Abstract
Objectives: To estimate reference data for the commonly performed sensory nerve conduction studies (NCS) using a cohort of healthy subjects from Saudi Arabia
Methods: This is a cross-sectional study conducted between May 2015, and June 2019. Sensory nerve action potential (SNAP) amplitude, conduction velocity (CV), and peak latency (PL) were recorded. Associations between these parameters and the covariates (age, sex, height, weight, and body mass index) were tested with Pearson correlations. Reference data were then derived using the lowest percentile that could be reliably determined for SNAP amplitude and CV. Reference data were derived using the highest percentile for PL.
Results: Upper and lower limb sensory NCS were performed in 127 and 137 participants, respectively. Age was the only covariate that required adjustment for estimation of SNAP amplitude. Therefore, a prediction model was generated for each nerve. Percentile estimation for PL and CV did not require adjustment for any of the covariates. Hence, it was derived for all the subjects pooled together.
Conclusion: The sensory NCS reference data were comparable to the data from other countries. However, minimal differences were observed. Further studies are required with a focus on the older age group. Received 2nd August 2019. Accepted 11th December 2019.
Nerve conduction studies (NCS) are commonly performed procedures during the diagnostic workup for peripheral nervous system disorders. Distinguishing normal NCS values from the abnormal values is a key to accurately interpreting the test, and to guiding further management. However, it requires comparison of patients’ NCS values with an established reference. Moreover, NCS values could vary in healthy individuals. Intrinsic factors that might affect NCS values include temperature, age, height, weight, body mass index (BMI), sex, and handedness.1⇓⇓–4 Conflicting evidence exists regarding the influence of ethnicity on the NCS values.5⇓⇓–8 Electrodiagnostic laboratories may use different protocols including filter setting and stimulation distance. Therefore, it has been suggested that the individual laboratories generate reference data for their own NCS protocol.9,10 However, it is a daunting process methodologically and statistically. In reality, most laboratories use reference data derived elsewhere, which is considered reasonable when the same technical standards as those from where the reference data were generated, are employed.11,12 However, it might not be suitable to use the reference data derived from the Western populations for Arabs living the Gulf region. We are not aware of any NCS reference data published from Saudi Arabia until now for the commonly performed nerve conduction studies.
In the present study, we reported the reference data for sensory NCS parameters. This study aimed to estimate the reference data for the commonly tested sensory nerves in healthy Saudi Arabian adults, and to identify the influence of age, sex, height, weight, and BMI.
Methods
Participants and setting
The study was conducted at King Saud University Medical City, Riyadh, Saudi Arabia between May 1, 2015 and June 30, 2019. The details of the inclusion and the exclusion criteria are listed in the recently published article about motor NCS reference data in the same journal.13 We included healthy adults who had no known current or previous neurological diseases, systemic diseases, or neurological symptoms. Neurological examination was performed for participants aged ≥ 50 years to exclude asymptomatic sensory deficit. Participants were recruited from the clinic waiting areas. We focused on recruiting patients’ family members, hospital staff, and medical students. Since the King Saud University Medical City is a tertiary hospital and accepts referral from the rural areas, a considerable number of the participants were from the rural areas and from different tribes.
