Effects of age, sex, and anthropometric factors on nerve conduction measures

Associations among measures of median, ulnar, and sural nerve conduc-tion and age, skin temperature, sex, and anthropometric factors were eval-uated in a population of105healthy, asymptomatic adults without occupa-tional exposure to highly repetitive or forceful hand exertions. Height wasnegatively associated with sensory amplitude in all nerves tested

(P

<

0.001

,

and positively associated with median and ulnar sensory distal la-tencies

(P

<

0.01) and sural latency

(P

<

0.001).

Index finger circumfer-ence was negatively associated with median and ulnar sensory amplitudes

(P

<

0.05).

Sex, in isolation from highly correlated anthropometric factorssuch as height, was not found to be a significant predictor of median or ul-nar nerve conduction measures. Equations using age, height, and fingercircumference for prediction of normal values are presented. Failure to ad-just normal nerve conduction values for these factors decreases the diag-nostic specificity and sensitivity of the described measures, and may resultin misclassification

of

individuals.

0

1992

John

Wiley

&

Sons,

Inc.

Key words: nerve conduction studies normal values age height an-thropometry

MUSCLE

&

NERVE

15:1095-1104 1992

EFFECTS OF AGE, SEX, ANDANTHROPOMETRIC FACTORS ONNERVE CONDUCTION MEASURES

DIANA

S.

STETSON, PhD, JAMES W. ALBERS,

Mb,

PhD, BARBARA A.SILVERSTEIN, PhD,

and

ROBERT

A.

WOLFE, PhD

Normal values for distal nerve conduction mea-sures are needed for clinical evaluation

of

individ-ual patients and as control data for epidemiologicstudies of work-related peripheral nerve compres-sion disorders such as carpal tunnel syndrome. Al-though many control opulations have been de-~cribed,5,~,'

*17"4'28729~3H)variation~

n methodology,unspecified interelectrode distances, choice of dig-its, and type (or lack)

of

skin temperature control,as well as small sample sizes, limit their usefulness.Furthermore, while some reports specified thatcontrol subjects had no peripheral nerve dis-

From the Department of Environmental and Industrial Health, School ofPublic Health (Dr. Stetson). and Department

of

Neurology, MedicalSchool (Dr. Albers). The University of Michigan, Ann Arbor, Michigan:Safety

&

Health Assessment Research Program, Washington Departmentof Labor and Industries, Olympia, Washington (Dr. Silverstein); and De-partment of Biostatistics. School of Public Health, The University of Mich-igan, Ann Arbor, Michigan (Dr. Wolfe)Acknowledgments: Partial funding was received from the General MotorsCorporation. The authors thank Cathy Olsewski, REDT for performing thenerve conduction studies. The employees who participated are thankedfor their time and good will.Address reprint requests to Dr.

J.

W. Albers, Department of Neurology,1C 325 UH,

Box

0032, The University of Michigan Medical Center, 1500

E.

Medical Center Drive, Ann Arbor, MI 48109.Accepted for publication March 2, 1992.CCC

01

48-639)(/92/101095-

10

0

1992 John Wiley

&

Sons, Inc.

ease,17*28*29one fully described the occupations

of

their subjects or specified a population with

low

ex-posure

to

known occupational risk factors for car-pal tunnel syndrome such as force, repetitiveness,andThis study describes the distribution of me-dian, ulnar, and sural nerve conduction measuresin a population of healthy, working adults withoutexposure to forceful or repetitive hand exertions,or segmental vibration. The hypotheses of no as-sociation between each nerve conduction measureand independent variables known or hypothesized

to

affect nerve conduction

were

tested. These fac-tors included age>,19326*38kin temperat~re,"'~'~.~~height,7,2

1,32935

and finger circumference.3

A

square-shaped wrist has been negatively associatedwith distal median nerve conduction." Factors forwhich associations with specific nerve conductionmeasures were found to be statistically significant,biologically plausible, and clinically importantwere used to develop formulas which predict nor-mal values.

MATERIALS

ANDMETHODS

Population.

Participants were salaried employeesin a large corporation. Prior to age- and sex-strat-

Nerve

Conduction

Measures

MUSCLE &NERVE

October

1992

1095

ified random selection, employees on jobs thoughtto have repetitive or forceful hand exertions wereexcluded (e.g., secretaries, word processors). Of

27

1 randomly selected empl~yees,~’ 49 wereavailable for participation (Table

1).

Of these, 75refused to participate and

69

were excluded, leav-ing 105 study participants. Exclusion criteria wereneurological symptoms, a self-reported history ofa systemic illness or disorder affecting the centralor peripheral nervous systems, or occupational ex-posure to forceful or repetitive hand exertions.Participants returned a self-administered ques-tionnaire which included information on age,height, medical history, occupational history, andsymptoms in the handlwrist during the precedingyear.

Physical Evaluation.

All subjects had a screeningneurologic examination performed by a neurolo-gist

(J.

A.) and an additional examination of thehand and wrist performed by a nurse

(B.

S.).

In-dex finger circumference was measured at themiddle of the proximal phalanx with a narrowcloth measuring tape. The wrist anterior- poste-rior and medial- lateral diameters were measuredwith engineering calipers at the distal wristcrease.

8

Wrist ratio was calculated as the ante-rior- posterior diameter divided by the medial-lateral diameter.

Table

1.

