The small fiber neuropathy screening list: Construction and cross-validation in sarcoidosis.

Abstract

Small fiber neuropathy (SFN) appears to be relatively common in sarcoidosis patients. However, there is no golden standard to establish SFN and diagnostic tests for SFN are not widely available. There is a need for an easy to administer SFN screening instrument for clinical assessment, research or therapeutic trials. The aim of the present study was to develop a screening list to identify sarcoidosis patients with SFN in general clinical practice.
We studied 139 sarcoidosis patients. The first consecutive 84 patients (Group 1) underwent temperature threshold testing (TTT) and completed an extensive SFN-symptoms-questionnaire. Based on data from Group 1 and using distribution measures and discriminant analyses, a screening list for SFN in sarcoidosis consisting of 21 questions was constructed: the Small Fiber Neuropathy Screening List (SFNSL). Subsequently, this SFNSL was crossvalidated in the next 55 consecutive patients (Group 2).
The same cut-off scores as found for Group 1 were appropriate in Group 2. The SFNSL was found to have high levels of internal consistency (Cronbach's alpha 0.90) and exploratory factor analysis showed that it measures only one underlying factor. Convergent validity seems good.
To assess the presence of SFN in clinical practice the SFNSL, a brief and easy to administer questionaire, was developed in a sarcoidosis population. The results of the present study support the idea that SFN is a serious problem in chronic sarcoidosis. Future studies are needed to establish the broad usefulness of this SFN screening list and expand knowledge on the psychometric properties.

Full text

Page 1

⏐129 The Small Fiber Neuropathy Screening List
Chapter 8

The Small Fiber Neuropathy Screening List:
Construction and cross-validation in sarcoidosis







E Hoitsma, J De Vries, CG Faber, M Drent

Submitted for publication

Page 2

130⏐Chapter 8
Abstract
Background and aim
Small fiber neuropathy (SFN) appears to be relatively common in sarcoidosis patients. However,
there is no golden standard to establish SFN and diagnostic tests for SFN are not widely available.
There is a need for an easily to administer SFN screening instrument for clinical assessment,
research or therapeutic trials. The aim of the present study was to develop a screening list to
identify sarcoidosis patients with SFN.

Methods
We studied 139 sarcoidosis patients. The first consecutive 84 patients (Group 1) underwent
temperature threshold testing (TTT) and completed an extensive SFN-symptoms-questionnaire.
Based on data from Group 1 and using distribution measures and discriminant analyses, a
screening list for SFN in sarcoidosis consisting of 21 questions was constructed: the Small Fiber
Neuropathy Screening List (SFNSL). Subsequently, this SFNSL was crossvalidated in the next 55
consecutive patients (Group 2).

Results
The same cut-off scores as found for Group 1 were appropriate in Group 2. The SFNSL was found
to have high levels of internal consistency (Cronbach`s alpha 0.90) and exploratory factor analysis
showed that it measures only one underlying factor. Convergent validity seems good.

Conclusions
We developed a brief and easy to administer screening list to assess SFN in sarcoidosis. The
results of the present study support the idea that SFN is a relatively uniform disorder. Future
studies in patient populations suffering from SFN associated with other causes are needed to
establish the broad usefulness of this SFN screening list and expand knowledge on the psycho-
metric properties.

Page 3

The Small Fiber Neuropathy Screening List ⏐131
Introduction
Sarcoidosis is a disseminated granulomatous disease of unknown origin.1 Depending on
the organs involved and the severity of granulomatous inflammation, patients suffer
from a broad range of persistent physical symptoms. Besides respiratory symptoms
such as coughing and dyspnoea on exertion, patients often suffer from systemic non-
specific symptoms such as fatigue and pain.2-6
Pain is considered to be a reflex response to underlying somatic pathology. In a
previous study we found that many sarcoidosis patients with peripheral pain appeared
to suffer from small fiber neuropathy (SFN) with involvement of autonomic nerve
fibers.7,8 SFN is a generalised peripheral neuropathy selectively involving Aδ and
C fibers. When the somatic small afferent fibers are affected, symptoms typically consist
of neuropathic pain. Furthermore, autonomic fibers may be involved, causing autonomic
dysfunction.
Routinely applied nerve conduction tests as well as tendon reflexes evaluate only large
nerve fiber function and consequently remain normal in isolated SFN. Besides,
symptoms of autonomic dysfunction are not always sufficiently severe to be mentioned
spontaneously by the patient. Furthermore, sarcoidosis patients are generally seen by
physicians, such as internists and pulmonologists, who may not be familiar with SFN.
Therefore, the diagnosis of SFN can easily be missed. Tests for assessment of small
nerve fibers include temperature threshold testing (TTT), quantitative sudomotor axon
reflex testing (QSART), intraepidermal nerve fiber density assessment in skin biopsy
and laser evoked potentials.9 These tests are not widely available, however.
There is a need for adequate means of assessing the presence of SFN, both for clinical
management and also for guidance of the development of further therapies. Assess-
ment of SFN may also be useful in epidemiological and pathophysiological studies. The
aim of the present study was to develop a short and easy to administer questionnaire
that screens for the presence of SFN in sarcoidosis patients.
Patients and methods
Participants
From 2001 to 2004, 139 (82 males and 57 females) sarcoidosis patients who visited the
outpatient clinic of the University Hospital Maastricht, a referral centre for sarcoidosis,
participated in the present study. Patients were diagnosed with sarcoidosis based on
consistent clinical features and bronchoalveolar lavage (BAL) fluid analysis or biopsy