NCS protocol
In our laboratory, NCS are performed following the standardized techniques published elsewhere.3,14,15 A trained technician with more than 20 years of experience in the field performed all the studies. Dr. MHA and Dr. NMK assessed the quality of the studies and reviewed the NCS waveforms. All the sensory studies were performed antidromically with the exception of the mixed palmar study, which involved orthodromic recording. The median nerve was stimulated at the wrist between the tendons of the flexor carpi radialis and the palmaris longus and was recorded slightly distal to the metacarpo-phalangeal joint of D2 (index finger) and D4 (ring finger). The ulnar nerve was stimulated just lateral to the flexor carpi ulnaris tendon and was recorded just distal to the metacarpo-phalangeal joint of D4 and D5 (little finger). The superficial radial nerve was stimulated on the distal radius and was recorded at the base of the thumb over the anatomical snuffbox. The medial antebrachial cutaneous nerve (MAC) was stimulated at the midpoint between the biceps tendon and the medial epicondyle and was recorded on the medial side of the forearm. The lateral antebrachial cutaneous nerve (LAC) was stimulated just lateral to the distal biceps tendon and was recorded on the lateral side of the forearm. The sural nerve was stimulated slightly lateral to the midline of the calf muscle and was recoded posteroinferior to the lateral malleolus. The superficial fibular nerve was stimulated on the anterolateral part of the lower leg and was recorded between the tendon of the tibialis anterior and the lateral malleolus. A distance of 14 cm was maintained between the stimulating cathode and the recording electrode for the median, the ulnar, the superficial fibular, and the sural nerves. A distance of 10 cm was maintained for the superficial radial, the MAC, and the LAC. A distance of 4 cm was maintained between the recording and the reference electrodes for all the sensory nerves. For the mixed palmar orthodromic study, the ulnar and median nerves were stimulated over the medial and lateral sides of the palm, and were recorded 8 cm proximally over the anatomic sites of the ulnar and the median nerves, just proximal to the wrist joint, respectively. Hand temperature was maintained at ≥ 32 °C and foot temperature was maintained at ≥ 30°C. All the sensory nerve action potentials (SNAP) were measured after a supramaximal stimulation had been achieved. The recorded parameters included the SNAP onset-to-peak amplitude, the peak latency (PL), and the conduction velocity (CV).
Instrument setting
The NCS were performed using Nicolet Viking version 11.1 (VIASYS Healthcare Inc., USA). Low and high frequency filters were set at 20 Hz and 3 kHz, respectively. Sweep speed was set at 2 milliseconds/division (ms/div). Gain was set at 10 microvolts (µV)/div. The study was approved by the institutional review board at the King Saud University Medical City. Signed informed consent was obtained from all the participants
Analysis
Descriptive statistics were used to summarize the data. Pearson correlation coefficients were calculated to evaluate the correlations of age, sex, height, weight, and BMI with the SNAP amplitude, the PL, and the CV. Quantile regression analyses were employed on the log-transformed data to identify the covariates that significantly contributed to the variance in the SNAP amplitude and the CV for each nerve. The goal was to determine and adjust for the covariates that showed significant associations with the NCS parameters of the related nerves. The covariates with inconsistent statistical significance across the NCS parameters of the related nerves may have been subjected to numerical artifacts rather than variations in the biology.16 A p-value <0.05 was considered statistically significant. We adjusted our reference data for age as recommended by the Normative Data Task Force (NDTF) of the American Association of Neuromuscular and Electrodiagnostic Medicine.12 We computed the third percentile for the SNAP amplitude and the CV, and the 97th percentile for PL. The 95% confidence intervals for these percentiles were generated to provide estimates of the upper and the lower reference limits, as deemed appropriate. Data were analyzed using Stata software version 12 (Stata Corp., College Station, Texas, USA).
Results
Upper and lower limb sensory NCS were performed in 127 and 137 participants, respectively. The number of participants was variable for each nerve. The characteristics of the study participants are identical to that listed in the recently published article for the motor NCS reference data (Table 1).13 A summary of the sensory NCS responses is presented in Table 2. The predicted reference data for the SNAP amplitude, the PL, and the CV are shown in Table 3. Correlations between the covariates (sex, height, weight, and BMI) and the NCS parameters were generally weak to moderate (Online Supplementary Table S1). Except age, none of the other covariates showed constant association with the sensory NCS parameters of related nerves in the quantile regression models (Online Supplementary Table S2). BMI, height, and sex were significantly associated with the ulnar SNAP amplitude, but not with the median SNAP amplitude, suggesting that these associations were probably due to numerical artifacts rather than due to differences in biology of the tested nerves. The MAC and the LAC SNAP amplitude, and the sensory CV and PL of all the nerves were not influenced by age, or by the other covariates. Therefore, the percentile estimations for these parameters were reported for the all participants pooled together.
The characteristics of the study participants are identical to that listed in the recently published article for the motor NCS reference data.13
Summary of the sensory NCS data presented as mean ± standard deviation.
Reference values for the sensory NCS parameters in the upper and the lower limbs.
