BY B. GUNNAR WALLIN*, DAVID BURKE AND S. C. GANDEVIA

Journal

of

Physiology

(1992),

455,

pp.

219-233

219

With

8

figures

Printed

in

Great

Britain

COHERENCE

BETWEEN

THE

SYMPATHETIC

DRIVES

TO

RELAXED

AND

CONTRACTING

MUSCLES

OF

DIFFERENT

LIMBS

OF

HUMAN

SUBJECTS

BY

B.

GUNNAR

WALLIN*,

DAVID

BURKE

AND

S.

C.

GANDEVIA

From

the

Department

of

Clinical

Neurophysiology,

The

Prince

Henry

and

Prince

of

Wales

Hospitals,

and

Prince

of

Wales

Medical

Research

Institute,

University

of

New

South

Wales,

Sydney

2036,

Australia

(Received

30

September

1991)

SUMMARY

1.

This

study

was

undertaken

to

quantify

the

simultaneous

sympathetic

drives

to

muscles

in

the

two

legs

of

human

subjects,

and

to

elucidate

the

extent

to

which

a

common

drive

determines

sympathetic

outflow

to

different

limbs

at

rest,

during

apnoea

and

during

voluntary

contractions.

2.

Sympathetic

efferent

activity

was

recorded

simultaneously

from

fascicles

of

both

peroneal

nerves,

innervating

the

pretibial

flexor

muscles.

At

rest

the

similarity

was

quantified

for

a

sample

of

records

by

manual

measurement

of

equivalent

bursts

in

the

two

recordings,

and

for

all

records

by

cross-correlation

and

power

spectral

analysis

of

the

two

recordings.

During

contractions,

only

the

latter

method

was

used.

3.

At

rest

the

correlation

coefficient

for

the

relationship

between

the

burst

amplitudes

for

the

two

recordings

was

0-72

(S.D.

0-1).

For

the

same

sequences,

the

computed

coherence

between

the

two

recordings

was

85-6

%

(S.D.

67

%)

at

the

cardiac

period.

There

was

a

statistically

significant

linear

relationship

between

these

two

measures

of

similarity,

and

this

was

stronger

when

data

from

sequences

recorded

during

apnoea

were

included

in

the

analysis.

At

rest

the

mean

difference

in

coherence

between

consecutive

sequences

with

no

intervening

manoeuvre

(apnoea,

contraction,

change

in

recording

site)

was

4-2

%

(S.D.

4.3

%).

In

only

two

of

forty-nine

such

instances

was

the

difference

in

coherence

>

10

%.

4.

Apnoea

at

end-expiration

increased

the

amplitude

and

frequency

of

sym-

pathetic

bursts

and

increased

the

similarity

between

the

two

recordings.

The

correlation

coefficients

increased

from

a

mean

of

0-72

at

rest

to 0-89

during

apnoea.

Coherence

increased

from

a

mean

of

82-1

%

at

rest

to

91-9

%

during

apnoea.

5.

On

the

right

side,

graded

voluntary

contractions

were

performed

at

5,

10,

20

or

30

%

maximal

force

using

the

muscle

innervated

by

the

fascicle

from

which

the

recording

was

made.

The

coherence

between

the

recordings

made

from

the

right

and

left

legs

decreased

by

>

10%

at

each

contraction

level.

Pooling

the

data

for

all

contractions,

there

was

a

significant

decrease

in

power

at

the

cardiac

frequency

in

the

*

Permanent

address:

Department

of

Clinical

Neurophysiology,

Sahlgrenska

Sjukhuset,

Goteborg,

Sweden.

MS

9780

B.

G.

WALLIN,

D.

BURKE

AND

S.

C.

GANDEVIA

sympathetic

recording

from

the

contracting

leg.

Contraction

of

a

synergist

or

antagonist

at

10%

maximum

produced

negligible

changes

in

coherence.

6.

It

is

concluded

that,

at

rest,

homologous

muscles

of

the

lower

limbs

are

subject

to

a

common

drive

and

that,

during

apnoea,

this

common

drive

can

dominate

the

sympathetic

outflow

to

the

virtual

exclusion

of

regional

drives.

During

voluntary

activity,

the

importance

of

this

common

drive

is

lessened,

presumably

because

of

regionally

specific

changes

involving

the

contracting

muscle.

The

data

suggest

that,

at

these

relatively

weak

contraction

levels,

voluntary

contraction

leads

to

a

reduction

in

sympathetic

outflow

to

the

contracting

muscle.

INTRODUCTION

At

rest,

there

is

a

remarkable

similarity

between

sympathetic

outflows

to

muscles

in

different

limbs

(Sundlbf

&

Wallin,

1977)

and

the

extent

of

the

similarity

suggests

that

muscle

sympathetic

activity

(MSA)

is

dominated

by

a

common

central

drive.

Differences

between

sympathetic

drives

to

arm

and

leg

muscles

have

been

demonstrated

during

mental

arithmetic

(Anderson,

Wallin

&

Mark,

1987)

and

post-

contraction

muscle

ischaemia

(Wallin,

Victor

&

Mark,

1989).

During

the

isometric

contraction

preceding

muscle

ischaemia

there

was

no

difference

between

sympathetic

outflows

to

resting

muscles

in

the

opposite

arm

and

the

leg

(Wallin

et

al.

1989),

but

no

such

recordings

have

been

made

from

nerves

innervating

contracting

muscles.

However,

there

is

indirect

evidence

that,

at

least

under

some

conditions,

there

may

be

increased

MSA

directed

to

exercising

muscle.

Strong

dynamic

contractions

of

thigh

muscles

to

50-100%

of

maximum

result

in

increased

noradrenaline

spillover

to

blood

from

the

working

muscle

(e.g.

Savard,

Strange,

Kiens,

Richter,

Christensen

&

Saltin,

1987),

suggesting

an

increase

in

sympathetically

mediated

vasoconstrictor

tone.

Similarly,

studies

in

intact

rats

and

following

acute

sympathectomy

suggest

that,

during

locomotion,

hindlimb

blood

flow

and

the

flow

to

individual

exercising

muscles

are

under

active

sympathetic

control

(Peterson,

Armstrong

&

Laughlin,

1988).

The

present

study

was

undertaken

to

quantify

the

degree

of

homogeneity

in

MSA

direct

to

muscles

of

both

legs

at

rest,

during

apnoea

and

during

isometric

voluntary

contractions

of

muscles

in

one

leg.

By

using

the

non-contracting

leg

as

a

control

it

was

possible

to

separate

regionally

specific

changes

in

MSA

from

generalized

changes.