Page 4

132⏐Chapter 8
results, according to the WASOG guidelines.1 Informed consent was obtained from all
participating patients. Relevant co-morbidity only included diabetes mellitus (n=7).
The first consecutive 84 patients seen before august 2003 (49 males and 35 females;
mean age 44.2±11.1) were evaluated with an extensive pilot questionnaire (Group 1).
Based on these data and using distribution measures and discriminant analyses, a
shorter screening list was constructed (see below). Patients seen after august 2003
(Group 2) were used for cross validation of this screening list. This group consisted of
55 patients (34 males and 21 females; mean age 45.5±10.7). Patient characteristics are
summarized in table 8.1.
Finally, 15 healthy controls (mean age 33.3 ± 9.8; 8 males, 7 females) were evaluated.

Table 8.1 Summary of the most relevant characteristics of the studied sarcoidosis population.

Group 1
(n=84)
Group 2
(n=55)

TTT normal TTT abnormal TTT normal TTT abnormal
Number of patients
Age (years)
Sex (males / females)
Diabetes (yes / no)
Time since diagnosis (years)
Chest X-ray stage (0/I/II/III/IV)
Prednison (yes / no)
31 53 19 36
37.3 ± 9.8
12 / 19
1 / 30
2.5 ± 3.9
6 / 4 / 6 / 12 / 3
14 / 17
47.3 ± 10.7*
34 / 19
2 / 51
5.8 ± 8.2*
8 / 9 / 18 / 13 / 5
20 / 33
38.1 ± 8.9
9 / 10
1 / 18
2.4 ± 2.6
3 / 2 / 4 / 4 / 6
8 / 11
45.1 ± 9.6
25 / 11
3 / 33
5.9 ± 6.6*
6 / 8 / 12 / 8 / 2
18 / 19
TTT: temperature threshold testing. Data are expressed as absolute numbers and, if appropriate, mean ±
standard deviation. No statistical differences were found between group 1 and 2. *p<0.05 between the subgroups
with a normal TTT or an abnormal TTT.

Pilot questionnaire of the small fiber neuropathy screening list
(SFNSL)
Based on clinical experience and existing neuropathy questionnaires,10-19 a pilot
questionnaire consisting of 93 questions, covering some 30 different complaints, was
constructed. It had three parts: questions in part I (35 questions) concerned presence or
absence of complaints; questions in part II (29 questions) were aimed at the frequency
of complaints; and part III (29 questions) concerned the severity of the complaints. The
response scale for part II ranged from 0 Never to 4 Always and for part III the scale went
from 0 Never to 4 Severe. The patients in Group 1 who reported pain also completed
the Neuropathic Pain Scale.20

