We investigated the application of rate-dependent depression (RDD) of the Hoffmann (H) wave as a predictor of treatment efficacy in patients with painful diabetic peripheral neuropathy (DPN). General medical information, scales, and nerve conduction data were collected from 73 healthy subjects, 50 subjects with type 2 diabetes and painless DPN, and 71 subjects with type 2 diabetes and painful DPN. The left tibial nerve was stimulated, and RDD was calculated by the decline in amplitude of the third H wave relative to the first one. Gabapentin treatment was initiated after baseline evaluation, and the RDD and visual analog scale (VAS) score were both evaluated regularly during the 2-week study period. At baseline, the painful DPN group exhibited significant RDD impairment across all stimulation frequencies. Gabapentin treatment significantly reduced the VAS score and restored RDD during the 2-week observation period. RDD was found to be an independent factor of minimal VAS score improvement, such that the benefit increased by 1.27 times per 1% decrease in the RDD value. In conclusion, this study demonstrates that diabetes-induced loss of RDD can be modified by gabapentin and suggests that RDD may be valuable for predicting the initial efficacy of gabapentin therapy in patients with painful DPN.

Diabetic peripheral neuropathy (DPN) is one of the most common complications of diabetes and is detected in 30–50% of patients with diabetes (1,2). Sensory complaints, especially so-called positive symptoms (tingling, burning, shock-like, etc.), are reported by up to one-third of patients with diabetes and markedly impede quality of life (3). Authoritative guidelines recommend medications such as gabapentin, pregabalin, and duloxetine as first-line treatments for painful DPN (46). However, whether a drug can relieve pain in a particular patient is unpredictable, and it has been argued that many reviews and guidelines overestimate the efficacy of these treatments (7), while relatively few people who use gabapentin or pregabalin achieve satisfactory pain relief (8). A meta-analysis showed that only approximately one in six patients with painful DPN receive “substantial benefits” (9), and in most clinical studies, the number of patients needed to treat for 50% pain relief is generally four or five or higher (1012). In other words, drugs are effective in only ∼20% of patients. This unpredictability may be due to the wide variety of mechanisms that have been proposed to precipitate pain in patients with diabetes (13).

Rate-dependent depression (RDD), a measurement of changes in the amplitude of the Hoffmann reflex (H-reflex) over consecutive stimulations, has been extensively studied in both humans and rodents since the 1960s (14). As a presumed monosynaptic reflex, loss of RDD has been reported in patients with spinal cord injury and poststroke spasticity as well as in physical exercise assessments (1518). Loss of RDD is also emerging as a potential biomarker for spinal involvement in patients with painful diabetic neuropathy (19) and may identify patients with pain caused by disinhibition of the spinal cord due to a shared pathogenic mechanism involving disrupted chloride homeostasis and modified γ-aminobutyric acid (GABA) receptor function (20). In the current study, we determined RDD values in subjects with type 2 diabetes and neuropathic pain and investigated the hypothesis that loss of RDD is predictive of the initial efficacy of gabapentin, a widely used antineuropathic agent proposed to work at the spinal level.

Ethics and Subjects

This single-center prospective study was performed according to the principles outlined in the Declaration of Helsinki and was approved by the Institutional Review Board of Renji Hospital, Shanghai Jiaotong University School of Medicine. Written informed consent for data and sample collection was obtained from the subjects on admission.

All subjects were individuals who visited, or were admitted to, Renji Hospital, Shanghai Jiaotong University School of Medicine between 2020 and 2021. Patients in the DPN group met the type 2 diabetes criteria of the American Diabetes Association (21). In addition, a further DPN diagnosis of small-fiber or large-fiber distal symmetric polyneuropathy was made using the American Diabetes Association guidelines on DPN (22) to determine the presence of typical distal symmetric sensory and motor symptoms as well as either abnormal nerve conduction study (NCS) values or low intraepidermal nerve fiber density (IENFD). DPN patients with pain and visual analog scale (VAS) scores ≥3 were assigned to the painful DPN (p+DPN) group, while patients without complaints of pain were assigned to the painless DPN (p−DPN) group. The exclusion criteria were as follows: 1) insulin or analgesic taken in the past month; 2) diabetic ketoacidosis or hyperosmolar nonketotic diabetic coma in the past month; 3) severe cardiorespiratory, liver, or renal function insufficiency; 4) cancer, autoimmune, or other metabolic disease, or other factors of neurological impairment (alcohol abuse, toxin or neurotoxin exposure); 5) history of gabapentin allergy; 6) pregnancy; and 7) refusal to participate. Healthy subjects were selected from volunteers who met the following criteria: 1) no neurological complaints such as weakness, numbness, or pain; 2) no positive neurological physical examination signs; and 3) no family history of neuromuscular disorders. Exclusion criteria nos. 3, 4, 6, and 7 for the DPN groups were also applied to screen the healthy subjects. Additional rejection criteria for all groups included the following: 1) an absent or poor H-reflex waveform or 2) L5-S1 spinal or radicular compression detected by lumbar MRI scan. Among the subjects meeting the abovementioned criteria, 72 patients with DPN were rejected for poor or absent H response, so a total of 50 p−DPN patients, 71 p+DPN patients, and 73 age- and sex-matched healthy subjects were assigned to the appropriate groups.

