THE

SIGNIFICANCE
PHILIP

OF

THE

TRENDELENBURG
SYDNEY NADE

TEST

HARDCASTLE,

From

tile University

of Western

Australia

Trendelenburgs four methods We examined order to determine enabled method The operation us to define of assessing major from result least

test

of function

of the hip joint

was

first

reported

before

radiology

was

available.

At

of performing it have since been described in the literature. 50 normal subjects and 103 people with disorders affecting the different responses that occurred when they were asked a standard hip abductor that method of performing function. in misinterpretation, the Trendelenburg or false-positive test,

either the spine or the hip, in to stand on one leg. This has and to interpret are pain, the test lack as a

pitfalls

result

responses,

of cothe hip

responses

the patient, and impingement from the patient using muscles side.

between the rib cage and the iliac crest. above and below the pelvis, and from leaning

False-negative
beyond

on the standing

In

1897

Friedrich

Trendelenburg

described

a test

which

he found useful in determining the integrity tor muscle function, with specific reference dislocation (Rang only other tioned textbooks two than in at a time 1966). years when his most of the after the ears, major hip and progressive report had and few and Trendelenburgs the discovery physician eyes, (Fig. aids The

ofhip abducto congenital atrophy I) appeared by Roentgen, to diagnosis test is menout in physiotherapy of the hip. at least four diftest, clear there the most ofit how was perof on
Positive
Trendelenburgs Fig. I Siga

muscular ofx-rays

fingers.

orthopaedic manoeuvre

as a diagnostic

to be carried

the assessment of function and In the standard textbooks ferent methods of performing usually described the test should little agreement formance them had vaguely, and be interpreted. among our

dysfunction we found the

Trendelenburg

it was seldom Furthermore, colleagues about

or interpretation routinely carried

of the test, although out their own version

many of their patients. Because the #tict/iod, ineatting and value

investigate it.

of this confusion as to of the test, we decided to first, request to record to stand the on

Negative

The responses one test; leg, and

aims of this study of normal people

were, to the

and hence to define a standard Trendelenburg second, to study people with various abnormaliand its muscles test in order to assess the value pitfalls of the as used in current orthopaedic

These drawings tive and negative and the alignment ( From Mercer

demonstrate what Trendelenburg described as a posisign. Note that both hands are held by an assistant of the pelvis with respect to the ground is observed. Rangs Ani/zologt n/orthopaedics, with permission).

ties of the hip and the practice.

SUBJECTS Normal between volunteers. A group 3 and 50 years were of 50 normal people examined to determine stood were flexed on one studied: to 30
.

aged what leg. first and perfor

P. 1-1. 1-Iardcastle, S. M. L. Nade. Queen Elizabeth Requests

I

FRACS. Senior Lecturer in Orthopaedic MD. FRACS. Professor ofOrthopaedic II Medical Centre, Nedlands. Western reprints should he sent to Professor and Joint

Surgery Surgery Australia 6009.

happened to their posture when they Two positions of the non-stance leg the test was performed with the hip then it was repeated with son was asked to maintain

for

S. M. L.Nade. Surgery

1985 British Editorial 0301 620X855150S2.00.

Society

of Bone

the hip flexed to 90 Each the one-leg stance posture

--1985

30seconds.
741

VOL.

67 B. No.

5. NOVEMBER

742 Volunteers 103 people age range They were with studied, of the divided neuromusculoskeletal 12 had bilateral

P. H. HARDCASTLE,

S. M. L. NADE Table I. Classification volunteers I . Neurological

ofthe

neuromusculoskeletal

disorders

ofthe

103

disorders. hip disorders. 6 to (Table with

Of

the The

103 patients was from into two major groups disorders hip or spine. DATA and those

82 years. I): those

conditions

with neurological disorders ofthe

mechanical

Incomplete paraplegia Muscular dystrophy Nerve root entrapment Cerebral palsy Poliomyelitis Hemiplegia

II 3
.5

.5

ANALYSIS and videotape. Initially, colour in order to study the movement

2.

