Electrophysiologic Characteristics of Nitrous-Oxide-Associated Peripheral Neuropathy: A Retrospective Study of 76 Patients

Abstract

Background and Purpose

The electrophysiologic characteristics of peripheral neuropathy secondary to nitrous oxide (N₂O) abuse remain unclear. The paper therefore aimed to summarize the electrophysiologic characteristics of N₂O-associated peripheral neuropathy and identify the risk factors of severe nerve injury.

Methods

The electrophysiologic results and clinical data of patients with peripheral neuropathy secondary to N₂O abuse at our hospital between 2018 and 2020 were analyzed retrospectively, and their electrophysiologic changes were summarized.

Results

Most patients exhibited decreased sensory and motor nerve conduction velocities (75% and 76%), decreased sensory nerve and compound motor action potentials (57% and 59%), and prolonged distal motor latency (59%), while a response was absent in 36%. These findings indicate that N₂O abuse can result in generalized injury to sensory and motor nerves. Electrophysiologic results indicated axonal neuropathy in 37 cases (49%), demyelinating peripheral neuropathy in 4 (5%), and mixed neuropathy in 12 (16%). Peripheral nerve injury was more common in the lower limbs (72%) than in the upper limbs (42%, p<0.0001). The upper and lower limbs were primarily affected by sensory nerve demyelination (35%) and motor axonal injury (67%), respectively. Subgroup analysis indicated that longer N₂O exposure and longer disease course were associated with more-severe motor axonal injury in the lower limbs.

Conclusions

N₂O-associated peripheral neuropathy can lead to sensory and motor nerve injury, with axonal injury being the most common. Injuries were more severe in the lower limbs. Prolonged N₂O exposure and disease course increased the severity of motor axonal injury in the lower limbs.

INTRODUCTION

Nitrous oxide (N₂O) is an oxidizing agent that is primarily used for anesthesia and in food manufacturing, such as in cream whipping.¹,² N₂O is also used as a recreational drug to produce stimulant and euphoric effects after inhalation. It acts rapidly, reaching its peak effect in approximately 1 minute, with its effects waning after about 2 minutes. N₂O trafficking and consumption have recently become increasingly common and occur primarily in bars, night clubs, and other entertainment venues; people can also find it in private homes and house parties, and most consumers are teenagers.³,⁴ Vitamin B12 metabolism is relevant to the development of N₂O-induced neuropathy.⁵,⁶,⁷ N₂O users have been reported to exhibit symptoms of peripheral neuropathy (most commonly), myelopathy, and encephalopathy.⁵,⁸,⁹ Layzer et al.¹⁰ were the first to report cases (n=3) of peripheral neuropathy caused by N₂O. They used nerve conduction testing to suggested that this type of peripheral neuropathy most commonly involved axonal injury. Hsu et al.¹¹ reported a patient with multiple peripheral demyelination-predominant polyneuropathy, and Morris et al.¹² reported a patient with severe axonal degeneration. These findings indicate that N₂O-associated peripheral neuropathy often manifests as acute sensorimotor polyneuropathy. However, only a few studies have summarized the electrophysiologic characteristics of N₂O-associated peripheral neuropathy.¹³,¹⁴ The present study therefore aimed to summarize the electrophysiologic characteristics of N₂O-associated peripheral neuropathy and determine the factors that affect severe nerve injury.

METHODS

Subjects

This study analyzed patients diagnosed with N₂O-associated peripheral neuropathy at Shengjing Hospital of China Medical University between February 2018 and August 2020 who had clinical symptoms of numbness and/or weakness in the distal limbs or electrophysiologic results that indicated peripheral neuropathy. The patients had various N₂O inhalation histories, and none had any personal or family history of neurologic or psychiatric diseases. Patients were excluded if their peripheral neuropathy was possibly caused by diabetes, alcohol poisoning, digestive system disease, long-term vegetarianism, infection and immunity, toxins, chemical substances, or genetic factors. All patients underwent electromyography (EMG). The study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Shengjing Hospital of China Medical University (IRB No. 2020PS047K).