Clinical Data Acquisition

The sex and height of all subjects were recorded. A detailed clinical history was collected to obtain information about the onset and duration of disease and pain and to determine neurological manifestations of weakness, numbness, and other symptoms. Every subject then underwent meticulous neurological physical examinations to determine positive neurological signs. For patients with DPN, the Michigan Diabetic Neuropathy Score (MDNS), Michigan Neuropathy Screening Instrument (MNSI), and VAS score were evaluated by a designated trained physician who was blinded to the study. Fasting blood glucose (FBG), 2-h postprandial blood glucose (2hPG), and hemoglobin A1c (HbA1c) were also monitored during the study.

Skin Innervation

Determination of IENFD was performed for all subjects with DPN. A 3-mm punch was used to obtain skin tissue from 10 cm above the left ankle. Protein Gene Product 9.5 (PGP9.5; no. MCA4750GA, Bio-Rad, Hercules, CA) was used to immunohistochemically stain 50-μm-thick tissue sections, which were subsequently observed by bright-field microscopy. Fibers were identified and counted using a previously published clinical method (23). At least four sections, excluding the first and last sections, were counted and used to calculate mean IENFD (number of fibers/mm). Because we did not collect skin biopsy specimens from control subjects, we adopted the previously established stratified normal ranges in our laboratory (lower limit of normal: 8.1 fibers/mm for 30–39 years old; 7.2 fibers/mm for 40–49 years old; 6.0 fibers/mm for 50–59 years old; 5.3 fibers/mm for 60–69 years old; and 4.7 fibers/mm for 70–79 years old) to determine the presence of small-fiber neuropathy as part of the study inclusion criteria.

Nerve Conduction and Needle Electromyogram Studies

All subjects underwent a nerve conduction test using a Keypoint.NET electromyograph (no. 9033A07, Natus, Middleton, WI) in a quiet room with a temperature >20°C. The operator was blinded to the patient diagnosis and study design. The limb temperature was maintained by a lamp warmer at ∼32°C. Surface electrodes (SEAg-J-22 × 32, Friendship Medical, Shanxi, China) were used to record the waveforms of fully relaxed muscles after stimulation. The stimulus intensity (duration: 0.2 ms) was gradually increased until the maximum amplitude of the elicited wave was reached (supramaximal stimulation). The distal motor latency (DML), compound muscle action potential (CMAP), and sensory nerve action potential (SNAP) were recorded. The CMAPs elicited at the distal stimulation point (dCMAP) were included. Median, ulnar, radial, peroneal, peroneal superficial, tibial, and sural nerves on either side were selected for evaluation. Multiple sites were stimulated to ensure that sufficient data were collected for each nerve and to calculate the motor conduction velocity (MCV) and sensory conduction velocity (SCV). The minimum F-wave and H-reflex latencies were measured by a standard stimulation method. The normal values and ranges established in our laboratory were adopted so that neurophysiology data could be used as part of the neuropathy inclusion criteria for individual study candidates, while group values were compared against the healthy subject group values to confirm the cohort neuropathy phenotype. A needle electromyograph test was administered to the DPN subjects to exclude other potential neuromuscular disorders.

RDD

RDD was measured with the subject in the prone position using the tibial nerve, which was stimulated using a surface electrode placed at the popliteal fossa with a recording electrode placed on the soleus muscle (middle posterior leg). The filter was set from 20 Hz to 10 kHz. A manual train of increasing monophasic stimuli (duration: 0.5 ms) with a gain of ∼1 mA or less was applied to identify the appropriate intensity that elicited the maximum H amplitude (Hmax stimulation intensity). Then, trains of five consecutive stimulations with the determined intensity were applied at ascending frequencies of 0.5 Hz, 1 Hz, 3 Hz, and 5 Hz. RDD was calculated by the decline in amplitude from the first to the third waveform (19).