METHODS

Thirteen

experiments

were

performed

on

twelve

normal

volunteers

(seven

male,

five

female,

aged

20-34

years)

with

their

informed

consent

and

with

the

approval

of

the

local

ethics

committee.

Successful

bilateral

recordings

of

sympathetic

activity

were

obtained

in

ten

experiments;

no

data

came

from

the

other

three

experiments.

The

subjects

lay

supine

on

a

comfortable

bed,

with

the

upper

body

supported

by

pillows,

the

left

leg

unrestrained

and

the

right

leg

secured

in

an

isometric

myograph.

At

the

start

of

each

experiment

maximal

dorsiflexion

and

plantar

flexion

forces

were

measured,

and

the

subject

was

trained

to

contract

tibialis

anterior,

extensor

digitorum

longus

or

triceps

surae

on

the

right

side.

The

electromyogram

(EMG)

of

appropriate

muscles

bilaterally

and

the

electrocardiogram

(ECG)

were

recorded

using

surface

electrodes.

Neural

recordings.

Recordings

of

MSA

were

made

bilaterally

using

tungsten

microelectrodes

introduced

manually

through

the

skin

into

the

right

and

left

peroneal

nerves

at

the

fibular

head

level.

Details

of

the

technique

and

evidence

for

the

sympathetic

nature

of

the

activity

have

been

220

SYMPATHETIC

ACTIVITY

DURING

CONTRACTION

described

elsewhere

(Sundl6f

&

Wallin,

1977;

Vallbo,

Hagbarth,

Torebj6rk

&

Wallin,

1979).

Fascicles

innervating

tibialis

anterior

were

sought

using

the

response

to

electrical

stimuli

delivered

through

the

microelectrode

as

a

guide.

In

nine

of

ten

experiments,

the

fascicle

from

which

MSA

was

recorded

on

the

right

side

innervated

tibialis

anterior;

in

the

tenth,

the

innervated

muscle

was

extensor

digitorum

longus.

When

the

tip

of

the

electrode

had

penetrated

the

desired

fascicle,

its

location

was

adjusted

carefully

under

auditory

control

to

bring

the

pulse-synchronous

bursts

of

MSA

into

focus.

On

request

subjects

remained

completely

relaxed,

held

their

breath

at

end-

expiration

('apnoea')

or

maintained

a

contraction

of

the

nominated

muscle

in

the

right

leg

for

up

to

6

min

at

a

specified

target

force

(5,

10,

20

or

30

%

of

maximal

voluntary

contraction,

MVC).

All

experimental

data

were

stored

on

tape

for

subsequent

analysis.

Sympathetic

activity

was

amplified

(x

10000-20000),

filtered

(bandpass

300

Hz-3

kHz)

and

stored

on

tape.

On

replay,

it

was

refiltered

prior

to

integration

using

a

RC

low-pass

filter

(time

constant

0-1

s).

This

produced

an

analog

signal

that

could

be

displayed

on

an

ink-jet

recorder

or

an

electrostatic

printer,

or

digitized

on

an

IBM-compatible

PC.

Sequences

were

rejected

when

contaminated

by

the

electromyogram

of

nearby

contracting

muscles.

Manual

quantification.

For

seven

subjects,

data

from

four

randomly

selected

2

min

sequences

when

the

subject

was

at

rest

and

one

when

apnoeic

were

replayed

on

an

ink-jet

recorder,

with

amplification

adjusted

such

that

the

sizes

of

corresponding

bursts

in

the

two

recordings

were

similar.

The

amplitudes

of

each

burst

of

sympathetic

activity

in

the

two

recordings

were

measured

for

the

five

sequences

for

the

seven

subjects,

expressing

each

burst

as

a

percentage

of

the

largest

burst

in

that

sequence.

The

amplitude

of

a

burst

in

the

right

peroneal

recording

was

then

plotted

against

the

amplitude

of

the

corresponding

burst

in

the

left

peroneal

recording

for

each

sequence.

Regression

analysis

was

performed

on

these

data,

and

the

correlation

coefficient

taken

as

an

index

of

the

degree

of

homology

of

the

two

recordings

(see

Wallin

et

al.

1989).

Computer

analysis.

The

integrated

neurograms

of

the

right

and

left

peroneal

nerve

sympathetic

activity

were

digitized

in

sequences

of

2

min

duration

using

a

sampling

rate

of

0-1

kHz.

Auto-

and

cross-correlograms

were

derived

from

the

digitized

data,

and

power-spectral

analysis

then

performed

on

these

correlograms

(see

Fig.

2).

The

cardiac

period

was

identified

from

the

ECG

recording

and

confirmed

in

the

auto-correlograms.

The

power

in

each

recording

was

measured

at

the

cardiac

period.

From

the

cross-correlated

data

the

coherence

between

the

two

recordings

and

the

gain

of

the

relationship

between

the

two

were

measured

at

the

cardiac

period

(assuming

that

the

left

side

was

the

input

to

and

the

right

the

output

from

the

cross-correlogram).

Gain

was

computed

as

cross-spectral

power

divided

by

the

input

(left

side)

spectral

power.

Coherence

was

calculated

as

the

square

of

the

cross-spectral

power

divided

by

the

product

of

the

input

and

output

powers.

Because

these

measurements

were

made

at

the

cardiac

period,

only

noise

at

the

cardiac

period

could

have

affected

the

measurements.

This

assumption

was

confirmed

in

control

studies

in

which

a

neurogram

was

correlated

with

itself

after

noise

was

added

to

one

input.

The

added

noise

was

bandpass-filtered

white

noise

(300

Hz-3

kHz).

In

different

control

runs

this

added

signal

was

held

constant,

varied

randomly

or

varied

in

time with

the

ECG.

Only

the

latter

decreased

the

coherence

between

the

two

signals

by

>

5

%.

This

finding

implies

that

increases

in

non-sympathetic

neural

traffic

during

a

contraction

(or

undetected

contamination

of

the

neural

recording

by

EMG

activity

would

affect

the

measurements

only

if

they

possessed

a

cardiac

periodicity.

Although

some

non-sympathetic

neural

activity

does

have

a

cardiac

periodicity

(McKeon

&

Burke,

1981),

such

contamination

would

increase

the

power

in

the

relevant

recording:

in

the

studies

reported

here,

contraction

was

associated

with

a

decrease

in

power

at

the

cardiac

frequency.

In

similar

controls,

changing

the

amplification

of

one

input,

slowly

or

abruptly,

produced

changes

in

coherence

of

<

5

%

provided

that

the

gain

change

was

not

locked

to

the

cardiac

period.