Methods

Clinical data

Clinical data of the patients were collected, including clinical symptoms and signs, laboratory test results (hemoglobin, vitamin B12, folate, and homocysteine), and electrophysiologic and imaging results. Vitamin B12 deficiency was defined as a vitamin B12 level lower than the lower normal limit (180 pg/mL), and hyperhomocysteinemia was defined as a homocysteine level higher than the upper normal limit (15 µmol/L).

Electrophysiologic study

The tested sensory nerves were the median, ulnar, superficial peroneal, and sural nerves. The electrophysiologic parameters included the sensory nerve conduction velocity (SNCV) and sensory nerve action potential (SNAP). The tested motor nerves were the median, ulnar, common peroneal, and tibial nerves. The electrophysiologic parameters included distal motor latency, motor nerve conduction velocity (MNCV), and compound motor action potential (CMAP).

Demyelinating neuropathy was considered to have occurred when two or more nerves met the following criteria: SNCV or MNCV <80% of the lower normal limit, or distal motor latency ≥120% of the upper normal limit. Axonal neuropathy was considered to have occurred when two or more nerves met the following criteria: SNAP or CMAP <80% of the lower normal limit, or no response. Mixed neuropathy was considered to have occurred if the criteria for both demyelinating and axonal neuropathy were met.¹⁴,¹⁵,¹⁶

Statistics

All statistical analyses were performed using SPSS (version 26.0, IBM Corp., Armonk, NY, USA). Continuous variables are expressed as mean±SD or median (interquartile range) values. All categorical variables are expressed as frequency (percentage) values. Measurement data were compared and analyzed using the t-test for independent samples or analysis of variance. Count data were analyzed using the chi-square test. Differences for which p<0.05 were considered significant.

RESULTS

Clinical data, and laboratory and imaging examinations

This study enrolled 76 patients (34 males, 42 females) aged 21.2±4.1 (range 14–33) years. The clinical data and laboratory test and imaging results of the patients are listed in Table 1. The most common clinical symptoms were limb numbness and/or weakness (74 cases, 98%), which were present in all limbs (53 cases, 70%) or only the lower limbs (21 cases, 28%). The most common positive neurologic sign was decreased limb muscle strength (60 cases, 79%). Of the 51 patients who had their hemoglobin tested, 86% (44/51 cases) exhibited anemia, 34% (15/44 cases) of which presented with megaloblastic anemia, with a corpuscular volume of 105.3±4.4 fL (normal value=83–101 fL). Vitamin B12 and folate tests were performed together in 44 patients, and 28 (64%) had vitamin B12 deficiency. The folic acid levels of those 44 patients were within the normal range (3–17 ng/mL). Homocysteine testing was performed on 35 patients, and 25 (71%) hyperhomocysteinemia cases were found. Spinal cord magnetic resonance imaging (MRI) was performed on 36 patients, and 20 cases (56%) were found to have high T2-weighted signals. Head MRI was performed on 13 patients; 6 patients (46%) of abnormal brain were found on MRI, comprising 3 cases each of brain atrophy and demyelination.

Table 1. Patient demographics and laboratory results.

			Value	Range
Sex
	Males		34 (45)
	Females		42 (55)
Age, yr			21.2±4.1	14–33
	Males		22.7±3.7	15–33
	Females		19.9±4.0	14–32
N2O exposure time, days			365 [90–365]	15–1460
	Males		240 [90–730]	30–1095
	Females		365 [90–365]	15–1460
Disease course, days			30 [8–30]	3–1095
	Males		18 [7–30]	7–1095
	Females		30 [7–45]	3–365
Clinical symptoms
	Neurologic symptoms
		Numbness and/or weakness in all limbs	53 (70)
		Numbness and/or weakness in lower limbs	21 (28)
		Walking instability	32 (42)
		Headache	5 (7)
		Urinary retention or dysphoria	10 (13)
		Constipation	3 (4)
		Hypomnesia	7 (9)
	Psychiatric symptom
		Illusion	4 (5)
	Other symptoms
		Chest distress	3 (4)
		Skin pigmentation	7 (9)
		Hearing loss	1 (1)
		Reduced or increased appetite	12 (16)
Positive neurologic signs
	Decreased limb muscle strength		60 (79)
	Hypesthesia of pain and temperature		46 (61)
	Reduced or absent tendon reflex		45 (59)
	Disturbance of deep sensations		44 (58)
	Ataxia		22 (29)
	Romberg sign		15 (20)
	Signs of upper motor neuron injury		10 (13)
Laboratory examinations
	Hemoglobin, normal range 110–150 g/dL		131.2±20.9	70–170
		Normal	7/51 (14)
		Anemia	44/51 (86)
	Vitamin B12, normal range 180–910 pg/mL		189.9±124.4	50–828
		Normal	16/44 (36)
		Decreased	28/44 (64)
	Homocysteine, normal range 0–15 μmol/L		43.0±39.4	3–138
		Normal	10/35 (29)
		Increased	25/35 (71)
Spine MRI
	Normal		16/36 (44)
	Abnormal		20/36 (56)
Brain MRI
	Normal		7/13 (54)
	Abnormal		6/13 (46)