Study Design

The clinical data of subjects who met the inclusion criteria of painful DPN were collected, and screening examinations and evaluations were performed to ensure that the subjects did not meet the exclusion criteria after written informed consent was obtained. The enrolled subjects underwent electrophysiological examinations (NCS and RDD at multiple frequencies) during the first visit (V0). Clinical scales (MNSI, MDNS, and VAS) and blood glucose indicators (FBG, 2hPG, and HbA1c) were also assessed. Basic medication regimens for DPN subjects included the following: 1) the patient’s current glycemic regimen and 2) vitamin B treatment consisting of methyl cobalamin (oral; 3 times a day, 1 tablet/0.5 mg each time; Eisai, Tokyo, Japan) and vitamin B1 (oral; 3 times a day, 1 tablet/10 mg each time; Xinhuanghe, Shanghai, China). Subjects in the p+DPN group then began a 2-week treatment period and follow-up. The intervention medication for the p+DPN subjects was gabapentin (oral, 3 times a day, 1 tablet/0.3 g each time; Selection, Hainan, China). The second visit (V1) and the third visit (V2) were 2 days and 1 week after V0, respectively, and RDD (1 Hz) and VAS assessments were performed. The fourth visit (V3) was 2 weeks after V0. Electrophysiological measurements (NCS and RDD at 1 Hz), clinical scales (MNSI, MDNS, and VAS), and blood glucose indicators (FBG, 2hPG, and HbA1c) were assessed.

Statistical Analysis

Normally distributed data are expressed as the mean ± SD, while data with a skewed distribution are expressed as the median (25% percentile–75% percentile). We used χ2 tests with Bonferroni correction to analyze the proportion and composition of different groups. Group t tests were used for comparisons of data with a normal distribution and a standard homogeneity of variance; otherwise, the Satterthwaite t test or Wilcoxon rank sum test were used. Data with a normal distribution and homogeneity of variance between random groups were compared by one-way ANOVA, and the Šidák method was used for pairwise comparison, while two-way ANOVA was used for comparisons of the variance between multiple paired groups. Data with a skewed distribution or heterogeneity underwent a rank conversion before ANOVA. The RDD values obtained under multiple frequencies were used to graph the receiver operator characteristic (ROC) curves to determine the area under the curve (AUC) and optimum values. Logistic regression was used to calculate and analyze the relative risk (odds ratio [OR]), 95% CI, and P value between selected factors and the effect variable. The values of the effect variable were binarily classified based on whether the patient was responsive or not. All statistical analyses were performed by using Stata 16.1 software (StataCorp, College Station, TX), and P < 0.05 was considered to indicate significant between-group differences. The statistically significant P value cutoff in multiple (n) comparisons was adjusted by the formula 1 − 0.051/n.

Data and Resource Availability

Data generated and/or analyzed in the current study are available from the corresponding author on reasonable request.

Clinical Information and Baseline NCS

The clinical information and nerve conduction findings are summarized in Table 1. Sex proportion, age, and height were matched among groups (P > 0.05). The p+DPN group exhibited a longer disease duration (median 11 years) than the p−DPN group (P = 0.002), while the differences in multiple diabetes scales and blood glucose evaluations were not significant (P > 0.05). A significant reduction in dCMAP, SNAP, and conduction velocity, accompanied by prolonged DML and F latency, were observed in both DPN groups compared with the healthy group (P < 0.01), while the reciprocal NCS comparisons between the p−DPN and p+DPN groups were not significant.

Skin Nerve Assessment

Subjects in the p+DPN group had an average IENFD of 4.7 fibers/mm (Table 1), which contributed to the identification of 63.3% (45 of 71) of recruited patients with a diagnosis of small- or mixed-fiber neuropathy (see research design and methods for laboratory normative values). The p−DPN group showed a similar average density of intraepidermal nerves (4.8 fibers/mm, P > 0.05).

