallerian degen-
eration or even nerve regeneration, leading to
the possibility of shrinkage in the extent of
sensory loss over time.
Förster acknowledged the phenomenon of
dermatomal overlap and individual variation,
concluding that sectioning of a single nerve
root was never accompanied by sensory loss.
However, his maps were notable for not showing
portions of the limbs or the posterior trunk.5
Keegan and Garrett’s map
Finally, Keegan and Garrett proposed their map
in 1947, 13 years after the first reported case of
back and leg pain from a herniated disc.7 Based
on their initial observations that herniated discs
compressing nerve roots were associated with
diminished sensation, they set about construct-
ing a dermatomal map based on observations
of hypoalgesia in patients with disc prolapse
(165 cervical and 1264 lumbosacral, of which
 Apok, Gurusinghe, Mitchell, et al 103

28% and 56%, respectively, were confirmed at
operation).1 This map consisted of neat, non-
overlapping dermatomes, despite the authors’
acknowledgement that dermatomes did over-
lap. These dermatomes almost always reached
the midline. Theirs is the most widespread der-
matome map in contemporary use but is argu-
ably the most flawed of the three main maps
• Aids to the examination of the peripheral
discussed thus far. The site of nerve root com-
pression was based on myelographic rather
than operative findings in the main. They also
claimed a high degree of reproducibility for
their map, with no more than 1 cm variation
between individuals. Importantly, Keegan sup-
ported the concept that intervertebral disc
compression of a single nerve root results in
an area of cutaneous sensory loss, contradict-
ing the work of Sherrington and Förster, per-
haps on account of the different physiological
characteristics of nerve root compression as
opposed to sectioning.8
its acknowledgement of common areas
of significant overlap—for example, C7
overlaps considerably with C6 and C8, the
dorsal surface of the hallux is commonly
innervated by L5 but can also be supplied
by L4 and the S1 dermatome extends as
far superiorly and posteriorly as the but-
tock and overlaps with S2.
nervous system12 is arguably one of the
most authoritative yet accessible mono-
graphs devoted to examination of the
peripheral nervous system (see page 106
for review of this book), its own rich his-
tory recently revisited.11 It does indeed
highlight the problems of overlap and
variability affecting dermatome maps.
For example, the current author—Michael
O’Brien—points out that while the usual
corresponding dermatomes for thumb,

A NEW ‘EVIDENCE BASED’

DERMATOME MAP
There have been few attempts at verifying
these original dermatome maps, especially in
more recent years:
• Of note is the prospective study of 403
patients by Kortelainen.9 Pain referral
patterns and neurological findings were
charted. Radiological findings (using CT)
and surgical root irritation did not always
correlate with anticipated pain referral
zones; this was a significant contribution
to the mounting evidence of the limited
utility of dermatomes in clinical practice.
• Nitta et al 10 selectively blocked nerve
roots with xylocaine injections under
fluoroscopic guidance in patients with
radicular pain due to disc herniation. They
concluded that the characteristic L4, L5
and S1 dermatomes were only present in
about 80% of patients. The implication
was that one in five patients had inner-
vation patterns other than those of the
traditionally described dermatomes.
• In their comprehensive review exploring
the controversies of dermatome maps, Figure
Lee et al 1 proposed an ‘evidence based’ The ‘evidence based’ dermatome map representing the most consistent tactile dermatomal areas for
each spinal dorsal nerve root found in most individuals, based on the best available evidence. The
dermatome map formed from the assimi-
dermatomal areas shown are not autonomous zones of cutaneous sensory innervation. Except across
lation of previous maps (figure); while the the midline where overlap is minimal, adjacent dermatomes overlap to a large and variable extent.
use of the term ‘evidence based’ may be Blank regions indicate areas of major variability and overlap. S3, S4 and S5 supply the perineum but
stretching a point, this map is at least are not shown for reasons of clarity. Note consecutive dermatomes shown in buff or blue for clarity.
an attempt to systematically distil the From Lee et al.1 Copyright Wiley-Blackwell (2008). This material is reproduced with permission of
Wiley-Blackwell, a subsidiary of John Wiley and Sons, Inc.
best available evidence. It is notable for