Mechanical (a) Ofthe

disorders spine

Clinical videotapes

photography were used

Scoliosis Ankylosing Iliac crest

anterior

Ii spondylitis defect after

spinal fusion

3 20

patterns by single-frame analysis. As we became more experienced, the responses were recorded on 35 mm colour projection slides. Electromyography. This was performed on three normal volunteers. Gluteus maximus, gluteus medius, gluteus minimus, activities sequently to occur walking matic tensor were with regularly subjects fascia recorded the pelvis lata with and the adductor subject magnus at rest positions muscle and subfound Three to trauC

(b) Ofthe hip (i) In children

Congenital dislocation Subluxation Coxa vara Perthes disease Slipped capital femoral Fractured neck of femur After arthroplasty Osteoarthritis Avascular necrosis congenital dislocation ofthe

.5 .5

5
epiphysis 4 15 I5 S 4 13)

(ii)

In adults

in the different

when the test was performed. with incomplete paraplegia due were also studied. With Council

Including

one

with

hip (Fig.

fracture-dislocation

there the subject ofthe hip was grading of side with the
Response Response

was (Fig.

a compensatory 2).
2. The 3. The

scoliosis

convex

to the

stance ground

Assessment of abductor muscle power. lying on one side, the strength ofabduction assessed, muscle using power. the Medical Research

minimal buttock with

pelvis remained spinal compensation

parallel to the (Fig. 3).

crease

pelvis dropped on the non-stance sic/c and moved downwards. This was associof the convex weight-bearing to the non-stance hip and side, a as

RESULTS Normal responses hip flexed at 30

ated

adduction scoliosis

compensatory

With the non-stance

. Three

different with side,

patthe the test;

terns of pelvic and spinal movement non-stance hip flexed to 30

Response

occurred

seen in the classical positive Trendelenburg test. Balance was achieved by moving the torso and centre of gravity directly over the weight-bearing hip (Fig. 4).

typical

1. The description

pelvis rose on the non-stance of a negative Trendelenburg

With
position

the

non-stance the pelvis

hip flexed rose

90

Response

1. In but not

this as

on the non-stance

side

Fig.

Fig.

Fig.

Three different responses were seen in norma/ volunteers when they were asked to stand on one leg with the hip on the non-stance side flexed at about 30 . Figure 2-The pelvis on the nonstance side rose above the stance side with the trunk centred over the stance hip (a negative response). Figure 3-The pelvis remained parallel to the ground. Figure 4-The pelvis on the non-stance side dropped below the level ofthe stance side (a positive Trendelenburg sign).

THE

JOURNAL

OF

BONE

AND

JOINT

SURGERY

THE

SIGNIFICANCE

OF

THE

TRENDELENBURG

TEST

743

high as when was because to the rib cage able.

the non-stance pelvic rotation and made spinal

hip was brought

flexed only 30. This the iliac crest closer uncomfort-

compensation

Response 2. The There was

90 -in no

pelvis remained parallel to the ground. no third response with the hip flexed did the findings. pelvis drop on the non-stance were

subject

side. Electromyographic obtained the third

present

Electromyograms It was test) gluteus found

from three volunteers. response (false-positive in gluteus maximus, or fascia adductor magnus; lata. If the subject, the pelvis the

that with no activity was medius, gluteus

minimus tensor tarily gradual reaching achieved; or and

activity was present in on command, volunside, there was a

raised

on the non-stance

increase of a maximum no activity maximus. activity was the non-stance

hip abductor muscle activity, when the first response was was recorded in adductor magnus With the the same non-stance as during hip at
9#{216}c

gluteus 2 with

abductor

Responses

hip at 30.

Figure 5-A false-negative test can occur if the trunk shifts excessively to reduce the abductor muscle force necessary to obtain equilibrium: note the wasting of the right thigh due to poliomyelitis. Figure 6When the trunk is re-aligned over the stance hip the pelvis drops rapidly on the non-stance side because there is not enough muscle power available to maintain the elevated position.

Responses Recording children conditions abnormal response standard was that Ia/se
Table Causes Use Use

in abnormal subjects and classifying the and adults with proved difficult subjects before patterns in normal way of performing some people with Trendelenburg
responses negatives muscles and rectus femoris of trunk the hip as

findings

.in

a group of different we examined the variable not defined

of

were

able

to stand

on the painful

hip for only

a very

short

a wide at first. we had people,

spectrum Because clarified we had What

period. In the absence of significant pain, subjects with weak hip abductor muscles could achieve balance by lifting a the pelvis on the non-stance side, using functioning muscles, kinetic energy Electromyography who could elevate the either groups leaning above well over the pelvis; and gravity. was performed on three patients non-stance side of the pelvis by the hip minimal joint or electrical using muscle activity was

the test. pathological test (Table

test

became disorders

clear had a

negative
II. False of false ofsuprapelvic ofpsoas

II).