Page 5

The Small Fiber Neuropathy Screening List ⏐133
Temperature threshold testing (TTT)
TTT was used to assess function of small calibre sensory fibers by measuring tempera-
ture sensation thresholds. TTT was done with a Medoc TSA-2001 device (Medoc,
Ramat Yishai, Israel). Thresholds for warm and cold sensation were determined on the
hand and dorsum of the foot on both sides using the method of levels (MLE) and the
method of limits (MLI) as described previously.6 Normative data according to Yarnitsky
were used.21 Temperature sensation was considered abnormal if at least on one side
both MLE and MLI testing resulted in Z values exceeding 2.5 (above the 99th percen-
tile).8
Statistical procedure and construction steps of the SFNSL
Frequencies were used for the characteristics of the patient groups. A number of steps
were performed using Group 1 to develop the SFNSL by reducing the number of
questions of the pilot questionnaire. First, missing values were examined to identify
questions with a percentage of missing values above 10%. Second, remarks from
patients concerning the questionnaire were recorded. Furthermore, three series of
discriminant analyses were performed starting with (i) the questions from Part II and (ii)
the questions from Part III of the pilot questionnaire, and (iii) the remaining questions
from Part II and III together. The criteria for the discriminant analyses were the size of
the discriminant function, the percentage predicted in the correct category, and reducing
the number of questions as much as possible. This resulted in the SFNSL. Subse-
quently, exploratory factor analysis (principle axis factoring) was performed using the
screetest criterion22 to establish the number of underlying factors measured by the
questionnaire and Cronbach’s alpha was employed to measure internal consistency.23
We used a criterion of 0.70-0.80 to indicate adequate internal consistency.23-25 In
addition, Pearson correlations and t-tests were performed between the SFNSL and the
neuropathic pain scale (NPS), depending on the questions of the latter questionnaire, to
provide some preliminary information on construct or convergent validity. It is usually
accepted that correlations above 0.40 indicate acceptable convergent validity.26 The
SFNSL was than completed by Group 2. The percentage of missing values was
checked. Again, exploratory factor analysis (principle axis factoring) was performed and
internal consistency was examined. The cut-off scores found in Group 1 were examined
on applicability in Group 2. Statistical analyses were performed using the SPSS11.0 for
Windows (SPSS, Chicago, IL, USA).

Page 6

134⏐Chapter 8
Results
Development of SFNSL in Group 1
The examination of the answers to the questions of the first consecutive 84 patients
(Group 1) revealed that seven questions had more than 10% missing values. One
question concerned the partner and whether he/she informed the patient about frequent
leg movement at night and the other questions concerned sexual intercourse related
questions. Therefore, we decided to remove these questions from further analyses.
The remaining remarks from the patients concerned part I of the questionnaire. Patients
indicated to find it difficult to answer these questions because of the yes/no response
category. They were not comfortable with it because of lack of sophistication (not
detailed enough). They frequently commented that they wanted to answer ‘sometimes’
and felt that yes was too strong and no was also not good. For this reason, all questions
in part I were left out of the subsequent analyses. Thus, for the statistical analyses 51
questions (25 part II and 26 part III) were used.
Subsequently, three series of discriminant analyses were performed to find out which
questions can correctly distinguish patients with normal and abnormal TTT results.
These specific questions were selected. In the first series of analyses the questions
from Part II, were used. The number of questions was reduced to 10. In the second
series of analyses, the same was done for the questions from Part III of the pilot
questionnaire. This resulted in 15 questions. In the final series, the remaining questions
from parts II and III together (25 questions) were used to group patients according to
their SFN status based on their TTT scores. This analysis showed that 83.9% of cases
could be correctly classified as having SFN using the discriminant coefficients
(Chi-square=38.93, p=0.037). Finally, we reduced the number of questions to 21
keeping the percentage correctly classified as having SFN at 83.9% (Chi-square=40.94,
p=0.006). This resulted in the Small Fiber Neuropathy Screening List (SFNSL)
questionnaire (see Appendix). Subsequently, a total score was made for the SFNSL by
summing the scores of the 21 questions. It appeared that a cut-off score of < 11 (25.8%
of patients) indicated only patients with normal TTT and a cut-off score of > 48 (18.9%
of patients) indicated only patients with abnormal TTT. A score range from 11 to 48
indicated a group of patients in which 33% had normal TTT and 67% had abnormal TTT
(Figure 8.1).

Page 7

The Small Fiber Neuropathy Screening List ⏐135














>48>48 >4837-48 37-4837-4824-36 24-3624-36 11-23 11-2311-23<11<11 <11
4040 40
30 3030
202020
1010 10
000
TTT TTTTTT
normal
normal
normalnormal
normal
abnormal
abnormal
abnormalabnormal
abnormal
SFNLS score SFNLS score
Number of patientsNumber of patients
Figure 8.1 Temperature threshold testing (TTT) results in a group of sarcoidosis pa-
tients with a small fiber neuropathy screening list (SFNSL) score below
11, in a group of sarcoidosis patients with a SFNSL score from 11 to 23,
from 24 to 36, from 37 to 48 and in a group of sarcoidosis patients with a
SFNSL score above 48.