This

finding

implies

that

coherence

would

change

minimally

with

minor

disturbances

to

electrode

position,

were

this

to

occur

on

the

right

side

during

contraction

sequences.

Such

changes

would,

however;

appear

as

a

change

in

gain

between

the

two

recordings,

much

as

is

shown

in

Fig.

3.

RESULTS

Satisfactory

bilateral

recordings

of

sympathetic

efferent

activity

to

pretibial

muscles

were

obtained

in

ten

experiments

on

nine

subjects.

In

all

recordings

there

was,

at

rest,

a

striking

similarity

in

the

pattern

and

amplitude

of

the

pulse-

synchronous

bursts

of

sympathetic

activity

destined

for

the

muscles

of

opposite

legs

221

B.

G.

WALLIN,

D.

BURKE

AND

S.

C.

GANDEVIA

Rest

5s

Apnoea

Fig.

1.

Bilateral

recordings

of

muscle

nerve

sympathetic

activity

from

the

right

and

left

peroneal

nerves

showing

the

similarity

of

the

integrated

neurograms.

In

each

panel

the

upper

trace

is

the

neurogram

from

the

right

leg,

the

second

trace

the

neurogram

from

the

left

leg,

and

the

third

trace

ECG.

Bursts

which

appear

only

on

one

side

are

indicated

by

asterisks.

In

the

lower

panel,

the

subject

was

apnoeic

for

the

duration

indicated

by

the

horizontal

bar.