Baseline H-Reflex Studies and RDD

Detailed electrophysiological studies on the H-reflex of the tibial nerve are summarized in Table 2. The p+DPN group had a significantly prolonged H latency (P < 0.001) and a higher Hmax stimulation intensity (P = 0.027) than the other groups. H1 amplitudes showed no significant difference between groups at most frequencies. The median RDD was significantly different among the groups at all stimulation frequencies studied (Fig. 1A). The healthy and p−DPN groups (P > 0.05) showed an ∼20–40% decline in H amplitude during consecutive stimulations, while RDD in the p+DPN group ranged from 10 to 30%, indicating a significant difference (P < 0.01). Typical RDD traces from the healthy, p−DPN, and p+DPN groups are illustrated (Fig. 1B).

ROC Curve

ROC curves were established using RDD values obtained at multiple frequencies for the p+DPN group versus the p−DPN group (Fig. 2). RDD at 1 Hz demonstrated the highest (P < 0.001) AUC value of 0.90 (95% CI 0.84–0.97, optimum cutoff: 28.3%) when these two groups were included.

Gabapentin Treatment and Follow-up

Two weeks of gabapentin treatment did not have any effect on diabetes or electrophysiological indices, including Hmax stimulation intensity or H1 amplitude, compared with pretreatment values (Table 3). The VAS score improved during gabapentin treatment, with significantly lower scores at all time points (Fig. 3A). The calculated needed to treat for 50% pain relief in this cohort was five. A significant (P = 0.002) improvement in RDD was observed between baseline (17.9% [25th–75th percentile 12.2–23.0]) and 2 weeks (25.4% [10.1–40.3]) of treatment (Fig. 3B).

Baseline indicators (diabetes duration, pain duration, FBG, 2hPG, HbA1c; IENFD, DML, dCMAP, MCV, F latency, H latency, and RDD of the left tibial nerve) were selected for further analysis as potential factors that might influence the efficacy of gabapentin treatment, which was set to a minimal (1-point) improvement in the VAS score. Subjects were grouped based on the presence (n = 59 [83.1%]) or absence (n = 12 [16.9%]) of the minimal VAS score increase, indicating a response to the treatment, and the abovementioned indicators were compared. RDD emerged as the only factor showing a difference between groups (P < 0.001) (Table 4). As a result of logistic regression, RDD demonstrated a significant influence on efficacy, with a P value of 0.014 for the single factor, a P value of 0.019 for the model, and an OR of 0.79 (95% CI 0.65–0.95), while the influences of other factors were not significant. The forest plot for the OR of multiple factors is shown in Fig. 4.

Current treatments for neuropathic pain work well in some patients but less so in others. The reasons for this diversity are not known but may reflect variability in the underlying pathogenic mechanisms, despite patients presenting with the same precipitating disease and pain symptoms. Recognition of this complexity has led to interest in a personalized medicine approach, for which it would be helpful to have biomarkers of the pain mechanism(s) present in individual patients to predict the relative efficacy of current and future analgesics. For example, previous work proposed that oxcarbazepine, a drug that is suspected to act by targeting sodium channels, would be most efficient in rodents and humans with an “irritable nociceptor” phenotype identified using quantitative sensory testing and electrophysiological techniques (24,25). Similarly, subjects with spontaneous peripheral nerve activity identified by microneurography were selectively included in a clinical trial of a T-type calcium channel antagonist against painful diabetic neuropathy (26). Loss of RDD has been proposed as a biomarker of pain in subjects with diabetes mediated by spinal disinhibition (19), but this idea has yet to be tested in terms of its predictive value for efficacy of spinally targeted therapies.

RDD may be measured in humans using a simple modification of the standard electrophysiological techniques used to measure nerve conduction parameters (14). In the current study, this is illustrated by the 30–40% drop in H wave amplitude following repeated peripheral nerve stimulation in the healthy subject group. Pharmacological studies in rodents indicate that RDD is mediated by GABAA receptors, whose inhibitory function is maintained by the activity of the K+-Cl co-transporter 2 (KCC2) membrane pump (27). In rat models of type 1 diabetes, spinal GABAergic inhibitory function is impaired secondary to a reduced spinal expression of the KCC2 pump (28). Reduced spinal KCC2 expression and subsequent loss of GABAergic inhibitory function impacts both RDD and concurrent behavioral indices of neuropathic pain such as allodynia to light touch and hyperalgesia in the paw formalin test (28). Interestingly, pharmacological interventions that restored RDD in diabetic rats also alleviated allodynia. The clinical relevance of these findings were subsequently confirmed by the identification of diminished RDD in cohorts of subjects with diabetes with painful diabetic neuropathy (19,29,30), prompting the suggestion that loss of RDD and painful diabetic neuropathy may share a common pathogenic mechanism in the spinal cord and that loss of RDD may serve as a biomarker for spinally mediated pain in diabetes (19).