Reliability and validity in Group 1
Exploratory factor analysis was employed to examine content validity. Herewith the
number of underlying factors measured by the questionnaire can be determined using
the screetest. The screetest criterion clearly showed that the SFNSL consisted of only
one underlying factor.
Furthermore, construct validity of the SFNSL was assessed. Construct validity is the
extent to which the SFNSL actually assesses what it is intended to assess.28 This is
examined by assessing the relationship of this questionnaire with other questionnaires.
For this purpose the relationship with the neuropathic pain scale (NPS) was examined.
Compared with patients without pain (mean SFNSL score=18.4, SD=12.1), patients who
indicated to have pain (mean SFNSL score=29.8, SD=15.4) scored significantly higher
on the SFNSL (t=-2.58, p=0.012). Furthermore, in patients with pain, the correlation
between pain at this moment (question of the NPS) and the total SFNSL score was 0.49
(p<0.001).
Internal consistency (Cronbach`s alpha) was assessed to find out to what extent the
different questions of the SFNSL were related with each other. The Cronbach’s alpha
was 0.89. Correlations between individual items (questions 1 through 21) and the total

Page 8

136⏐Chapter 8
score of the questionnaire (minus that item) were always significant and positive. The
item-total correlations ranged from 0.25 to 0.70.
Cross validation of the SFNSL in Group 2
The SFNSL was then completed by the following consecutive 55 patients (Group 2).
There were no missing values. The cut-off scores provided by Group 1 were also useful
in Group 2. Now, 19.4% of the patients had an SFNSL above 48. They all had an
abnormal TTT. The percentage of patients with a SFNSL score below 11 was 38.1%.
They all had a normal TTT.
Again exploratory factor analysis showed that the SFNSL measured one construct,
since the screetest clearly showed one factor. The Cronbach’s alpha was 0.90 and the
item-total correlations ranged from 0.29 to 0.72.
Total group
Cronbach`s alpha for the total scale was 0.90 and item-total correlations ranged from
0.32-0.67. Based on TTT results, sensitivity and specificity of the SFNSL was 100% and
31%, respectively, when a cut off score of 11 is used and 19% and 100%, respectively,
when a cut off score of 48 is used (figure 8.1).
Healthy controls
A All the 15 tested healthy controls had a normal TTT. Moreover, the SFNSL scores of
all of them were below 11.
Discussion
We have developed a short and easy to administer questionnaire to screen for small
fiber neuropathy (SFN) in sarcoidosis patients. The questionnaire was cross-validated in
an other sarcoidosis patient group. Cut-off scores of below 11 for certainly no SFN and
above 48 for certainly SFN were established based on temperature threshold testing
(TTT) results. The reliability and validity analyses revealed results that exceeded
minimum quality standards for an instrument of this kind. Internal consistency revealed
that this scale was highly unified, a conclusion supported by content validity assessment
that revealed that the SFNSL measured only one underlying factor. These results
strongly argue that SFN is a unified condition and they argue against a psychogenetic
component to the symptoms of SFN.

Page 9

The Small Fiber Neuropathy Screening List ⏐137
Strict TTT criteria were used to diagnose SFN. Both MLE and MLI test results had to
exceed the 99% value of a normal population to score TTT results as abnormal. These
sharp cut-off scores were used as we preferred a high specificity of TTT in order to be
relatively certain which patients do have SFN. Consequently, based on the present
results patients with a SFNSL score above 48 can almost certainly be suspected of
suffering from SFN. We strived to develop a screening list with the highest possible
sensitivity because SFN is a disorder that probably develops over time with increasing
symptom severity. Since a strict criterion for SFN was defined using TTT, it is possible
that a considerable number of patients are suffering from SFN-related symptoms, but do
not yet meet the TTT criteria. With these strict cut-off values it is possible to indicate
which patients certainly have no SFN (SFNSL score below 11) and definitive do have
SFN (SFNSL score above 48). In the intermediate group of patients 33% had normal
TTT and 67% had abnormal TTT. These patients should be carefully monitored and
tested for the presence of SFN. All 15 tested healthy controls had normal TTT results
and SFNSL scores were all below 11. Thus, based on the results in healthy controls and
the results in both groups sarcoidosis patients a SFNSL score below 11 seems to
exclude the presence of SFN.
One of the limitations that this and other studies have to face is that unfortunately a
proper gold standard to diagnose SFN is still lacking. SFN is clinically defined by
paresthesias (abnormal sensations), pain, and sometimes numbness, in combination
with normal strength and normal nerve conduction studies.9 A number of investigative
tools are now available for confirming the diagnosis. TTT reaches sensitivity of 60-85%,
and the test is validated and standardized.31,32 Epidermal nerve fiber density analysis
has been used extensively by several groups, and sensitivities of 74.0-87.5% are
reported.32 This latter method, however, is technically difficult and not widely available.
The sparse availability of diagnostic tests for SFN urges for an easy to administer
screening instrument. In the present study we used TTT to diagnose SFN, which, as
already mentioned, does not reach 100 percent sensitivity. Therefore, the diagnostic
value and cut-off scores of the SFNSL should be examined further in future studies,
using also other clinical tests that can be used to diagnose SFN, such as intra-
epidermal nerve fiber density assessment in skin biopsy, QSART, laser evoked
potentials and cardiovascular autonomic function testing.
In the present study only sarcoidosis patients participated. Presumably, the SFNSL is
also useful in idiopathic SFN or those due to other causes. However, this assumption
has to be examined.
As SFN has only recently gained more attention, exact data on prognosis, clinical
course and treatment efficacy are lacking. Furthermore, the condition may be easily
missed as symptoms of autonomic dysfunction may not always be recalled sponta-
neously by the patient. Moreover, even if reported, symptoms such as diarrhoea,