Data are mean±SD, n (%), median [interquartile range] values, or n/N values. The n in n/N is the number of patients who underwent the test or examination; Disease course: duration between initial symptom onset and admission.

N₂O, nitrous oxide.

Electrophysiologic results

Most patients exhibited decreased SNCV and MNCV (75% and 76%, respectively; p=0.850) and decreased SNAP and CMAP (57% and 59%, respectively; p=0.743). The prolonged distal motor latency and absent response rates were 59% and 36%, respectively. Among the 76 patients, 284 sensory and 276 motor nerves were tested. The EMG results for these nerves are listed in Table 2. The motor nerve potentials of the lower limbs were significantly decreased, by 76% and 63% in the common peroneal and tibial nerves, respectively. Based on the electrophysiologic diagnostic criteria, the most common neuropathy type was axonal (37 cases, 49%), followed by mixed demyelinating and axonal (12 cases, 16%) and demyelinating (4 cases, 5%). Mild electrophysiologic abnormalities were observed in 21 patients (28%) who did not meet the diagnostic criteria for axonal or demyelinating neuropathy. Two patients (3%) had normal electrophysiologic findings.

Table 2. Electromyography results for various nerves.

Sensory nerves		Median	Ulnar	SPN	Sural	Total
Total		76	68	75	65	284
SNCV<NV		18 (24)	33 (49)	38 (51)	23 (35)	112 (39)
	m/s	47.13 [6]	46.38 [7]	36.29 [9]	36.36 [9]
SNAP<NV		20 (26)	16 (24)	23 (31)	13 (20)	72 (25)
	μV/mV	11.09 [45]	8.86 [48]	4.17 [31]	3.81 [37]
No response		0	0	13 (17)	9 (14)	22 (8)
Motor nerves		Median	Ulnar	CPN	Tibial	Total
Total		76	57	75	68	276
MNCV<NV		18 (24)	18 (32)	51 (68)	46 (68)	137 (50)
	m/s	46.60 [7]	46.67 [7]	34.72 [21]	34.99 [17]
CMAP<NV		4 (5)	6 (11)	32 (43)	34 (50)	76 (28)
	μV/mV	2.450 [39]	4.017 [33]	0.730 [76]	1.494 [63]
DML>NV		6 (8)	9 (16)	25 (33)	30 (44)	70 (25)
	m/s	4.620 [5]	3.523 [7]	6.216 [24]	5.829 [21]
No response		0	0	17 (23)	13 (19)	30 (11)

Data are n, n (%), or mean [percentage change] values.

CMAP, compound muscle action potential; CPN, common peroneal nerve; DML, distal motor latency; MNCV, motor nerve conduction velocity; NV, normal value; SNAP, sensory nerve action potential; SNCV, sensory nerve conduction velocity; SPN, superficial peroneal nerve.