In order to define optimal measurement parameters for identifying individual subjects with impaired RDD, we measured RDD over a range of stimulus frequencies. While all frequencies tested identified impaired median RDD in subjects with neuropathy and pain compared with those with painless subjects, 1 Hz proved to be the most efficient in discrimination due to its highest AUC. This agrees with a recent study in subjects with type 1 diabetes (19). Despite a clear impact of diabetes on various routine indicators of neuropathy, such as large-fiber electrophysiological function and small-fiber density in the skin, only the presence or absence of pain, significantly impacted RDD. Prior clinical studies in subjects with type 1 and type 2 diabetes also showed that RDD values were only impaired in the cohort with concurrent neuropathy and neuropathic pain and not in a cohort of subjects with diabetes with matched neuropathy but no pain (19,29,30). It is notable that within cohorts of subjects with type 1 (19) and type 2 (Fig. 1A) with painful diabetes, only a subset of RDD values fall outside the normative range of control subjects or subjects with diabetes with neuropathy but no pain. It has been proposed that individual subjects with abnormal RDD may have pain of spinal pathogenesis (19,29,30). RDD values of subjects with diabetes may therefore be useful as a potential indicator in identifying those patients with painful DPN who may have an optimal response to therapies that target spinal function (30).

Once impaired RDD was identified in our cohort of subjects with type 2 diabetes with neuropathy and pain, we were interested in the impact of initiating treatment with gabapentin. Peak plasma gabapentin concentration is generally attained 2–3 h after administration (31), with steady state reached in 2 days, while initial clinical efficacy against neuropathic pain is usually significant in 1–2 weeks (32,33). Therefore, in addition to pretreatment baseline values, we also collected data at 2 days, 1 week, and 2 weeks after onset of treatment to assess the initial period of pain relief, as gabapentin has been reported to achieve significant reductions in VAS score over this time period (34,35). As anticipated, gabapentin treatment significantly improved pain, indicated by VAS scores, without impacting blood glucose levels, nerve conduction velocity, or any clinical index of diabetic neuropathy. More surprising was the observation that gabapentin treatment also significantly restored RDD. Despite widespread use, it remains unclear how gabapentin acts in alleviating neuropathic pain. Gabapentin was designed as a GABA-mimetic but does not have a high infinity for either the GABAA or GABAB receptors and does not directly influence GABAAR function (36). It does inhibit the axonal transport of the α2δ-1 subunit of voltage-gated calcium channels (VGCC), and this is widely accepted as one mechanism by which it alleviates neuropathic pain via a presynaptic modulation (37). Other hypotheses include modulation of transient receptor potential channels, the N-methyl-d-aspartic acid (NMDA) receptor, protein kinase C, and inflammatory cytokines along with upregulating the expression of the GABAAR δ subunit (38,39). As a reflection of the complete reflex arc, RDD potentially represents modulation of multiple pre- and postsynaptic sites. How such actions are fulfilled awaits further investigation.

In the current study, the capacity of gabapentin to promote pain relief (40,41) correlated with its impact in restoring RDD. Thus, when reviewing baseline characteristics of subjects, we observed that patients in whom 2 weeks of gabapentin was ineffective for pain relief tended to also present with less marked RDD impairments. This could indicate that the pain of DPN subjects with markedly impaired RDD is predominantly spinal driven and could be alleviated by a treatment that impacts spinal nociceptive processing. Regression analysis provided further evidence to support this presumption and indicated that the difference in therapeutic benefit between individuals with varying RDD values could be predicted; the benefit increases by 1.27 times (1/OR) per 1% decrement in the RDD value. Other than sensitization and disinhibition at the spinal cord level, both peripheral sensitization (42) and intracranial sensitization (43) may concurrently initiate or amplify pain states, and further complexity may arise from pain chronification, whereby the dominant pain generator sites evolve over time and with increasing disease stage (44). The utility of specific treatments may therefore depend not only on the capacity to target specific pain mechanisms but also on a specific time window as pain generator mechanisms evolve. Owing to its minimally invasive nature, RDD has the potential to serve as an iterative biomarker that identifies a site of pain generation at a specific time in the evolution of pathogenic mechanisms to guide therapeutic selection.