to Trendelenburg

Causes Pain Poor Lack

of false

positives

recorded in the abductor muscle groups. Thus, simply looking at the final position of the trans-iliac line of the pelvis gives no assessment ofabductor muscle function. It was our experience could not be performed. test could that children under assessed, the age of four years reliably however

balance ofco-operation

or

Wide lateral translocation to allow balance over a fulcrum

understanding Costo-pelvic impingement

the test was Trendelenburg

In children over four, the be assessed only if the children fully; unless this was and different responses at short time intersign. Several people had

This patients could this, bearing activity raise first,

false with

negative neurological

was

particularly many

evident of whom

in

could understand and co-operate possible, assessment was invalid, were seen at separate examinations vals on the same positive day. Trendelenburg Delayed

disorders, on the non-stance the torso well

the pelvis by moving to

side. They did over the weight-

hip, thereby necessary

reducing maintain

the

amount of abductor this posture (Fig. 5).

an initial negative with the non-stance began to fall and initial posture. Trendelenburg time at which be recorded. Trendelenburg they are asked they fatigue

test. but after standing for a short time side of the pelvis raised, it gradually they were not able to maintain their has been Mitchell pelvis people called (1973) began with a a delayed suggested to drop delayed positive that the should positive when that

Secondly. by supporting the non-stance side, they tion of especially muscle contraction latissimus dorsi, side, side, that

the hand on a table or wall on were then able, by a combinaof the shoulder adductors, and (possibly) psoas major on

This sign; the In

the non-stance weight-bearing pelvis above also found the centre the abductors.
VOL. 67- B. No.

with quadratus lumborum on the to raise the non-stance side of the weight-bearing hip (Fig. 6). We

sign the gait to walk quickly, easily and a limp,

can be normal, but it becomes apparent with all the characteristics gait, becomes scoliosis had between

of the

that people with pain in the hip tended to shift of gravity over the hip to decrease the pull of Other
S. NOVEMBER

of the so-called Trendelenburg Two subjects with severe responses because of

obvious. false-positive the lower

patients
1985

with

a painful

hip or spine

impingement

744

P. M. HARDCASTLE,

S. M. L. NADE

Fig.

Fig.

Fig.

Fig.

10

Fig.

II

Fig.

12

The suggested method ofperforming the Trendelenburg test: Figure 7-The examiner stands behind the patient. Figure 8-The patient is asked to raise one leg off the ground with the hip flexed between 0 and 30 and to balance herself. Figure 9-The patient is asked to raise the non-stance side as high as possible. Figure lO-The examiner may support the patient by holding the arm on the other side. Figure I I-The examiner may Pzot support the non-stance arm as this may act as a fulcrum for latissimus dorsi. quadratus lumborum, and the paraspinal muscles. Figure 1 2--If there is significant trunk shift to the stance side this is corrected by gently aligning the trunk over the stance-side hip and watching the relation of the pelvis to the ground.

costal elevate A

margin

and

the

iliac

crest

when

they

attempted

to

weight-bearing

hip,

the

examiner

corrects

this

by

the pelvis. STANDARD TRENDELENBURG TEST a test;

gentle pressure on the shoulders prominens approximately over joint and the weight-bearing foot we have formulated the Trendelenburg be interpreted

to bring the vertebra the centre of the hip (Fig. I 2).

As a result of our observations, standard method of performing if this cally take is used meaningful time, and the response can way. To its accurate and

Interpretation
(a)

in a clini-

The

response

is NORMAL

(i.e.

the

test

is nega-

perform the assessment ofthe

test properly does demands the full patient.

understanding How 1
.

co-operation

tive) if the pelvis on the non-stance as high as hip abduction on the and providing this posture can seconds with the vertebra prominens and foot. (b) positive) responses side above
maximal.

side can be elevated stance side will allow, be maintained for 30 centred over the hip (i.e. the test is This includes

to do the test the patient and (the line joining 7). from holding the ground the The hip knee observes the iliac the foot joint should of the rectus at

The examiner stands behind the angle between the pelvis the ground (Fig. is asked to raise not being and tested, 30

The if

response this

cannot pelvis side,

is ABNORMAL be done. is elevated but where

crests) and 2. The patient of the between side

where the the stance

on the non-stance this elevation is not

neutral

of flexion.