Electrophysiologic characteristics

The types and rates of sensory and motor nerve injury were further analyzed based on the distributions in the upper and lower limbs (Table 3). The injury rate was significantly higher in the lower limbs (72%) than in the upper limbs (42%) (χ²=51.345, p<0.0001). The primary type of injury in the upper limbs was demyelination (35%), and that in the lower limbs was axonal injury (54%) (χ²=20.654, p<0.0001). This indicates that both the nerve injury rate and injury type exhibited different distributions between the limbs. In the lower limbs, the injury rate was higher in the motor nerves (76%) than in the sensory nerves (68%) (χ²=1.700, p=0.192). In contrast, that in the upper limbs was higher in the sensory nerves (48%) than in the motor nerves (35%) (χ²=4.494, p=0.034). This indicates that the nerve injury rate in the same limb was associated with the nerve type. We also found that the nerve injury type in the lower limbs was associated with the nerve type, unlike in the upper limbs. The primary injury type in the motor nerves of the lower limbs was axonal injury (67%), and that in the sensory nerves was demyelination (44%) (χ²=7.503, p=0.023).

Table 3. Distribution differences of the rates and types of neuroelectrophysiologic abnormalities.

	Demyelination	Axonal injury	Mixed*	Total	χ2, p
Sensory nerve (upper limb)	51 (35)	36 (25)	18 (13)	69 (48)	12.44, 0.0004
Sensory nerve (lower limb)	61 (44)	58 (41)	23 (16)	96 (69)
χ2, p		1.134, 0.567
Motor nerve (upper limb)	45 (34)	10 (8)	8 (6)	47 (35)	45.199, <0.0001
Motor nerve (lower limb)	77 (54)	96 (67)	65 (45)	108 (76)
χ2, p		31.617, <0.0001
Sensory nerve (upper limb)	51 (35)	36 (25)	18 (13)	69 (48)	4.494, 0.034
Motor nerve (upper limb)	45 (34)	10 (8)	8 (6)	47 (35)
χ2, p		8.978, 0.011
Sensory nerve (lower limb)	61 (44)	58 (41)	23 (16)	96 (69)	1.700, 0.192
Motor nerve (lower limb)	77 (54)	96 (67)	65 (45)	108 (76)
χ2, p		7.503, 0.023
Sensory and motor nerves (upper limb)	96 (35)	46 (17)	26 (9)	116 (42)	51.345, <0.0001
Sensory and motor nerves (lower limb)	138 (49)	154 (54)	88 (31)	204 (72)
χ2, p		20.654, <0.0001

Data are n (%) values.

^*Both demyelination and axonal neuropathy.

Analysis of factors influencing nerve injury severity

Since the results described above indicated that motor nerve axonal injury in the lower limbs was the most-prominent injury, the patients were divided into two groups based on the degree of motor nerve axonal injury in this region. The inclusion criterion for the severe group (n=24) was CMAP ≤1 mV or an absent response; the remaining patients constituted the mild group (n=52). Factors that influenced the degree of motor nerve axonal injury in the lower limbs were further analyzed (Table 4). There were significant differences in disease course and exposure time between the two patient groups (p<0.001), suggesting that these factors influence the degree of motor nerve axonal injury in the lower limbs. We therefore further applied logistic regression analysis to the two independent variables of disease course and exposure time with the outcome set as severe motor nerve axonal injury in the lower limbs (no=0, yes=1). This analysis indicated that a longer exposure time and disease course were associated with more-severe motor axonal injury in the lower limbs (exposure time: odds ratio [OR]=1.003, 95% confidence interval [CI]=1.000–1.005, p=0.02; disease course: OR=1.042, 95% CI=1.012–1.072, p=0.005) (Table 5).

Table 4. Analysis of factors affecting the severity of nerve injury according to electrophysiologic results.

	Severe group			Mild group
	Total	Females	Males	Total	Females	Males	p
Number	24	13 (54)	11 (46)	52	29 (56)	23 (44)	0.896
Age, years	21.5±4.5	19.2±3.1	24.2±4.5	21.0±4.1	20.1±4.4	22.2±3.3	0.687
p	0.004			0.062
Exposure time, days	365 [180–821]	365 [180–730]	365 [180–1095]	365 [90–365]	365 [90–365]	365 [180–730]	<0.0001
p	0.580			0.432
Disease course, days	30 [26.3–97.5]	30 [30–90]	30 [7–120]	15 [7–30]	15 [7–30]	15 [7–30]	<0.0001
p	0.541			0.740
Hemoglobin, g/L	128.9±18.6	122.1±17.0	137.6±17.9	132.3±22.1	123.1±19.7	146.1±18.5	0.429
p	0.077			0.002
Vitamin B12, pg/mL	188.0±145.2	190.6±186.1	185±111.8	190.7±112.9	185.9±122.2	195.7±106.5	0.628
p	0.470			0.629
HCY, μmol/L	35.8±22.9	39.9±23.0	24.9±23.1	50.8±45.2	54.1±48.3	48.1±44.4	0.721
p	0.497			0.999
Spinal MRI	8/13	4	4	12/23	7	5	0.587
Brain MRI	3/6	1	2	3/7	1	2	0.797