Subject selection criteria.Consent form and questionnaireLess:

n

=

271

I

ent

to

randomly selected employeesUnavailable for participation

(n

=

22)Refused

to

participate

(n

=

75)Medical exclusion*

(n

=

28)>21

oz.

ethanol/week orself-described problem

(n

=

2)Numbness

or

paresthesia in handoccurring

>3

imes or lasting >1 week oroccurring within prior week

(n

=

18)

n

=

249

n

=

174Excluded

(n

=

69):Pain in volar hand/wrist

(n

=

2)

Current job has

24

h/d withrepetitive/forceful hand exertions

(n

=

14)Prior job, within previous 2.5 years, withrepetitive/forceful hand exertions

(n

=

5)

v

Study population

n

=

105

~ ~

*Medical reasons for exclusion were diabetes mellitus, rheumatoidarthritis, abnormal thyroid function within previous

10

years, carpaltunnel syndrome, cervical disk disease (or symptoms of paresthesias inhands radiating from the neck) pregnancy, renal disease, hemophilia,prior radial mononeuropathy, prior cerebral infarction, chemotherapy, oruse of anticonvulsant medication

Nerve Conduction Studies.

Nerve conductionstudies were performed on the dominant hand ofall subjects by the same certified electromyogra-phy technician. Using a TECA

TD-5

electromyo-graphy machine and standard TECA surface elec-trodes, standard techniquesofsupramaximalpercutaneous nerve stimulation and surface re-cording were employed.’ Anatomic landmarksand standardized stimulation to recording elec-trode distances are shown in Table 2. Skin tem-peratures were monitored at the proximal creaseof digits

I1

and

V

and at the midcalf. Tempera-ture was maintained above 32°C in the hand, andabove 30°C in the calf, with an electric heatingpad.Sensory nerve action potential and compoundmuscle action potential amplitudes were measuredfrom the baseline to the negative peak. Sensory la-tencies were measured to the onset of the negativedeflection and the negative peak. Motor latencieswere measured to the onset of the initial negativedeflection. Median motor conduction velocity forthe forearm was calculated using proximal and di-stal onset latencies. Terminal sensory conductionvelocities were calculated by dividing the distal dis-tance by the onset latency. F-wave latencies weremeasured as the minimal latency in a series of fiveresponses.

Statistics.

Student’s

t

test was used to comparethe difference between means for quantitativevariables. Least squares regression was used formodel building and hypothesis testing. The Pear-son product moment was calculated for correla-tions of continuous variables. A P-value of c0.05was used to define statistical significance. Formodel building, criteria for including an addi-tional independent variable were that its coeffi-cient have a P-value

<0.05

and the partial

R2

asso-ciated with adding the variable be

>0.05.

Data formany of the nerve conduction measures did notfollow a normal distribution. Prior to hypothesistesting, several transformations (natural log,square root, and inverse) were evaluated. Trans-formations resulting in a more normal distribu-tion (Kolmogorov D-statistic and visual inspection

of

histogramsbox plots) were used in the regres-sion models. A model building approach was usedfor hypothesis testing. The independent variableswere added in an order based on the strength ofpreviously reported evidence for biologically plau-sible and significant associations: age, skin temper-ature, height, dominant index finger circumfer-ence, dominant wrist ratio, and sex.

1096

Nerve Conduction Measures

MUSCLE

&

NERVE

October 1992

Table

2.

Description

of

nerve conduction measurements

(n

=

105)

mean

*

standard deviatlon (range)Recording electrodes(interelectrode Conduction velocity(m/s)erve Stimulate distance) Amplitude

(pV

or mV) Latency (ms)Median sensory WristWristMidpalm, midthenarcreaseMedian motor WristAntecubital fossa, overbrachial pulseproximalMedian

F

response Wrist; cathodeUlnar sensory WristWristMedial midpalmSural sensory MidcalfDigit

/I,

proximal anddistal phalanxes;

30-40

mm apart

(140

mm)Digit

IV,

proximal anddistal phalanxes;

30-40

mm apart

(140

mm)Wrist, between

PL'

and FCR tendons

(80

mm)Thenar muscle

(70

mm)Thenar muscle(measured)Thenar muscleDigit

V,

proximal anddistal phalanxes;

30-40

mm apart

(1

40

rnm)

Digit

IV,

proximal distalphalanxes;

30-40

mm apart

(140

mm)Medial wrist

(80

mm)Lateral malleolus

(140

rnm)

32.7

i

1.4(14-68)17.7

*

8.6(5-50)115

f;

41.5(40- 225)12.5

f;

3.1

(4-19)28.9

*

11.6(9-60)19.0

c_

7.3(6-38)20.8

2

8.417.5

t

.7(8-52)(6-48)3.0

t

0.2(2.6

-

.8)

3.0

t

0.3

(2.5-4.1)1.8

*

0.2(1.4-2.4)3.2

i

.4

(2.4

-

.2)27.1

*

2.22.8

*

0.2

(21.6-32.6)(2.5

-

.7)2.9

2

0.2

(2.4- 3.9)1.7

2

0.13.4

2

0.3(1.5

-

.2)(2.9

-

4.9)60.2

2

4.9(47-70)56.7

t

2.9(50-64)63.0

-+

4.1(48

-

4)52.2

-t

5.3(36-64)

*PL

=

palmaris longus,

FCR

=

flexor carpi radfalfs

RESULTS

Population Description.

The mean

(+SD)

age formen was

42.7

(&

12.6)

years and for women

it

was

37.8 (k10.0)

years

(P

<

0.05);

a e and sex distri-butions are reported elsewhere: Distributions ofheight, dominant hand index finger circumfer-ence, and dominant hand wrist ratio, stratified bysex, are summarized in Table

3.

As

a group,women were shorter and had a smaller finger cir-

3$

'

.

Table

3.

Population distribution of height, dominant hand indexfinger circumference,

and

wrist ratio,

by