In conclusion, we found significant impairment in RDD in patients with type 2 diabetes and painful DPN. Both the pain VAS score and impaired RDD were improved after 2 weeks of gabapentin treatment. Patients with painful DPN are likely to achieve better initial gabapentin efficacy in the event of lower RDD values. Among various indicators, RDD serves as an independent factor that could indicate minimal VAS score improvement; namely, the treatment benefit was found to increase by 1.27 times per 1% decrement in the RDD value. RDD may therefore be of value in guiding selection of front-line analgesics in individual patients with DPN.

It should be noted that interpretation of this study may be limited by the relatively small sample size (n = 50–73 per arm), and our findings require confirmation in a larger cohort. RDD measurement was also limited to subjects in whom H-reflexes were demonstrable, so that a number of potential subjects were excluded due to the loss of H-reflexes arising from severe neuropathy. Further, our studies were restricted to the initial period of gabapentin-induced pain relief and cannot be extrapolated to longer periods of treatment until we have completed such long-term studies. It would also be of interest to expand out observations beyond VAS to include multiple methods of pain evaluation and to investigate various doses and durations of gabapentin and the relative efficacy of other front-line analgesics such as duloxetine.

X.Z. and Y.Z. contributed equally to this work.

Clinical trial reg no. ChiCTR2000038848, www.chictr.org.cn

Funding. This work was supported by the Shanghai Municipal Health Commission (grant no. ZHYY-ZXJHZX-1-201701).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. X.Z. contributed to conceptualization, methodology, and resources, and to writing the original draft. Y.Z. contributed to data curation, formal analysis, resources, and to writing the original draft. Z.W. contributed to conceptualization, data curation, and methodology. Z.L. contributed to data curation, methodology, and formal analysis. D.Z. contributed to formal analysis and reviewed and edited the manuscript. C.X. contributed to resources and data curation, and reviewed and edited the manuscript. N.A.C. contributed to conceptualization, reviewed and edited the manuscript, and supervised the study. Y.G. supervised the study and was project administrator. Y.G. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