For the SNAP amplitudes in which age contributed significantly to their prediction model, we estimated reference values for ages 20, 40, and 60 years (Table 3). The regression coefficients generated in the quantile regression model for the third percentile can be used to estimate the log (predicted SNAP amplitude) for other ages. For example, the predicted third percentile of the ulnar SNAP amplitude for a 50-year-old subject would be estimated as follows:
log (ulnar SNAP amplitude) = β0 + β1 * age; where β0 is the constant coefficient and β1 is the coefficient for age.
= 4.125 + (– 0.027) * 50
= 4.125 – 1.35
= 2.775
Hence:
ulnar SNAP amplitude = exp (2.775) = 16 µV
The predicted lower limit of the normal amplitude for the third percentile can also be estimated using β0 and β1 of the lower bound of the 95% confidence intervals (Table 3). The effect of age on the SNAP amplitudes was variable for different nerves. Comparing the SNAP amplitudes for the age of 60 years with those for the age of 20 years, it was observed that the SNAP amplitudes decreased by approximately 50-80% for the sural and the superficial fibular nerves, 67-75% for the ulnar nerve, 30-46% for the median nerve, and 24-34% for the superficial radial nerve.
Discussion
The present study followed the recommendations of the NDTF. More than 100 healthy participants were recruited for the majority of the nerves to support the reliability of the generated data.12 We presented our data as mean±2 standard deviations and derived the cut-off values using percentiles. Researchers have discouraged the use of mean ± 2 standard deviations to generate NCS reference data. This statistical approach is hampered by the inherent skewness (non-Gaussian distribution) of the NCS values, whereby the lower limit of normal (e.g., for amplitude) would not be meaningful in a heavily right-skewed curve. The same holds true for the upper limit of normal (e.g., for latency) in a heavily left-skewed curve.17 On the contrary, percentile analyses produce a more reliable reference data regardless of the shape of the data distribution curve. Thus, it has been considered a preferred method for NCS reference data estimation.12
The reference data generated in this study were comparable to those of previous studies that fulfilled the NDTF criteria.3 However, minor differences were observed in the generated data for some of the NCS parameters. The 97th percentile reference limit for the PL of the median, the ulnar, the radial, and the sural sensory nerves were 3.5 ms, 3.5 ms, 2.6 ms, and 4.3 ms, respectively. These latencies are slightly shorter than the previously reported values of 4.0 ms, 4.0 ms, 2.8 ms, and 4.5 ms, respectively.1,2,18,19 Direct comparison of the SNAP amplitude was not considered feasible as we used a prediction model with age as a covariate, while the previous reports were stratified for different age ranges. However, the SNAP amplitudes of the median, the ulnar, the radial, and the sural nerves in this study had higher mean amplitudes, resulting in slightly higher cut-off values. The antebrachial cutaneous nerves are not commonly studied in the electrodiagnostic laboratory. However, their assessment is crucial, especially when a lesion of the brachial plexus if suspected. For the MAC nerve, the PL was 2.7 ms, and the amplitude was 4.0 µV. These values are almost identical to the values of 2.6 ms and 4.0 µV reported by Prahlow and Buschbacher.20 For the LAC nerve, the PL was 2.5 ms, and the amplitude was 8.0 µV. The reported reference limits for the LAC nerve by Buschbacher et al21 were 2.5 ms, and 5.0 µV, respectively.
Notably, we implemented the same NCS protocol described in the previous studies. The differences in the results may be in part due to the younger age, the disproportionate sex ratio, and the lower average height of our study population. Height, male gender, and age have been shown in some studies to have positive associations with sensory distal latency.22,23 Fujikama et al24 reported that women have greater SNAP amplitudes in the upper limbs than men. On the other hand, Stetson et al. did not find any correlations between sex and NCS parameters except the ones that could be explained by physiological factors such as height and finger circumference.23 Salerno et al25 observed that, after adjusting for finger circumference, women still had higher median SNAP amplitudes. Ethnicity might have contributed to the differences seen in the present study as all the participants in this study were Arabs. However, the role of ethnicity might be controversial.