A

8

100

10

a)~

~

~~~ne

0

oo

O

C

D

100-

100-

Apnoea

OR~~~~~~~~~~~~~~~~~~~a

g0

0

a

T

ApeApnoea

.50

0L50

o

0

Rest0

est

0

1-5

3

0

0

1-5

3.0

Frequency

(Hz)

Frequency

(Hz)

Fig.

2.

Computer

analysis

of

2

min

recordings,

samples

of

which

are

illustrated

in

Fig.

1.

Apnoea

resulted

in

a

marked

increase

in

power

bilaterally

but

only

slight

changes

in

gain

and

coherence.

During

apnoea

the

heart

rate

increased

from

092

to

1-02

Hz.

222

SYMPATHETIC

ACTIVITY

DURING

CONTRACTION

223

(Fig.

1,

upper

panel).

There

were,

however,

small

bursts

that

could

be

identified

only

on

one

side

(indicated

by

asterisks

in

Fig.

1),

and

often

the

amplitude

of

a

single

burst

did

not

have

the

same

relationship

to

those

of

other

bursts

in

the

corresponding

sequence

on

the

other

side.

For

the

subject

illustrated

in

Fig.

1,

the

correlation

New

site

I'

!

,

o.

.

,

.

Nt

''

I

t

I

.

I

t

I

.

'

1

/

.

1:

|

__.

_

TA

TA

*10%'

20%

if

\

21

.d

I

\,

C)

1'N

!.

'

.

.

:

!:

TA

Tri

EDL.

30%

10%

1

o%

,/

'

'a

Fig.

3.

The

effect

of

different

manoeuvres

on

coherence

and

gain

plotted

sequentially

throughout

an

experiment.

*

represent

values

obtained

at

rest,

*

represent

values

obtained

when

apnoea

was

performed

by

an

otherwise

resting

subject,

0

represent

values

obtained

during

contraction,

and

[1

represent

values

obtained

when

the

subject

was

apnoeic

during

the

contraction.

The

resting

values

for

coherence

changed

little

during

the

experiment,

as

indicated

by

the

error

bars

on

either

side

of

the

plot

(±

1

S.D.).

Coherence

changed

little

with

contractions

of

right

tibialis

anterior

(TA),

right

triceps

surae

(Tri)

or

right

extensor

digitorum

longus

(EDL)

at

10%.

There

were

significant

decreases

in

coherence

with

contractions

of

right

TA

at

20

and

30%

MVC.

In

the

initial

part

of

the

experiment

there

is

a

steady

increase

in

gain,

due

to

progressive

dislodgement

of

the

microelectrode

on

the

left

side.

Despite

this,

coherence

values

at

rest

remained

quite

reproducible.

At

the

dotted

line

the

experiment

was

interrupted

until

a

stable

recording

was

obtained

bilaterally.

coefficient

between

corresponding

bursts

on

the

two

sides

was

083

at

rest

and

0-92

during

apnoea.

Figure

2

illustrates

the

power

spectral

analyses

for

these

two

recordings.

The

computed

coherences

for

these

two

recordings

were

92-4

and

96-4

%,

at

the

cardiac

frequencies

092

Hz

and

1P02

Hz,

respectively.

2.0

r

15

1-

c

U7

10

1

05

F

O

L

100

r

I

80

60

I-

0)

O

c

0)

0)

0

U

40

_

20

F

0

I

B.

G.

WALLIN,

D.

BURKE

AND

S.

C.

GANDEVIA

Homogeneity

of

the

sympathetic

drive

to

muscles

of

different

limbs

at

rest

In

each

experiment

the

coherence

between

the

two

recordings

was

computed

for

the

subject

at

rest,

over

five

to

fifteen

2

min

sequences

(average

10-2

sequences/experiment).

The

degree

of

coherence

between

the

two

recordings

varied

1-0r

08

F

061-

._

0-41

02

I

0

100

.XII

TA

.

10%,

,19.

S

I

z

1:

./

90

I

a,

1-.

a,

a,

0

u

80

F

701-

New

site

*

*

*

|

TA

TA

Tri

*

EDL

t

5%

20%

10% 10%

s

r

o

I!

\

\

I

,

,

60

F

50

Fig.

4.

The

evolution

of

coherence

and

gain

values

throughout

an

experiment

in

a

second

subject.

The

symbols

are

as

in

Fig.

3.

The

reproducibility

of

the

coherence

when

at

rest

is

indicated

by

the

error

bars

on

either

side

of

the

coherence

plot

(±1

S.D.).

Coherence

and

gain

decreased

in

the

three

contractions

of

right

TA

but

changed

little

in

the

contractions

of

right

triceps

surae

and

EDL.

The

recording

site

on

the

right

side

was

lost

abruptly

at

the

vertical

arrow,

so

that

the

subsequent

recordings

had

a

different

gain.

However,

the

coherence

values

at

rest

were

quite

reproducible.

throughout

an

experiment.

Some

of

this

variability

appeared

to

be

related

to

a

preceding

manoeuvre,

particularly

apnoea,

following

which

return

to

a

truly

basal

state

took

several

minutes

(Figs

3

and

4).

In

the

ten

experiments

the

average

coherence

of

all

sequences

at

rest

was

65A4-92-2

%

(mean

82-1

%,

S.D.

7

9

%).

In

some

subjects

the

coherence

values

for

all

rest

sequences

in

the

experiment

had

S.D.S

of

2-4

%

despite

many

manoeuvres.

The

experiments

with

the

highest

variability

had

S.D.S

of

9

5

and

12-8%,

respectively.

224

SYMPATHETIC

ACTIVITY

DURING

CONTRACTION

The

difference

in

coherence

between

consecutive

rest

sequences

with

no

intervening

apnoea,

contraction

or

change

in

recording

site

was

determined

to

quantify

the

reproducibility

of

coherence

measurements

in

the

absence

of

these

manoeuvres.

The

mean

difference

in

coherence

in

forty-nine

consecutive

pairs

of

measurements

was

100-

50

.

0~~~~~~~~~4

-D

0

0

50-

0-25

0-5

0-75

10

Correlation

coefficient

Fig.

5.

Relationship

between

coherence

and

correlation

coefficient

for

seven

subjects.

The

values

for

sequences

when

the

subject

was

at

rest

are

indicated

by

*.

The

values

for

the

same

subjects

when

apnoeic

are

shown

by

0.

The

line

represents

the

regression

line

for

all

data

(r

=

0-60,

P

<

0-01).

4-2

%

(S.D.

4-3%).

In

nine

of

the

ten

experiments,

the

mean

differences

were

0

1-6

5

%;

in

the

tenth

experiment,

one

of

only

three

values

was

25-2

%.

Of

the

forty-

nine

consecutive

differences,

all

but

two

were

<

10%.

For

seven

subjects,

the

amplitudes

of

corresponding

bursts

were

measured

manually

for

four

sequences

at

rest,

and

linear

regression

was

performed

on

these

measurements

to

provide

correlation

coefficients

as

a

measure

of

the

similarity

of

the

activity

on

the

two

sides.

The

mean

correlation

coefficient

of

the

four

rest

sequences

for

the

seven

subjects

was

0-72

(S.D.

0410)

indicating

that,

in

these

sequences,

52

%

of

the

variance

in

the

two

recordings

could

be

attributed

to

common

drives.

For

these

same

twenty-eight

sequences,

the

mean

coherence

was

85-6

%

(S.D.

6-7

%).

The

sequences

analysed

for

each

subject

were

usually

taken

from

different

stages

of

the

experiment.

However,

there

were

seven

pairs

of

consecutive

rest