1.
Li
S
,
Chen
X
,
Li
Q
, et al
.
Effects of acetyl-L-carnitine and methylcobalamin for diabetic peripheral neuropathy: a multicenter, randomized, double-blind, controlled trial
.
J Diabetes Investig
2016
;
7
:
777
785
2.
Pirart
J
.
Diabetes mellitus and its degenerative complications: a prospective study of 4,400 patients observed between 1947 and 1973
.
Diabetes Care
1978
;
1
:
252
263
3.
Abbott
CA
,
Malik
RA
,
van Ross
ERE
,
Kulkarni
J
,
Boulton
AJM
.
Prevalence and characteristics of painful diabetic neuropathy in a large community-based diabetic population in the U.K
.
Diabetes Care
2011
;
34
:
2220
2224
4.
Attal
N
,
Cruccu
G
,
Baron
R
, et al
.
EFNS guidelines on the pharmacological treatment of neuropathic pain: 2010 revision
.
Eur J Neurol
2010
;
17
:
1113
1123.e67-e88
5.
Dworkin
RH
,
O’Connor
AB
,
Kent
J
, et al
.
Interventional management of neuropathic pain: NeuPSIG recommendations
.
Pain
2013
;
154
:
2249
2261
6.
National Institute for Health and Care Excellence
.
Neuropathic pain in adults: pharmacological management in non-specialist settings
.
Clinical guideline [CG173]. Published: 20 November 2013. Last updated: 22 September 2020. Accessed 14 April 2021. Available from https://www.nice.org.uk/guidance/CG173
7.
Goodman
CW
,
Brett
AS
.
A clinical overview of off-label use of gabapentinoid drugs
.
JAMA Intern Med
2019
;
179
:
695
701
8.
Wiffen
PJ
,
Derry
S
,
Moore
RA
, et al
.
Antiepileptic drugs for neuropathic pain and fibromyalgia— an overview of Cochrane reviews
.
Cochrane Database Syst Rev
2013
:
CD010567
9.
Wiffen
PJ
,
Derry
S
,
Bell
RF
, et al
.
Gabapentin for chronic neuropathic pain in adults
.
Cochrane Database Syst Rev
2017
;
6
:
CD007938
10.
Mathieson
S
,
Lin
CC
,
Underwood
M
,
Eldabe
S
.
Pregabalin and gabapentin for pain
.
BMJ
2020
;
369
:
m1315
11.
Moore
RA
,
Wiffen
PJ
,
Derry
S
,
Toelle
T
,
Rice
ASC
.
Gabapentin for chronic neuropathic pain and fibromyalgia in adults
.
Cochrane Database Syst Rev
2014
:
CD007938
12.
Finnerup
NB
,
Attal
N
,
Haroutounian
S
, et al
.
Pharmacotherapy for neuropathic pain in adults: a systematic review and meta-analysis
.
Lancet Neurol
2015
;
14
:
162
173
13.
Calcutt
NA
.
Diabetic neuropathy and neuropathic pain: a (con)fusion of pathogenic mechanisms?
Pain
2020
;
161
(
Suppl. 1
):
S65
S86
14.
Ishikawa
K
,
Ott
K
,
Porter
RW
,
Stuart
D
.
Low frequency depression of the H wave in normal and spinal man
.
Exp Neurol
1966
;
15
:
140
156
15.
Nielsen
J
,
Petersen
N
,
Ballegaard
M
,
Biering-Sørensen
F
,
Kiehn
O
.
H-reflexes are less depressed following muscle stretch in spastic spinal cord injured patients than in healthy subjects
.
Exp Brain Res
1993
;
97
:
173
176
16.
Aymard
C
,
Katz
R
,
Lafitte
C
, et al
.
Presynaptic inhibition and homosynaptic depression: a comparison between lower and upper limbs in normal human subjects and patients with hemiplegia
.
Brain
2000
;
123
:
1688
1702
17.
Sabatier
MJ
,
Wedewer
W
,
Barton
B
,
Henderson
E
,
Murphy
JT
,
Ou
K
.
Slope walking causes short-term changes in soleus H-reflex excitability
.
Physiol Rep
2015
;
3
:
e12308
18.
Lamy
J-C
,
Wargon
I
,
Mazevet
D
,
Ghanim
Z
,
Pradat-Diehl
P
,
Katz
R
.
Impaired efficacy of spinal presynaptic mechanisms in spastic stroke patients
.
Brain
2009
;
132
:
734
748
19.
Marshall
AG
,
Lee-Kubli
C
,
Azmi
S
, et al
.
Spinal disinhibition in experimental and clinical painful diabetic neuropathy
.
Diabetes
2017
;
66
:
1380
1390
20.
Lee-Kubli
C
,
Marshall
AG
,
Malik
RA
,
Calcutt
NA
.
The H-reflex as a biomarker for spinal disinhibition in painful diabetic neuropathy
.
Curr Diab Rep
2018
;
18
:
1
21.
American Diabetes Association
.
2. Classification and diagnosis of diabetes: Standards of Medical Care in Diabetes—2020
.
Diabetes Care
2020
;
43
(
Suppl. 1
):
S14
S31
22.
Pop-Busui
R
,
Boulton
AJM
,
Feldman
EL
, et al
.
Diabetic neuropathy: a position statement by the American Diabetes Association
.
Diabetes Care
2017
;
40
:
136
154
23.
Lauria
G
,
Hsieh
ST
,
Johansson
O
, et al.;
European Federation of Neurological Societies; Peripheral Nerve Society
.
European Federation of Neurological Societies/Peripheral Nerve Society Guideline on the use of skin biopsy in the diagnosis of small fiber neuropathy. Report of a joint task force of the European Federation of Neurological Societies and the Peripheral Nerve Society