Data are mean±SD, median [interquartile range], n (%), or n/N values. The n in n/N is the number of patients with abnormal MRI results; Disease course: duration between initial symptom onset and admission.

HCY, homocysteine.

Table 5. Logistic regression analysis of factors influencing severity of motor nerve axonal injury in the lower limbs.

Variable	B	SEM	Wald	p	Odds ratio	95% CI
Exposure time	0.003	0.001	5.372	0.020	1.003	1.000–1.005
Disease course	0.041	0.015	7.887	0.005	1.042	1.012–1.072

CI, confidence interval; SEM, standard error of the mean.

DISCUSSION

Recreational N₂O use has recently become increasingly popular in both Western and Asian countries, especially among teenagers and young adults.¹⁷,¹⁸,¹⁹,²⁰,²¹ The results of the present study reaffirm that N₂O users tend to be young. Myelopathy and (particularly) peripheral neuropathy are the most common N₂O-associated neuropathies. All of the patients in the present study were diagnosed with peripheral neuropathy, of which 20 exhibited abnormalities on spinal cord MRI, suggesting concomitant myelopathy. Consistent with previous studies, the common clinical symptoms of N₂O abuse are numbness and/or limb weakness. Neurologic signs often include weak or lost tendon reflexes, decreased sensitivity to pain and temperature, and decreased muscle strength in the limbs, which are consistent with peripheral neuropathy.²²,²³,²⁴,²⁵,²⁶,²⁷,²⁸,²⁹

The exact mechanism underlying N₂O-associated peripheral neuropathy has not been elucidated. Previous studies suggested that the pathologic effect of N₂O is related to vitamin B12, which plays a key role in maintaining myelin production and metabolism. N₂O irreversibly induces vitamin B12 oxidation, thereby inhibiting the methionine synthase activity and blocking methylation and DNA synthesis, which affects myelin methylation and leads to the demyelination of central and peripheral nerves.⁵,⁶,⁷ N₂O can also exert neurotoxic effects by antagonizing the accumulation of NMDA receptors and homocysteine. Injury due to oxidative stress and changes in cytokine and growth factor levels are also considered to be related to N₂O-associated neuropathy.¹⁷,³⁰,³¹,³²,³³

Homocysteine and vitamin B12 levels were therefore assessed in the present study. The results indicated that the hyperhomocysteinemia rate (71%) was higher than the vitamin B12 deficiency rate (64%). Consistent with previous studies, the present study also found that the low vitamin B12 deficiency rate may be due to many patients taking vitamin B12 supplements before visiting the clinic.³⁴ Even if serum vitamin B12 levels are normal, there may be vitamin B12 deficiency in the cells. Buizert et al.⁸ reported a case of subacute combined spinal degeneration, in which hematologic tests indicated normal serum vitamin B12 and folic acid levels but significantly elevated methylmalonic acid levels; elevated homocysteine or methylmalonic acid levels indicate a lack of functional vitamin B12 at the cellular level.⁸,³⁵ Homocysteine is therefore more suitable as a test indicator than vitamin B12 when examining patients suspected of N₂O abuse.