Page 10

138⏐Chapter 8
micturation disturbances and sweating may not always be recognised as such. In this
respect, recognition of SFN is important because it may prevent extensive investigations
such as colonoscopy in the case of diarrhoea or urodynamic investigation in case of
micturation disturbances. And finally, tests for assessment of SFN are not widely
available while routinely applied nerve conduction studies and EMG remain normal in
SFN patients. Moreover, the SFNSL can be easily used by physicians including
pulmonologists who often are involved in the follow-up of sarcoidosis patients and are
not so familiar with the clinical picture of SFN. Especially, the SFNSL might be helpful to
distinguish those patients without SFN from those with probably or highly likely SFN.
Therefore, the SFNSL is recommended as first screenings tool for various disciplines
seeing patients with possible SFN.
Conclusion
In conclusion, the SFNSL is aimed to be a practical tool in screening for the presence of
SFN. It is brief and offers the possibility of ready use in clinical practice, in epidemiologi-
cal and pathophysiological research, and in clinical trials.

Page 11

The Small Fiber Neuropathy Screening List ⏐139
References
1. Hunninghake GW, Costabel U, Ando M, Baughman RP, Cordier JF, du Bois RM, Eklund A,
Kitaichi M, Lynch J, Rizzato G, Rose C, Selroos O, Semenzato G, Sharma OP.
ATS/ERS/WASOG statement on sarcoidosis. American Thoracic Society/European Respira-
tory Society/World Association of Sarcoidosis and other Granulomatous Disorders. Sarcoido-
sis Vasc Diffuse Lung Dis 1999;16:149-73.
Wirnsberger RM, De Vries J, Breteler MHM, van Heck GL, Wouters EFM, Drent M.
Evaluation of quality of life in sarcoidosis patients. Respir Med 1998;92:750-6.
Sharma OP. Fatigue and sarcoidosis. Eur Respir J 1999;13:713-4.
James DG. Complications of sarcoidosis. Chronic fatigue syndrome. Sarcoidosis 1993;10:
1-3.
Newman LS, Rose CS, Maier LA. Sarcoidosis. N Engl J Med 1997;336:1224-34.
Hoitsma E, De Vries J, van Santen-Hoeufft M, Faber CG, Drent M. Impact of pain in a Dutch
sarcoidosis patient population. Sarcoidosis Vasc Diffuse Lung Dis 2003;20:33-9.
Hoitsma E, Marziniak M, Faber CG, Reulen JPH, Sommer C, de Baets M, Drent, M. Small
Fiber Neuropathy in Sarcoidosis. Lancet 2002;359:2085-6.
Hoitsma E, Drent M, Verstraete E, Faber CG, Troost J, Spaans F, Reulen JPH. Abnormal
warm and cold sensation thresholds suggestive of small-fibre neuropathy in sarcoidosis. Clin
Neurophysiol 2003;114:2326-33.
Hoitsma E, Reulen JPH, de Baets M, Drent M, Spaans F, Faber CG. Small fiber neuropathy:
a common and important clinical disorder. J Neurol Sci 2004;15;227:119-30.
10. Polydefkis M, Yiannoutsos CT, Cohen BA, Hollander H, Schifitto G, Clifford DB, Simpson DM,
Katzenstein D, Shriver S, Hauer P, Brown A, Haidich AB, Moo L, McArthur JC. Reduced
intraepidermal nerve fiber density in HIV-associated sensory neuropathy. Neurology 2002;
58:115-9.
11. Novak V, Freimer ML, Kissel JT, Sahenk Z, Periquet IM, Nash SM, Collins MP, Mendell JR.
Autonomic impairment in painful neuropathy. Neurology 2001;56:861-8.
12. Grootenhuis PA, Snoek FJ, Heine RJ, Bouter LM. Development of a type 2 diabetes