.
Eur J Neurol
2010
;
17
:
903
912.e44-49
24.
Patel
R
,
Kucharczyk
M
,
Montagut-Bordas
C
,
Lockwood
S
,
Dickenson
AH
.
Neuropathy following spinal nerve injury shares features with the irritable nociceptor phenotype: a back-translational study of oxcarbazepine
.
Eur J Pain
2019
;
23
:
183
197
25.
Demant
DT
,
Lund
K
,
Vollert
J
, et al
.
The effect of oxcarbazepine in peripheral neuropathic pain depends on pain phenotype: a randomised, double-blind, placebo-controlled phenotype-stratified study
.
Pain
2014
;
155
:
2263
2273
26.
Serra
J
,
Duan
WR
,
Locke
C
,
Solà
R
,
Liu
W
,
Nothaft
W
.
Effects of a T-type calcium channel blocker, ABT-639, on spontaneous activity in C-nociceptors in patients with painful diabetic neuropathy: a randomized controlled trial
.
Pain
2015
;
156
:
2175
2183
27.
Toda
T
,
Ishida
K
,
Kiyama
H
,
Yamashita
T
,
Lee
S
.
Down-regulation of KCC2 expression and phosphorylation in motoneurons, and increases the number of in primary afferent projections to motoneurons in mice with post-stroke spasticity
.
PLoS One
2014
;
9
:
e114328
28.
Lee-Kubli
CA
,
Zhou
X
,
Jolivalt
CG
,
Calcutt
NA
.
Pharmacological modulation of rate-dependent depression of the spinal H-reflex predicts therapeutic efficacy against painful diabetic neuropathy
.
Diagnostics (Basel)
2021
;
11
:
283
29.
Worthington
A
,
Kalteniece
A
,
Ferdousi
M
, et al
.
Optimal utility of H-reflex RDD as a biomarker of spinal disinhibition in painful and painless diabetic neuropathy
.
Diagnostics (Basel)
2021
;
11
:
1247
30.
Worthington
A
,
Kalteniece
A
,
Ferdousi
M
, et al
.
Spinal inhibitory dysfunction in patients with painful or painless diabetic neuropathy
.
Diabetes Care
2021
;
44
:
1835
1841
31.
Bockbrader
HN
,
Wesche
D
,
Miller
R
,
Chapel
S
,
Janiczek
N
,
Burger
P
.
A comparison of the pharmacokinetics and pharmacodynamics of pregabalin and gabapentin
.
Clin Pharmacokinet
2010
;
49
:
661
669
32.
McLean
MJ
.
Clinical pharmacokinetics of gabapentin
.
Neurology
1994
;
44
(
Suppl. 5
):
S17
S22
;
discussion S31–S32
33.
Beydoun
A
,
Uthman
BM
,
Sackellares
JC
.
Gabapentin: pharmacokinetics, efficacy, and safety
.
Clin Neuropharmacol
1995
;
18
:
469
481
34.
Gholyaf
M
,
Sheikh
V
,
Yasrebifar
F
,
Mohammadi
Y
,
Mirjalili
M
,
Mehrpooya
M
.
Effect of mirtazapine on pruritus in patients on hemodialysis: a cross-over pilot study
.
Int Urol Nephrol
2020
;
52
:
1155
1165
35.
Pirbudak
L
,
Ciçek
H
,
Işik
M
,
Zer
Y
.
The effect of tramadol and tramadol + gabapentin combination in patients with lumbar disc herniation after epidural steroid injection
.
Turk J Med Sci
2015
;
45
:
1214
1219
36.
Davies
JA
.
Mechanisms of action of antiepileptic drugs
.
Seizure
1995
;
4
:
267
271
37.
Hoppa
MB
,
Lana
B
,
Margas
W
,
Dolphin
AC
,
Ryan
TA
.
α2δ expression sets presynaptic calcium channel abundance and release probability
.
Nature
2012
;
486
:
122
125
38.
Kukkar
A
,
Bali
A
,
Singh
N
,
Jaggi
AS
.
Implications and mechanism of action of gabapentin in neuropathic pain
.
Arch Pharm Res
2013
;
36
:
237
251
39.
Yu
J
,
Wang
D-S
,
Bonin
RP
, et al
.
Gabapentin increases expression of δ subunit-containing GABAA receptors
.
EBioMedicine
2019
;
42
:
203
213
40.
Chu
LC
,
Tsaur
ML
,
Lin
CS
, et al
.
Chronic intrathecal infusion of gabapentin prevents nerve ligation-induced pain in rats
.
Br J Anaesth
2011
;
106
:
699
705
41.
Jiang
Y
,
Li
J
,
Lin
H
, et al
.
The efficacy of gabapentin in reducing pain intensity and morphine consumption after breast cancer surgery: a meta-analysis
.
Medicine (Baltimore)
2018
;
97
:
e11581
42.
Costigan
M
,
Scholz
J
,
Woolf
CJ
.
Neuropathic pain: a maladaptive response of the nervous system to damage
.
Annu Rev Neurosci
2009
;
32
:
1
32
43.
Baliki
MN
,
Chialvo
DR
,
Geha
PY
, et al
.
Chronic pain and the emotional brain: specific brain activity associated with spontaneous fluctuations of intensity of chronic back pain
.
J Neurosci
2006
;
26
:
12165
12173
44.
Ossipov
MH
,
Morimura
K
,
Porreca
F
.
Descending pain modulation and chronification of pain
.
Curr Opin Support Palliat Care
2014
;
8
:
143
151
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.