Some differences exist in the electrophysiologic characteristics of the different etiologies of peripheral neuropathy. Peripheral neuropathy due to chemotherapeutic agent use manifests primarily as a pathology in sensory nerve axons,³⁶,³⁷,³⁸,³⁹,⁴⁰,⁴¹,⁴² whereas peripheral neuropathy due to copper and folic acid deficiency manifests as sensory rather than motor peripheral neuropathy.⁴³,⁴⁴ These peripheral neuropathies primarily manifest as sensory neuropathy. However, the present study found that peripheral neuropathy due to N₂O generally affects sensory and motor nerves, and the rate and type of nerve injury are associated with limb distribution and nerve type. The nerve injury rate is higher in the lower limbs, and injury is more severe in the distal than in the proximal limbs. Consistent with the neuroelectrophysiologic findings, we also observed that weakness and numbness were more severe in the distal and lower limbs than in the proximal and upper limbs. This length-dependent characteristic is also common in peripheral neuropathy with other etiologies.¹⁴,³⁶,⁴⁵,⁴⁶ The motor nerve potentials of the lower limbs exhibited the most-significant changes, suggesting that motor nerve axonal injury is most severe in the lower limbs.

As mentioned above, peripheral neuropathy associated with N₂O exposure is often attributed to vitamin B12 deficiency. Tani et al.²⁸ compared the EMG presentations of peripheral neuropathy caused by N₂O exposure and isolated vitamin B12 deficiency, and found that the rate and degree of motor nerve axonal injury were more significant in the N₂O group. The prominent changes in motor axons of patients with N₂O exposure may indicate a distinct pathophysiologic mechanism. The nerve excitability test in patients with N₂O exposure performed by Tani et al.²⁸ indicated a decreased accommodation toward depolarizing current and increased superexcitability, which may be related to the loss of or conduction block of the largest nerve fibers caused by N₂O toxicity in the axonal paranodal region. The most common injury type in the motor nerves was axonal, and that in the sensory nerves was demyelination. Motor nerves contain more large nerve fibers than the sensory nerves, which may have a closer relationship with this difference. EMG and nerve biopsies may be more helpful to for explaining this problem from the perspective of pathophysiology, which requires further study.

The present study found an association between the degree of nerve injury and the duration of N₂O exposure and disease course. A longer exposure time and disease course were associated with more-significant nerve potential changes, suggesting that motor axonal injury was more severe in the lower limbs. Exposure duration plays an important role in predicting the degree of nerve injury. Previous studies have also found that long-term N₂O consumption is more likely to result in neurologic symptoms. The present study also found significant symptoms and rapid progression even in patients with short exposure durations. This may be related to consumption of high doses of N₂O over a short period, and may also differ among individuals based on their sensitivity to N₂O.¹⁷ The degree of nerve injury is also related to the amount of N₂O consumed. Alt et al.⁴⁷ suggested that an N₂O dose exceeding 80 g/day increases the risk of permanent nerve injury. The present study did not calculate the amount of N₂O consumed overall or during a single inhalation, and so could not analyze the relationship between N₂O consumption and the degree of nerve injury. However, the exposure duration somewhat reflects consumption. The present study did not find any association between the vitamin B12, homocysteine, or hemoglobin level and the degree of nerve injury. Li et al.¹⁴ compared the EMG findings of patients with normal and abnormal vitamin B12 and/or homocysteine levels, and found no association between nerve injury and vitamin B12 or homocysteine level. In the present study, the electrophysiologic results illustrated the characteristics and extent of injury in N₂O-associated peripheral neuropathy and provided evidence for a better understanding of the pathogenesis and pathophysiology of this neuropathy.

The present study had some limitations. First, it had a retrospective design and some patients lacked vitamin B12, homocysteine, and hemoglobin data. Second, some patients lacked a complete medical history regarding exposure time, quantity of N₂O consumed, and whether they inhaled large N₂O doses within a short period of time, all of which may have affected the analysis of factors influencing the degree of injury. A prospective experimental design with a larger sample is therefore necessary to clarify these issues.

In conclusion, in N₂O-associated peripheral neuropathy, SNCV, MNCV, SNAP, and MNAP were significantly decreased, and distal motor latency was significantly prolonged. Axonal neuropathy was the most common N₂O-associated peripheral neuropathy. Motor nerve axonal injury in the lower limbs was the most-characteristic injury. Motor axonal injury in the lower limbs was more severe in the presence of a longer N₂O exposure duration and disease course.

Availability of Data and Material

The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.