symptom checklist: a measure of symptom severity. Diabet Med 1994;11:253-61.
13. Dyck PJ, Davies JL, Litchy WJ, O'Brien PC. Longitudinal assessment of diabetic polyneuro-
pathy using a composite score in the Rochester Diabetic Neuropathy Study cohort. Neurology
1997;49:229-39.
14. Dyck PJ, Karnes JL, O'Brien PC, Litchy WJ, Low PA, Melton LJ. The Rochester Diabetic
Neuropathy Study: reassessment of tests and criteria for diagnosis and staged severity. Neu-
rology 1992;42:1164-70.
15. Dyck PJ, Litchy WJ, Lehman KA, Hokanson JL, Low PA, O'Brien PC. Variables influencing
neuropathic endpoints: the Rochester Diabetic Neuropathy Study of Healthy Subjects. Neu-
rology 1995;45:1115-21.
16. Dyck PJ, Kratz KM, Lehman KA, Karnes JL, Melton LJ 3rd, O'Brien PC, Litchy WJ,
Windebank AJ, Smith BE, Low PA, et al. The Rochester Diabetic Neuropathy Study: design,
criteria for types of neuropathy, selection bias, and reproducibility of neuropathic tests. Neu-
rology 1991;41:799-807.
17. Dyck PJ. Detection, characterization, and staging of polyneuropathy: assessed in diabetics.
Muscle Nerve 1988;11:21-32.
18. Dyck PJ, Karnes J, O'Brien PC, Swanson CJ. Neuropathy Symptom Profile in health, motor
neuron disease, diabetic neuropathy, and amyloidosis. Neurology 1986;36:1300-8.
19. Bril V. NIS-LL: the primary measurement scale for clinical trial endpoints in diabetic peripheral
neuropathy. Eur Neurol 1999;41(S1): 8-13.
20. Galer BS, Jensen MP. Development and preliminary validation of a pain measure specific to
neuropathic pain: the Neuropathic Pain Scale. Neurology 1997;48:332-8.
2.
3.
4.
5.
6.
7.
8.
9.

Page 12

140⏐Chapter 8
21. Yarnitsky D, Sprecher E. Thermal testing: normative data and repeatability for various test
algorithms. J Neurol Sci 1994;125:39-45.
22. Cattell RB. The scree test for the number of factors. Multivariate Behavioral Research 1966;1:
140-61.
23. Cronbach LPJ. Coefficient alpha and the internal structure of tests. Psychometrika 1951;16:
297-334.
24. Tabachnik BG, Fidell LS. Using multivariate statistics. New York, Harper Collins, 1996.
25. Gross-Porney L, Watkins MP. Foundations of clinical research: Applications to practice.
Upper Saddle River, NJ, Prentice Hall, 2000.
26. Campbell DT, Fiske DW. Convergent and discriminant validation by the multitrait-multimethod
matrix. Psychol Bull 1959;56:81-105.
27. Zechmeister JS, Zechmeister EB, Shaughnessy JJ. Essentials of Research Methods in
Psychology. New York: McGraw-Hill 2001:118-21.
28. Jamal GA, Hansen S, Weir AI, Ballantyne JP. The neurophysiologic investigation of small
fiber neuropathies. Muscle Nerve 1987;10:537-45.
29. Holland NR, Crawford TO, Hauer P, Cornblath DR, Griffin JW, McArthur JC. Small-fiber
sensory neuropathies: clinical course and neuropathology of idiopathic cases. Ann Neurol
1998;44:47-59.
30. Periquet MI, Novak V, Collins MP, Nagaraja HN, Erdem S, Nash SM, Freimer ML, Sahenk Z,
Kissel JT, Mendell JR. Painful sensory neuropathy: prospective evaluation using skin biopsy
[see comments]. Neurology 1999;53:1641-7.
31. Tobin K, Giuliani MJ, Lacomis D. Comparison of different modalities for detection of small
fiber neuropathy. Clin Neurophysiol 1999;110:1909-12.
32. Lacomis D. Small-fiber neuropathy. Muscle Nerve 2002; 26: 173-88.