be flexed enough ground in order

to allow the foot to be clear to nullify the effect of the

can that pelvis ment,

(c) The be lifted position

response is also ABNORMAL if the pelvis on command, but can not be maintained in for 30 seconds. The time taken before the a time elerecorded

femoris muscle. The position of the pelvis is again noted (Fig. 8). A supporting stick can be used in the hand only on the side of the weight-bearing hip; alternatively.
examiner

starts to fall is recorded. By introducing the Trendelenburg test can be objectively purposes. constitutes In the Obviously a zero time presence

(Fig. 3. Once

so 12). balanced,

both shoulders as to maintain the patient

can be supported balance without is then asked

by the a stick the

for comparison described in (b) test. Non-valid or of because responses.

the response Trendelenburg or leg pain

of back

to raise

non-stance side of the pelvis as high as possible (Fig. 9). The examiner may support the patient by holding the arm on the stance side (Fig. 10; compare
4.

deformity. or if the patient is ofage or mental status, inappropriate

uncooperative responses

withFig. If the

II). patient

leans

too

far

over

to

the

side

of

the

may arise (Table II). An abnormal response (positive test) in these circumstances can be misleading. However, ifthe test is negative that is significant-it means that the subject does not have abnormal hip mechanics.
THE JOURNAL OF BONE AND JOINT SURGERY

THE

SIGNIFICANCE

OF THE

TRENDELENBURG

TEST

745 respQnse when examined the in

Significance Neurological strength

subjects

disorders. was
whose

MRC

Grade

5 abductor

muscle

arthroplasty three years Subluxating

importance

had a normal after surgery.

of

required to produce a normal response. All hip abductor power was Grade 4 or less responses at patients in this not to the full times group extent. between 0 and 25 were able to elevate We considered this

hips. Mitchell thedelayed

(1973) has described (timed) Trendelenburg

test

had abnormal seconds. Some the pelvis, but

assessing clinical abular dysplasia.

deterioration in adolescents with Our two patients with subluxating tests, or negative one

acethips at IS

to be a positive Trendelenburg test at 0 seconds. One subject who had Grade 5 strength on clinical testing had a delayed (or timed) positive Trendelenburg test at 1 5 seconds. Only one subject with Grade 3 hip abductors had an initial negative response which became positive good at 5 seconds rectus femoris (a delayed muscle isolated responses. which nerve positive probably root test). entrapment He helped. had The of L5 a

both had positive (timed) Trendelenburg and one at 20 seconds (Fig. 14). (oxa vara. The test may be positive ing on the femoral degenerative changes. 100 the response can of 90 after a varus capital by the seconds. Slippedjemoral was not altered to the negative disease. size relation all had Perthes by the
incongruity

depend-

neck angle and the presence of With femoral neck angles of up to be normal. One child with an angle osteotomy epiphrsis. rotation had The a positive Trendelenburg femoral head examined not altered test at 20 test in

two subjects with or SI had negative

of the

Mechanical disorders. Congenital dislocation of the hip. The Trendelenburg test was always positive in subjects with congenital dislocation of the hip (Fig. 13). One patient who had been treated by bilateral replacement

femoral neck and the subjects responses if they were painfree. The Trendelenburg test was head. However, the test can response responses was

of the femoral or hinge abduction positive Variable

if there is become posiin one observed subbut

tive, and ject(Fig. Arthritis the

when

a delayed

seen

15).
ofthe hip. were not vary in the same individual times. Obviously pain or probe expected to reduction to alter the of time of patients

type of response did studied at different of the disease particularly response. inequality

gression response, the normal Leg length with line

tests

would in respect qfter

hip arthroplastv.

Two

up to 2 cm ofshortening after hip replacement

(i.e. normal responses).

above the intertrochanteric had negative Trendelenburg trochanter qfter the Trendelenburg hip arthroplastr. test was parthe a

A iulsion

In the ticularly osteotomy

Fig. 13

of the greater absence of pain,

valuable some time after operation. Where gap was greater than 2 cm the Trendelenburg positive either immediately (zero time) or with

test

was

A 48-year-old woman with congenital dislocation of the hip showing a positive Trendelenburg response despite supporting her body weight with her hands.

delayed positive response. Fractured neck of/emur. Patients tion with Ender or Zickel nails

who had unstable fixahad a positive response

Fig. Anteroposterior hip. His initial

14 a subluxating but within 15

Fig.

15

radiograph of a nine-year-old boy with Trendelenburg response was negative seconds he had a delayed positive test.

Anteroposterior radiograph ofa 10-year-old boy with coxa magna and hip joint incongruity as a sequel of Perthes disease. He had a delayed positive test at 20 seconds.

VOL.

67-B.

No.