www.practical-neurology.com
104 Practical Neurology
Box 2 Testing dermatomes
The history will usually determine whether examination of dermatomal
sensation is required. The patient should be asked to indicate any area of
altered sensation, including its limits. It is usually not necessary to test all
dermatomes, with the examination focusing instead on the region suggested
by the history. For sampling dermatomes, it is customary to move from
distal to proximal along the long axis of the medial and lateral borders of
the limbs, and ascending vertically on both sides of the trunk. If there is a
reported area of sensory impairment to pinprick the examination should
proceed from the centre of the area of maximum abnormality towards the
normal area to define the borders of the area of altered sensation. If there is
an area of enhanced sensation, usually hyperalgesia, the examination should
proceed in the reverse direction. The patient is asked to confirm that the
stimulus is perceived as sharp in each dermatome. Temperature sensation,
often omitted if pain sensation is normal, is undertaken in a similar
sequence. Usually the metal of the tuning fork is the most readily accessible
cold stimulus in the clinic. Arguably, light touch, tested with a wisp of
cotton wool or a light finger touch on the skin, and otherwise following the
same sequence, adds little additional information, although it is said that the
area of deficit can be somewhat larger than that to pinprick in dermatomal
sensory loss.
PRACTICE POINTS
• Examination of cutaneous loss over dermatomes is necessary only when
suggested by the history.
• Dermatome maps are approximations, subject to various methodological
weaknesses, and serve only as a guide.
• There is significant overlap between adjacent dermatomes.
• There is less, if any, overlap between non-consecutive dermatomes,
and therefore these boundaries give more reliable and clinically useful
borders.
middle finger and little finger are C6,
C7 and C8, respectively, the index and
ring fingers are too variable to be clini-
cally useful. Thus there must be some
concern that dermatome maps depicting
the distributions to the ends of the limbs
(including the ‘evidence based’ map) are
flawed in this respect. It is also important
to bear in mind that the ‘evidence based’
map does not show the dermatomes as
autonomous areas. O’Brien refers to there
being less, if any, overlap between non-
consecutive dermatomes, and therefore
these boundaries give more reliable and
clinically useful borders. In practice it is
only necessary to know the approximate
centre of a dermatome and, if appropri-
ate, map the boundary with the principles
outlined in box 2 (based on the sensory
examination sequence ‘Aids’12).
10.1136/jnnp.2011.242222
DERMATOME MAPPING: AN
EXERCISE IN FUTILITY?
The greatest flaw in seeking to map der-
matomes has been the assumption that the
correlation of CNS to skin is a direct and static
one. We now know that neural elements are
continuously being suppressed, facilitated and
reorganised in a dynamic fashion. Moreover, as
far back as 1893, Sherrington demonstrated in
his early experiments on monkeys that the dis-
tribution of sensory fibres is less dense towards
the periphery of a dermatome, hence maps can
only reflect the regions of most intense cutane-
ous innervation. Almost a hundred years later in
1989, Moriishi’s work on cadavers demonstrated
the presence of intrathecal intersegmental con-
nections between the dorsal spinal rootlets,
with the greatest variation in the upper limb
dermatomes.13 This has effectively obviated the
idea of dermatomes being the cutaneous repre-
sentation of dorsal root ganglia.
The most significant work in recognising the
vast complexity of cutaneous innervation was
by Denny-Brown and colleagues.14–16 Through
their experiments on monkeys, they effectively
demonstrated that cutaneous innervation of
dorsal root ganglia is different to that of the
dorsal root. The role of adjacent dorsal root
ganglia and the spinal cord in determining the
cutaneous innervation of a given spinal nerve
root was recognised for the first time. By mod-
ifying Förster’s method, they sectioned nerves
either proximal or distal to dorsal root ganglia
and studied patterns of cutaneous sensibil-
ity. Their most important finding was that the
Lissauer tract (near the substantia gelatinosa
of the spinal cord) is a key mediator of dor-
sal root transmission. The medial and lateral
parts of this tract potentiate and inhibit sen-
sory impulse transmission, respectively. Hence
corresponding lesions result in dermatomal
shrinkage or expansion. In effect, this finding
established that the previously accepted idea
of a direct correlation between neural element
and skin was an overly simplistic one.
That dermatomes can expand and shrink,
depending on the anatomical and physiologi-
cal characteristics of adjacent spinal cord seg-
ments and dorsal root ganglia, has led to an
emerging recognition of the dynamic nature of
cutaneous innervation. This work has yet to be
translated into clinical practice but it heralds a
new page in the history of attempts at under-
standing cutaneous innervation.