sequences,

and

the

mean

difference

in

correlation

coefficient

for

consecutive

sequences

was

0-08.

The

mean

difference

in

coherence

for

the

same

seven

pairs

of

analyses

was

5-7

%.

When

the

data

for

the

seven

subjects

were

pooled,

there

was

a

significant

linear

relationship

between

coherence

and

correlation

coefficient

for

the

same

sequences

at

rest

(Fig.

5,

*;

r

=

0-55,

P

<

0-01).

This

relationship

was

improved

when

the

larger

bursts

that

occur

during

apnoea

(see

later)

were

included

(Fig.

5,

0;

r

=

0-60,

P

<

00

1).

Including

the

data

obtained

during

apnoea

did

not

alter

the

slope

of

the

relationship.

A

positive

correlation

between

coherence

and

correlation

coefficient

was

225

226

B.

G.

WALLIN,

D.

BURKE

AND

S.

C.

GANDEVIA

apparent

in

the

five

data

points

for

six

individual

subjects,

but

there

was

no

such

trend

in

the

data

from

the

seventh

subject.

Excluding

the

data

for

this

subject

improved

the

correlation

(r

=

07

1,

P

<

00

1).

It

can

be

concluded

that

coherence

and

correlation

coefficient

represent

valid

measures

of

the

homogeneity

of

sympathetic

drive

to

different

muscles,

differing

only

in

sensitivity.

TABLE

1.

The

effects

of

different

manoeuvres

on

coherence

Coherence

during

Control

Manoeuvre

manoeuvre

n

coherence

n

Apnoea

919+213

9

821+250

10

TA

5%

71

8+791

3

904+3

17

3

TA

10%

672+580

8

837+225

8

TA

20%

74-1+3-20

4

85-9+3-0

4

TA

30%

48-5+740

3

84-5+1-5

3

EDL

10%

88-9+2-47

3

91-5+1-39

3

Triceps

surae

10%

86-3

+

2-84

4

88-0

+

2-88

4

Absolute

coherence

is

expressed

as

a

percentage,

with

mean+

S.E.M.

n

=

number

of

subjects.

Effects

of

apnoea

In

nine

experiments,

subjects

held

their

breath

in

expiration

('apnoea')

and,

as

illustrated

in

Figs

1

and

2,

apnoea

increased

the

size

and

frequency

of

sympathetic

bursts

in

both

recordings.

This

increased

the

coherence

between

the

recordings

(Figs

2A,

3

and

4).

Power

increased

in

both

recordings

to

a

similar

extent

(Fig.

2

C

and

D)

and

consequently

there

were

only

small

changes

in

gain

(i.e.

the

common

power

in

the

two

recordings

divided

by

the

power

in

the

recording

from

the

left

leg),

as

shown

in

Figs

2B,

3

and

4.

Apnoea

increased

coherence

in

all

but

one

experiment

over

the

corresponding

values

at

rest,

the

means

being

82-1

%

(S.D.

7-9%)

at

rest

and

91

9°%

(S.D.

6-4

%)

during

the

manoeuvre

(Table

1).

Different

manoeuvres

were

not

equally

efficacious

in

the

same

subject,

presumably

because

of

variability

in

performance.

The

highest

coherence

values

obtained

during

individual

apnoeic

sequences

were

82-4-99-2

%

(mean

94-9

%,

S.D.

5.7

%).

These

findings

suggest

that,

under

some

circumstances,

sympathetic

activity

with

a

cardiac

periodicity

may

be

determined

almost

exclusively

by

a

central

drive

that

is

common

to

different

body

regions,

to

the

virtual

exclusion

of

regional

differences.

The

correlation

coefficients

for

the

relationship

between

the

amplitudes

of

corresponding

bursts

in

the

two

recordings

increased

during

apnoea

from

072

+

O-10

at

rest

to

089

+

007

during

apnoea.

For

the

same

sequences

coherence

increased

from

85-6

+

6-7

%

at

rest

to

93

0

+

6-2

%

during

apnoea.

As

mentioned

above

inclusion

of

these

seven

data

points

in

Fig.

5

improved

the

regression

without

altering

the

slope

of

the

relationship.

Muscle

sympathetic

activity

to

contracting

muscle

In

eight

experiments,

steady

isometric

contractions

were

performed

on

the

right

side

using

the

muscle

innervated

by

the

nerve

fascicle

from

which

MSA

was

being

recorded

(tibialis

anterior,

seven

experiments;

extensor

digitorum

longus,

one

SYMPATHETIC

ACTIVITY

DURING

CONTRACTION

227

experiment).

Simultaneous

recordings

of

sympathetic

activity

were

obtained

from

the

left

(control)

leg

and

from

the

contracting

right

leg.

The

voluntary

contractions

increased

the

background

neural

noise

generated

by

muscle

afferents

and

non-

sympathetic

efferents

in

the

nerve

fascicle

innervating

the

contracting

muscle.

This

150

-

*

Mean

-M

No.

1

_

EI

No.

2

0

0

0

80_

0

410

C

~0

0

Coherence

Gain

Power

left

Power

right

leg

leg

Fig.

6.

The

changes

in

coherence,

gain

and

power

for

three

subjects

during

contractions

of

right

TA

at

5%

MVC.

Data

are

normalized

as

percentages

of

the

pre-contraction

control

values.

For

each

data

set,

the

filled

bar

represents

the

mean

of

the

data

for

the

three

subjects.

Coherence

decreased

in

all

three

subjects,

as

did

gain.

The

power

in

the

recording

from

the

contracting

(right)

leg

decreased

in

all

three

subjects

but,

on

average,

there

was

no

change

in

the

power

from

the

non-contracting

(left)

leg.

increase

in

noise

invalidated

manual

measurements

of

burst

amplitude

from

that

side

but

did

not

alter

the

spectral

analysis

significantly

because

it

did

not

contribute

power

at

the

cardiac

frequency

(see

Methods).

Furthermore,

measurements

of

gain

are

dependent

only

on

the

cross-correlated

(common)

power

in

the

recording

from

that

leg.

Coherence

decreased

in

all

contractions,

no

matter

what

the

contraction

strength

(Table

1).

In

five

of

six

experiments

in

which

contractions

of

different

strength

were

tested,

the

stronger

the

contraction

the

greater

was

the

decrease

in

coherence

(Fig.

3A).

However,

this

relationship

disappeared

in

pooled

data

(see

below),

because

different

individuals

were

tested

with

different

contraction

levels

and

there

was

marked

interindividual

variability

in

the

extent

of

the

decrease.

With

contractions

of

different

strength

(5,

10,

20

and

30%

MVC),

the

average

decrease

in

coherence

exceeded

10%

at

each

level

(Figs

6-8).

The

decrease

in

coherence

was

accompanied

by

a

decrease

in

power

in

the

recording

from

the

contracting

leg

at

each

contraction

level.

This

change

in

power

was

not

statistically

significant

at

any

of

the

levels,

but

when

the

data

were

pooled

for

all

contraction

levels

in

all

subjects,

it

was

statistically

significant,

whether

the

power

during

228

B.

G.

WALLIN,

D.

BURKE

AND

S.

C.