5, NOVEMBER

1985

746 until were the fracture tested initially pain at which positive was united 8 to radiologically. 10 weeks after

P. H. HARDCASTLE,

S. M. L. NADE

These operation

people and

pelvis vis,

to elevate or shifts the

the torso called noted the

non-weight-bearing well over the trick in some test

side weight-bearing

of the

pelside;

had no response, also had

the time of was positive.

their initial Trendelenburg Two nialunited fractures

these can be responses were years of age, and

movements. patients less

Variable than seven under

responses. A vascular necrosis of the jemoral head. Hip pain made proper assessment of these cases difficult. However, one ofour four patients had only mild symptoms; his test was
positive at 20 seconds

is of no value

in children

on the first In the

examination, other patients

and pain

at 25 pre-

seconds vented Spinal

losing

on the adequate disorders.

spondylitis,

next day. assessment.

four. Nevertheless, if the Trendelenburg test is carefully performed, it is an accurate clinical sign with prognostic implications. Inman (1947) measured the torque strength about the hip with the pelvis in different postures with respect to the ground. Our electromyographic results confirm his is pelvis of is findings necessary dropped side. As abductor centred that little abductor muscle strength/activity to maintain a balanced posture with the (as in Response 3) on the the pelvis rises on this side muscle activity provided

Stiffliess. Totally did not affect of the hip

stiff spines, as in ankythe test unless there was spinal deformity.

also

abnormality

or gross

non-weight-bearing there is increase that the torso

over the hip. Functional assessment of a joint is important clinical assessment of patients. Observation of probably
Fig. 16 This 43-year-old ladys hip abductors had normal strength. However, she was unable to raise her pelvis on the left above the horizontal because of impingement between the iliac crest and costal margin as the result of severe scoliosis.

in the gait is because

performed of space. assessment clinical sign

less

often

than

is desirable

of limitation functional valuable

The Trendelenburg in a confined space, than many static

test allows for and is a more It can an abnormal in this becomes is not diffialso

tests. has

be easily recorded on film or videotape. It is our belief that a patient who response to the Trendelenburg paper has an inefficient gait,
easily fatigued.

test and

as described therefore the test

With

a little

practice,

cult to perform and interpret. Timing is an essential part of the test; it provides an objective measure of improvement or deterioration in the neuromuscular or mechanical function of the hip.
Deformiti. scoliosis,

Kyphosis
however,

may

lower
positive

costal
test
Nerve

margin (Fig.
root

not affect the outcome. Severe lead to impingement between the and the iliac crest and give a falsecan

did

Trendelenburgs original observations and clear, and his interpretations accurate. the need for meticulous clinical examination provide correct the standardised ment offunction

were precise We Support in order to

16).
irritation

Pain.
results, responses.

but

back

pain

itself

did

lead to not lead

false-positive to abnormal

diagnosis and we recommend the use of timed Trendelenburg test in the assessand malfunction ofthe hip.

DISCUSSION When
routine same response

asked side,
is

to stand hip test

normal;

on may

one

leg,

and to

follow contraction the

a standard on the one An standardised

to ensure

abductor

muscle respond

Conclusions I The Trendelenburg test is a useful part of clinical examination if performed and interpreted correctly. We have described a standard method for performing
.

patients

Trendelenburg ability and to assume

in one of three ways. Only the other two are abnormal. normal on response must be absolute, the non-stance test is positive.
of the

2.

the test. False-positive occur,

test

and their
performed.

false-negative can performing of

responses be clarified the test

may if the is essenabnor-

but
is properly

interpretation when measurement

the

if the pelvis drops seconds the Trendelenburg

timer is an essential part

side within 30 The use of a test, severity

balance

3. The tial, mal

use

of a timer

indeed, altered
either patient cannot negative

makes

it an

objective

Trendelenburg measure of
poor

and, of and

and allows response.

a delayed

hip mechanics. However, the lack

can

REFERENCES

presence

of pain,

Inman

be
tests

of co-operation or lead to false-positive properly performed.

is that

understanding tests, because The reason uses muscles

by the for above

the test falsethe

VI. Functional aspects of the abductor Bonefoint Surg[Br] l947;29:607-19. delayed of Trendelenburg orthopaedics. hip test.

muscles

of the Congr Ser

hip. 1973;

Mitchell GP. The 291: 1113. Rang M. Antho/ogy 1966; 139-43.

mt

Edinburgh:

E&S

Livingstone,

the

subject

THE JOURNAL

OF BONE AND

JOINT

SURGERY