AND IN PRACTICE….
There is no place for dogmatic adherence to
classical dermatome maps—rather, in the
teaching and practice of sensory examination,
we all need to be aware of the considerable
frailties of the methods used to derive the
maps. They should serve merely as a guide, with
the examiner clearly understanding their many
limitations such as variation between individu-
als, as well as overlap between dermatomes. It
is however refreshing that an evidence based
approach has been taken, and arguably we
should be embracing more recent anatomical
work when considering dermatomes in clinical
practice rather than perpetuating the weak-
nesses of the original dermatome maps of
Head and Campbell, Förster, and Keegan and
Garrett, while at the same time acknowledg-
ing the crucial contributions of these earlier
workers.
ACKNOWLEDGEMENTS
This article was reviewed by Richard Hughes
and Michael O’Brien, London.
Competing interests None.
Provenance and peer review Not commissioned;
externally peer reviewed.
REFERENCES
1. Lee MW, McPhee RW, Stringer MD. An evidence-
based approach to human dermatomes. Clin Anat
2008;21:363–73.
2. Pellerin M, Kimball Z, Tubbs RS, et al. The prefixed
and postfixed brachial plexus: a review with surgical
implications. Surg Radiol Anat 2010;32:251–60.
Apok, Gurusinghe, Mitchell, et al 105
3. Greenberg SA. Henry Head (1861–1940). J Neurol
2004;251:1158–9.
4. Head H, Campbell AW. The pathology of herpes
zoster and its bearing on sensory localization. Brain
1900;23:353–523.
5. Greenberg SA. The history of dermatome mapping.
Arch Neurol 2003;60:126–31.
6. Head H. On disturbances of sensation with especial
reference to the pain of visceral disease. Brain
1893;16:1–133.
7. Mixter WJ, Barr JS. Rupture of the intervertebral disc
with involvement of the spinal canal. N Engl J Med
1934;211:210–15.
8. Keegan JJ. Neurosurgical interpretation of
dermatome hypalgesia with herniation of the
lumbar intervertebral disc. J Bone Joint Surg
1944;26:238–48.
9. Kortelainen P, Puranen J, Koivisto E, et al. Symptoms
and signs of sciatica and their relation to the
localization of the lumbar disc herniation. Spine
1985;10:88–92.
10. Nitta H, Tajima T, Sugiyama H, et al. Study on
dermatomes by means of selective lumbar spinal nerve
block. Spine 1993;18:1782–6.
11. Compston A. Aids to the investigation of peripheral
nerve injuries. Medical Research Council: Nerve
Injuries Research Committee. His Majesty’s Stationery
Office: 1942; pp. 48 (iii) and 74 figures and 7 diagrams;
with aids to the examination of the peripheral nervous
system. By Michael O’Brien for the Guarantors of
Brain. Saunders Elsevier: 2010; pp. [8] 64 and 94
Figures. Brain 2010;133:2838–44.
12. O’Brien MD. Aids to the examination of the peripheral
nervous system, 5th Edn. London: Saunders Elsevier
(on behalf of the Guarantors of Brain), 2010.
13. Moriishi J, Otani K, Tanaka K, et al. The intersegmental
anastomoses between spinal nerve roots. Anat Rec
1989;224:110–16.
14. Denny-Brown D, Kirk E. Hyperesthesia from
spinal and root lesions. Trans Am Neurol Assoc
1968;93:116–20.
15. Kirk EJ, Denny-Brown D. Functional variation in
dermatomes in the macaque monkey following dorsal
root lesions. J Comp Neurol 1970;139:307–20.
16. Denny-Brown D, Kirk EJ, Yanagisawa N. The tract of
Lissauer in relation to sensory transmission in the
dorsal horn of spinal cord in the macaque monkey.
J Comp Neurol 1973;151:175–200.
www.practical-neurology.com