GANDEVIA

contraction

was

compared

with

that

in

the

preceding

rest

period

(mean

decrease

10-0

%;

P

=

0-0079;

two-tailed

paired

t

test)

or

with

the

mean

of

the

power

values

for

the

preceding

and

subsequent

rest

periods

(mean

decrease

13-8

%

P

=

0-00

13;

two-

tailed

paired

t

test).

100

-v

01)

0)

c

a)

_C

0

u

50

20

::

.

.n

A:

.

.:>:

>:

.....

>

-:

-:.

::>

Is

--

n

An.+.

so.

Is

sesns'nsn

:.:.:.:.:.:.:.:.:.-.:.:.-

->:-'-F::

F:

:-F:

:-

:S

>:

:->:

e:::>::

::2:-::2:

.:

::.:-::::.:::::.-e.

:.-

:.:.:.:.:.:.:.:.:.:

:::

.;;:.:

:.:.

:.:

>.-:

>:-F>:-'-:-'-'-:

:*:

:-X

:-:-

-.

;:-

>::

>:'

>:

->:

>:

IS

nt..

In

FS-

::S::::;-:::S::::S:

::

F.:

':F,:::::

S:

.............

.............

......

..............

............

...............

........

..........

.............

..

..............

..............

........

:E

lectro

chemi

cal

0

1

00

01

cm

...........

..

..............

........

....

...

......

....

..

... ... ...

..

100

CD

C

...........

Co

............

cm

-T-

......

...............

.......

...........

...

.............

.............:

...........

.............

............. ............. .............

0

C

0

C:

M

0

M

M

0

CI4

CD

U

(1)

M

Q)

-

CY)

0

0

0

M

lo-o

C

C

x

C

co

CD

CL

0

CL

cc

cc

<

5

<

T

Cr

Fig.

7.

Data

for

contractions

of

right

TA

at

IO

%

MVC.

Each

column

represents

the

mean

for

eight

subjects

(±

I

S.E.M.).

There

was

no

significant

contraction-induced

change

in

power

for

the

non-

contracting

leg

when

the

power

value

during

the

contraction

was

compared

to

that

in

the

preceding

rest

period

(mean

decrease

5-7

%;

P

=

0-2366;

two-tailed

paired

t

-T

SYMPATHETIC

ACTIVITY

DURING

CONTRACTION

229

test).

When

compared

to

the

mean

of

the

power

values

for

the

preceding

and

following

rest

periods,

there

was

a

significant

decrease

in

power,

though

to

a

lesser

extent

than

on

the

contracting

side

(mean

decrease

7-2

%;

P

=

0-0345;

two-tailed

paired

t

test).

As

a

result

the

gain

between

the

two

sides

generally

decreased

(fifteen

120

U,

0

00

Cu

80

0

80

Cy)

C

~0

Co60-

co

CD

Co

40

40

~0

Coherence

Gain

Power

left

Power

right

leg

leg

Fig.

8.

Pooled

data

for

contractions

of

right

TA

at

20

and

30%

MVC,

expressed

as

percentages

of

the

pre-contraction

control

values

(n

=

4

at

20%;

n

=

3

at

30%).

Each

column

represents

the

mean

(±

1

S.E.M.).

There

are

decreases

in

coherence,

gain

and

the

power

for

the

contracting

(right)

leg,

but

no

change

in

the

power

for

the

non-contracting

(left)

leg.

of

eighteen

trials).

This

change

could

be

seen

in

the

mean

data

for

each

contraction

level,

except

at

10

%

MVC

(Figs

6-8),

and

was

significant

when

the

data

for

all

trials

were

pooled

(P

=

0-0085).

In

determining

the

effects

of

contraction

on

power,

the

preferred

comparison

was

with

the

power

in

the

pre-contraction

rest

period

rather

than

with

the

mean

power

in

the

pre-

and

post-contraction

rest

periods.

The

usual

experimental

protocol

involved

performance

of

apnoea

at

the

end

of

a

contraction

and

again

immediately

after

relaxation,

prior

to

the

post-contraction

rest

period

(see

Figs

3

and

4).

In

some

sequences,

the

increase

in

power

due

to

apnoea

subsided

slowly,

continuing

into

the

post-contraction

rest

period.

Similarly,

apnoea

prior

to

contraction

(see

Fig.

6)

may

have

falsely

elevated

the

mean

power

during

the

contraction

in

some

sequences,

but

this

would

have

minimized

the

contraction-induced

change,

not

enhanced

it.

Apnoea

during

contraction

Sustained

apnoea

during

a

contraction

of

the

homonymous

muscle

restored

coherence,

often

above

the

control

values

at

rest

but

below

values

achieved

with

apnoea

when

not

contracting

(Figs

3,

4

and

7).

The

mean

increases

in

coherence

were

18-7,

16-5

and

15-7

%

with

contraction

levels

of

10,

20

and

30%

MVC

respectively.

However,

when

the

apnoea

was

repeated

on

cessation

of

the

contraction,

coherence

increased

by

a

further

6-3,

3-3

and

19-3

%,

respectively.

These

changes

occurred

in

all

trials.

During

contraction,

apnoea

increased

the

power

in

recordings

from

both

the

B.

G.

WALLIN,

D.

BURKE

AND

S.

C.

GANDEVIA

contracting

and

the

non-contracting

legs,

generally

in

parallel

(Fig.

7),

but

it

increased

the

gain

in

ten of

the

fifteen

trials

due

to

a

slightly

greater

effect

on

the

contracting

leg.

As

a

result

apnoea

reduced

the

asymmetry

in

sympathetic

outflow

produced

by

contraction.

Contraction

of

other

muscles

The

effects

of

contracting

a

synergist

(extensor

digitorum

longus,

at

10

%

MVC;

three

experiments)

or

an

antagonist

(triceps

surae,

at

10

%

MVC;

four

experiments)

were

tested

(Figs

3

and

4;

Table

1).

With

contraction

of

the

synergist,

there

was

little

change

in

coherence

(on

average

a

decrease

of

2-6

%),

even

though

one

subject

found

it

difficult

to

prevent

contraction

of

tibialis

anterior

when

holding

a

steady

contraction

of

extensor

digitorum

longus.

With

contractions

of

triceps

surae,

there

was

on

average

a

decrease

in

coherence

of

1-7

%.

With

neither

muscle

were

these

changes

significant;

both

changes

fall

well

within

the

mean

reproducibility

of

coherence

measures

for

consecutive

sequences

(4-2

%,

see

earlier).

Apnoea

when

contracting

a

synergist

or

antagonist

had

similar

effects

as at

rest.

DISCUSSION

The

present

results

have

quantified

the

similarity

between

the

sympathetic

efferent

drives

to

homologous

muscles

in

different

limbs

at

rest,

and

demonstrated

even

greater

similarity

when

subjects

hold

their

breath

in

expiration.

There

is

a

significant

decrease

in

this

similarity

when

one

of

the

homologous

muscles

contracts,

and

this

is

largely

due

to

a

decrease

in

sympathetic

drive

to

the

contracting

muscle.

Contraction

of

a

synergist

or

antagonist

at

10

%

maximal

force

has

little

effect

on

the

symmetry

of

sympathetic

outflow.

MSA

has

a

prominent

cardiac

rhythmicity,

and

power,

gain

and

coherence

were

therefore

measured

at

the

cardiac

period.

The

analysis

would

not

have

measured

all

sympathetic

activity

because

power