The 97th percentile of the PL difference between median D2-ulnar D5, median D4-ulnar D4, and median-ulnar mixed palmar studies were 0.4 ms, 0.4 ms, and 0.3 ms, respectively. These tests, in addition to the lumbrical-interossei study (not assessed in the present study), are important for evaluating patients with carpal tunnel syndrome. Our findings are consistent with the median-ulnar digit SNAP comparison study reported by Grossart et al26 and the median-ulnar mixed palmar PL difference reported by Buschbacher.27 Another study reported a slightly larger median-ulnar D4 PL difference (0.5 ms) than the values in our study.28
In this study, all the SNAPs of the tested nerves were recordable except 6% of the superficial fibular SNAPs were absent. While unrecordable potentials from the sural nerve have been reported in 2.6% of healthy subjects, it seems that this phenomenon is more frequently encountered with the superficial fibular nerve (8%).29 It is not known whether an unrecordable potential reflects a truly absent response, or whether the response is present but is below the threshold of detection. Although we did not plan to investigate this a priori, further analysis of our data revealed that patients with absent superficial fibular SNAP had a higher mean BMI (36.3±8.7) than those with recordable potentials (27.5±5.7, p<0.001). Therefore, it is possible that failure to record the superficial fibular potentials was due to physical factors such as adipose tissue interfering with the ability to achieve supramaximal nerve stimulations and to record the provoked potentials. As suggested by the NDTF,12 participants with absent responses in this study were not included in the analysis. Caution is still required while interpreting the NCS values of a nerve that could have an unrecordable response in healthy subjects.
Among the covariates, only age had a significant negative association with the SNAP amplitudes. The effect was seen in all the studied sensory nerves except the MAC and the LAC. We did not find any significant association between sex (adjusted for height) and sensory NCS parameters. These findings are consistent with the findings of the study by Stetson et al23 but not consistent with the findings of other studies.7,22,24,25
One limitation of this study is that the side-to-side differences were not accounted for. This could potentially affect the accuracy of the reference values, since side-to-side differences have been reported in several publications including those with absent responses.1,2,19 Although we did not find a correlation between sex and any of the NCS parameters, the relatively small number of male participants (≈30%) could have potentially influenced our conclusion. Moreover, the older age group was underrepresented due to difficulties with recruitment. The study had several strengths. It is the first study of normative sensory NCS data in Saudi Arabia. The study had an adequate total sample size of more than 100 healthy adults. NCS were performed following the standardized protocol of previous studies. The data were reported using percentiles and the covariates were accounted for. Future studies should focus on obtaining the NCS reference data in the older age group.
Supplementary Table 1
Correlations of age, sex, height, weight, and BMI with sensory nerve action potential amplitude, peak latency, and conduction velocity.
Supplementary Table 2
Quantile regression analysis output (3rd percentile unless otherwise indicated).
1. Median SNAP amplitude
2-Ulnar SNAP amplitudes
3-Superficial radial SNAP amplitude
1. Median antebrachial SNAP amplitude
1-Lateral antebrachial SNAP amplitude.
2-Sural amplitude
3-Superficial fibularamplitude
.qreg supP_log age, quantile (3)
Iteration 1: WLS sum of weighted deviations = 29.298231
Iteration 1: sum of abs. weighted deviations = 34.19511
Iteration 2: sum of abs. weighted deviations = 16.343988 Iteration 3: sum of abs. weighted deviations = 7.5662585 Iteration 4: sum of abs. weighted deviations = 6.7079784 Iteration 5: sum of abs. weighted deviations = 6.6780002
4-Conduction velocity (97rd percentile unless otherwise indicated) Median
95th percentile
5-Ulnar
6-Superficial radial
7-Sural
8-Superficial fibular
Footnotes
Disclosure. Authors have no conflict of interests, and the work was not supported or funded by any drug company.
- Received August 2, 2019.
- Accepted December 11, 2019.
- Copyright: © Neurosciences
Neurosciences is an Open Access journal and articles published are distributed under the terms of the Creative Commons Attribution-NonCommercial License (CC BY-NC). Readers may copy, distribute, and display the work for non-commercial purposes with the proper citation of the original work.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.