at

multiples

of

the

cardiac

period

was

ignored.

However,

the

effects

of

this

omission

were

small:

the

peak

at

the

cardiac

period

dominated

the

power

spectrum

(see

Fig.

2),

and

inspection

of

the

plots

indicated

that

the

peaks

at

the

cardiac

period

and

at

twice

the

period

behaved

in

much

the

same

way

during

contractions.

During

apnoea

more

heart

beats

were

associated

with

sympathetic

bursts

than

in

the

control

situation,

and

this

change

may

have

accentuated

the

increase

in

power

at

the

cardiac

period.

Homogeneity

of

sympathetic

drives

Total

sympathetic

outflow

presumably

depends

on

a

balance

between

drives

that

are

common

to

all

muscle

groups,

and

regionally

specific

drives

of

central

or

peripheral

(reflex)

origin.

In

subjects

at

rest

there

is

probably

no

specific

stimulus

for

regional

variations

in

muscle

blood

flow

and,

accordingly,

the

coherence

between

sympathetic

drives

to

muscles

of

different

limbs

was

high

in

relaxed

subjects.

During

apnoea

the

sympathetic

outflow

to

muscle

increases

but

again

there

is

no

reason

why

this

drive

should

be

distributed

unevenly

to

the

two

legs.

Coherence

decreased

when

one

of

the

innervated

muscles

was

contracted.

This

implies

that

the

common

drive

had

become

less

dominant

and,

by

inference,

that

230

SYMPATHETIC

ACTIVITY

DU

RING

CONTRACTION1

regional

drives

were

more

important.

However,

this

inference

is

valid

only

if

a

change

in

noise

can

be

excluded

as

the

cause

for

the

decreased

coherence.

Inevitably

contraction

is

associated

with

increased

afferent

and

efferent

traffic

in

the

nerve

fascicle

innervating

the

contracting

muscle.

but

this

is

unlikely

to

have

affected

the

measurements

for

two

reasons.

First,

an

increase

in

noise

would

be

expected

to

increase

the

power

in

the

recording

from

the

contracting

leg.

not

decrease

it

as

was

consistently

found.

Secondly,

the

measurements

of

power,

gain

and

coherence

were

made

at

the

cardiac

period

because

MSA

occurs

in

pulse-synchronous

bursts.

Control

studies

indicate

that

only

changes

in

noise

that

occur

with

a

cardiac

periodicity

affect

coherence

(see

Methods).

Thus,

regional

factors,

central

or

peripheral,

assume

greater

importance

in

determining

sympathetic

outflow

when

subjects

contract a

muscle

than

when

they

are

at

rest.

In

anaesthetized

and

decerebrate

cats

with

all

baroreceptors

denervated.

Koesis.

Gebber,

Barman

&

Kenney

(1990)

determined

the

coherence

between

different

sympathetic

nerve

pairs

using

a

similar

analysis

to

that

used

here.

There

was

correlated

activity

in

the

2-6

Hz

range

with

different

phase

relationships

that

could

be

altered

experimentally

(e.g.

by

altering

ventilation

rate).

These

analyses

indicate

that

the

relationship

between

different

sympathetic

outflows

is

not

immutable,

a

conclusion

supported

by

the

present

findings.

Their

coherence

values

were

generally

lower

than

in

the

present

study,

some

50-70

%,

and

this

raises

the

question

whether

the

present

analysis,

restricted

to

power

at

the

cardiac

period,

might

have

resulted

in

high

coherence

values

that

exaggerate

the

degree

of

similarity

between

the

correlated

recordings.

In

this

context,

it

is

notable

that

the

correlation

coefficient

of

the

relationship

between

the

amplitudes

of

equivalent

sympathetic

bursts

suggests

that,

at

rest,

52

%

of

the

variance

of

MSA

in

the

two

recordings

was

due

to

a

common

drive,

less

than

suggested

by

the

coherence

value.

On

the

other

hand,

the

baroreflex

is

responsible

for

the

pulse-synchronous

bursts

seen

in

human

MSA

(Wallin

&

Fagius,

1988),

and

this

would

impose

a

high

degree

of

similarity

for

nerves

innervating

muscle

throughout

the

body.

Vasoconstrictor

or

vasodilator

activity

There

is

strong

evidence

that

the

pulse-synchronous

multiunit

bursts

of

MSA

are

dominated

by

vasoconstrictor

impulses

which

are

influenced

by

arterial

baroreflexes

(Vallbo

et

al.

1979;

Vissing,

Scherrer

&

Victor,

1989).

Vasodilator

efferents

are

present

in

the

skin

of

the

human

forearm

and

lower

leg

(Grant

&

Holling,

1938)

and

foot

(Lundberg,

Norgren,

Ribbe,

Rosen.

Steen,

Thorne

&

Wallin,

1989)

and,

although

the

question

is

not

settled,

some

observations

suggest

that

human

muscle

receives

an

active

vasodilator

innervation

(Blair,

Glover,

Greenfield

&

Roddie,

1959).

However,

muscle

vasodilator

neurones

in

the

cat

are

neither

spontaneously

active

at

rest

nor

influenced

by

arterial

baroreceptors

(Horeyseck,

Jhnig,

Kirchner

&

Thhmer,

1972,

1976).

Thus,

even

if

a

putative

vasodilator

innervation

of

human

muscle

were

important,

the

present

analyses

have

probably

avoided

such

activity

by

focusing

on

pulse-synchronous

activity

and

activity

in

correlograms

with

a

cardiac

periodicity.

231

B.

G.

WALLIN,

D.

BURKE

AND

S.

C.

GANDEVIA

Changes

in

sympathetic

activity

associated

with

voluntary

contraction

The

findings

of

decreased

coherence

between

the

recordings

from

the

two

legs

and

decreased

power

in

the

recording

from

the

contracting

leg

suggest

that

contraction

of

a

lower

limb

muscle

at

5-30

%

MVC

induced

a

local

reduction

of

sympathetic,

presumably

vasoconstrictor,

drive

to

that

muscle.

The

specificity

of

this

alteration

is

illustrated

by

the

fact

that

no

such

change

was

found

when

the

contraction

involved

not

the

innervated

muscle

but

a

neighbour

in

the

same

leg,

be

it

a

synergist

or

antagonist.

A

decrease

in

MSA

would

certainly

add

to

metabolic

factors

tending

to

produce

vasodilatation

in

the

contracting

muscle.

Whether

the

underlying

mechanism

for

the

MSA

reduction

is

a

descending

supraspinal

influence

or

an

inhibitory

spinal

reflex

is

unclear.

There

has

been

no

previous

study

of

changes

in

MSA

to

contracting

muscles,

but

one

study

of

noradrenaline

spillover

led

to

the

opposite

conclusions

(Savard

et

al.

1987).

However,

that

study

involved

dynamic

(rather

than

static)

contractions

of

a

large

muscle

mass

performed

at

50-100

%

of

maximal

power

for

10-20

min.

Compared

to

our

weak,

fairly

short-lasting

contractions

such

efforts

should

be

expected

to

produce

a

much

greater

cardiovascular

load

and

therefore

it

would

be

reasonable

to

expect

a

difference

in

vasoregulatory

neural

outflow.

A

number

of

studies

have

addressed

changes

in

MSA

to

lower

limb

muscles

during

sustained

contractions

of

muscles

of

the

forearm

or

jaw

(Delius,

Hagbarth,

Hongell

&

Wallin,

1972;

Mark,

Victor,

Nerhed

&

Wallin,

1985;

Seals,

1989;

Elam,

Johansson

&

Wallin,

1991).

The

common

finding

was

an

increase

in

MSA,

particularly

after

contractions

lasting

more

than

1

min.

The

degree

of

MSA

increase

is

load

and

time

dependent

(Saito,

Mano,

Abe

&

Iwase,

1986;

Seals

&

Enoka,

1989)

and

has

been

found

to

correlate

to

a

decrease

of

intramuscular

pH

in

the

contracting

muscles

(Victor,

Bertocci,

Pryor

&

Nunnally,

1988).

In

contrast,

recent

studies

suggest

that

static

leg

exercise

does

not

increase

MSA

to

the

non-contracting

leg

(Ray,

Rea,

Clary

&

Mark,

1990,

1991).

This

finding

is

consistent

with

the

present

results.

This

work

was

supported

by

the

National

Health

and

Medical

Research

Council

of

Australia,

the

Ramaciotti

Foundations

and

the

Swedish

Medical

Research

Council

(grant

B91-04X-03546-20A).

We

would

like

to

thank

G.

Macefield,

J.

P.

Hales

and

R.

B.

Gorman

for

programming

and

for

assistance

with

experiments

and

analysis.

REFERENCES

ANDERSON,

E.

A.

WALLIN,

B.

(".

&

MARK,

A.

L.

(1987).

Dissociation

of

sympathetic

nerve

activity

in

arm

and

leg

muscle

during

mental

stress.

Hypertension

6,

suppl.

III,

114-119.

BLAIR,

D.

A..

GLOVER,

W.

E.,

GREENFIELD,

A.

D.

M.

&

RoDDIE,

I.

C.

(1959).

Excitation

of

cholinergic

vasodilator

nerves

to

human

skeletal

muscles

during

emotional

stress.

Journal

of

Physiology

148.

633-647.

DELIUS.

W.,

HAGBARTH.

K.-E.,

HONGELL,

A.

&

WALLIN,

B.

G.

(1972).

Manoeuvres

affecting

sympathetic

outflow

in

human

muscle

nerves.

Acta

Physiologica

Scandinavica

84,

82-94.

ELAM,

M.,

JOHANSSON.

G.

&

WALLIN,

B.

G.

(1992).

Do

patients

with

primary

fibromyalgia

have

an

altered

muscle

sympathetic

nerve

activity?

Pain

48,

371-375.

GRANT,

R.

T.

&

HOLLING,

H.

E.

(1938).

Further

observations

on

the

vascular

responses

of

the

human

limb

to

body

warming:

evidence

for

sympathetic

vasodilator

nerves

in

the

normal

subject.

Clinical

Science

3,

273-285.

232

SYMPATHETIC

ACTIVITY

DURING

CONTRACTION

HOREYSECK,

G.,

JANIG,

W.,

KIRCHNER,

F.

&

THXMER,

F.

&

THXMER,

V.

(1972).

Activation

of

muscle

vasodilator

neurons

by

hypothalamic

stimulation.

Brain

Research

48,

394-396.

HOREYSECK,

G.,

JANIG,

W.,

KIRCHNER,

F.

&

THXMER,

V.

(1976).

Activation

and

inhibition

of

muscle

and

cutaneous

postganglionic

neurones

to

hindlimb

during

hypothalamically

induced

vasoconstriction

and

atropine-sensitive

vasodilation.

Pfluigers

Archiv

361,

231-240.

Kocsis,

B.,

GEBBER,

G.

L.,

BARMAN,

S.

M.

&

KENNEY,

M.

J.

(1990).

Relationships

between

activity

of

sympathetic

nerve

pairs:

phase

and

coherence.

American

Journal

of

Physiology

259.

R549-560.

LUNDBERG,

J..

NORGREN,

L.,

RIBBE,

E.,

ROSEN,

I.,

STEEN,

S.,

Tti6RNE.

J.

&

WALLIN,

B.

G.

(1989).

Direct

evidence

of

active

sympathetic

vasodilatation

in

the

skin

of

the

human

foot.

Journal

of

Physiology

417,

437-446.

MCKEON,

B.

&

BURKE,

D.

(1981).

Component

of

muscle

spindle

discharge

related

to

arterial

pulse.

Journal

of

Neurophysiology

46,

788-796.

MARK,

A.

L.,

VICTOR,

R.

G.

&

NERHED,

C.

&

WALLIN,

B.

G.

(1985).

Micro-neurographic

studies

of

the

mechanisms

of

sympathetic

nerve

responses

to

static

exercise

in

humans.

Circulation

Research

57,

461-469.

PETERSON,

D.

F.,

ARMSTRONG,

R.

B.

&

LAUGHLIN,

M.

H.

(1988).

Sympathetic

neural

influences

on

muscle

blood

flow

in

rats

during

submaximal

exercise.

Journal

of

Applied

Physiology

65,

434-440.

RAY,

C.

A.,

REA,

R.

F.,

CLARY,

M.

P.

&

MARK,

A.

L.

(1990).

Static

and

dynamic

leg

exercise

decreases

muscle

sympathetic

nerve

activity.

Circulation

82,

suppl.

III,

692.

RAY,

C. A.,

REA,

R.

F.,

CLARY,

M.

P.

&

MARK,

A.

L.

(1991).

Contrasting

responses

of

arm

and

leg

muscle

sympathetic

nerve

activity

to

static leg

exercise

and

postexercise

circulatory

arrest.

FASEB

Journal

5,

A382.

SAITO,

M.,

MANO,

T.,

ABE,

H.

&

IWASE,

S.

(1986).

Responses

in

muscle

sympathetic

nerve

activity

to

sustained

hand-grips

of

different

tensions

in

humans.

European

Journal

of

Applied

Physiology

55,

493-498.

SAVARD,

G.,

STRANGE,

S.,

KIENS,

B.,

RICHTER,

E.

A.,

CHRISTENSEN,

J.

N.

&

SALTIN,

B.

(1987).

Noradrenaline

spillover

during

exercise

in

active

versus

resting

skeletal

muscle

in

man.

Acta

Physiologica

Scandinavica

131,

507-515.

SEALS,

D.

R.

(1989).

Sympathetic

neural

discharge

and

vascular

resistance

during

exercise

in

humans.

Journal

of

Applied

Physiology

66,

2472-2478.

SEALS,

D.

R.

&

ENOKA,

R.

M.

(1989).

Sympathetic

activation

is

associated

with

increases

in

EMG

during

fatiguing

exercise.

Journal

of

Applied

Physiology

66,

88-95.

SUNDL6F,

G.

&

WALLIN,

B.

G.

(1977).

Human

muscle

nerve

sympathetic

activity

at

rest.

Relationship

to

blood

pressure

and

age.

Journal

of

Physiology

272,

383-397.

VALLBO,

A.

B.,

HAGBARTH,

K.-E.,

TOREBJ6RK,

H.

E.

&

WALLIN,

B.

G.

(1979).

Somatosensory,

proprioceptive,

and

sympathetic

activity

in

human

peripheral

nerves.

Physiological

Reviews

59,

919-957.

VICTOR,

R.

G.,

BERTOCCI,

L.

A.,

PRYOR,

S.

L.

&

NUNNALLY,

R.

L.

(1988).

Sympathetic

nerve

discharge

is

coupled

to

muscle

cell

pH

during

exercise

in

humans.

Journal

of

Clinical

Investigation

82,

1301-1305.

VISSING,

S.

F.,

SCHERRER,

U.

&

VICTOR,

G.

(1989).

Relation

between

sympathetic

outflow

and

vascular

resistance

in

the

calf

during

perturbations

in

central

venous

pressure.

Circulation

Research

65,

1710-1717.

WALLIN,

B.

G.

&

FAGIUS,

J.

(1988).

Peripheral

sympathetic

neural

activity

in

conscious

humans.

Annual

Review

of

Physiology

50,

565-576.

WALLIN,

B.

G.,

VICTOR,

R.

G.

&

MARK,

A.

L.

(1989).

Sympathetic

outflow

to

resting

muscles

during

static

handgrip

and

post-contraction

muscle

ischemia.

American

Journal

of

Physiology

256,

H10S-110.

233