Article Text

Download PDFPDF

Nitrous oxide-induced subacute combined degeneration of the cord: diagnosis and treatment
  1. Alvar Paris1,2,
  2. Luke Lake2,
  3. Albert Joseph2,
  4. Anna Workman2,
  5. Joseph Walton2,
  6. Tom Hayton3,4,
  7. Nikos Evangelou5,6,
  8. James B Lilleker7,8,
  9. Ruth M Ayling2,
  10. David Nicholl9,
  11. Alastair J Noyce1,2
  1. 1Preventive Neurology Unit, Centre for Prevention, Diagnosis and Detection, Wolfson Institute of Population Health, Faculty of Medicine and Dentistry, Queen Mary University of London, London, UK
  2. 2Royal London Hospital, Barts Health NHS Trust, London, UK
  3. 3Neurology, University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK
  4. 4Surgical Reconstruction and Microbiology Research Centre, National Institute for Health Research, Birmingham, UK
  5. 5Academic Neurology, Nottingham University Hospitals NHS Trust, Nottingham, UK
  6. 6Mental Health and Clinical Neurosciences Academic Unit, School of Medicine, University of Nottingham, Nottingham, UK
  7. 7Muscle Diseases Unit, Manchester Centre for Clinical Neurosciences, Manchester Academic Health Science Centre, Northern Care Alliance NHS Foundation Trust, Salford, UK
  8. 8Centre for Musculoskeletal Research, Division of Musculoskeletal and Dermatological Sciences, The University of Manchester, Manchester, UK
  9. 9Sandwell and West Birmingham NHS Trust, Birmingham, UK
  1. Correspondence to Prof Alastair J Noyce, Preventive Neurology Unit, Centre for Prevention, Diagnosis and Detection, Wolfson Institute of Population Health, Faculty of Medicine and Dentistry, Queen Mary University of London, London, EC1M 6BQ, UK; a.noyce{at}


Recreational use of nitrous oxide (N2O) has increased rapidly in recent years and is now the second most commonly used recreational drug among young people in the UK. There has been a corresponding rise in cases of nitrous oxide-induced subacute combined degeneration of the cord (N2O-SACD), a pattern of myeloneuropathy usually associated with severe vitamin B12 deficiency. This can cause serious and permanent disability in young people but, if recognised early, may be effectively treated. All neurologists should be aware of N2O-SACD and its treatment; however, there are currently no agreed guidelines. Based on our experience in East London, an area of high N2O use, we provide practical advice on its recognition, investigation and treatment.

  • GAIT

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Nitrous oxide-induced subacute combined degeneration of the cord (N2O-SACD) was first described in 1978 in 14 dental practitioners abusing nitrous oxide (N2O).1 Vitamin B12 inactivation was soon established as the likely mechanism of toxicity, as N2O oxidises the cobalt atom integral to vitamin B12 function.2 3 N2O-SACD was subsequently reported as an infrequent complication of N2O anaesthesia in patients with pre-existing subclinical B12 deficiency.4–6 However, the exponential rise in recreational N2O use in the last decade has made it now the second most commonly used drug among 16–24 year olds in the UK.7 8 Around 3.4% of N2O users experience neurological symptoms consistent with subacute combined degeneration (SACD)9—but this may be an underestimate—and there has been a rapid rise in reports of N2O-SACD both in the UK and worldwide.10–14 Despite this, there is no consensus or guideline on how best to diagnose and treat N2O-SACD in practice.

At the Royal London Hospital, we diagnose and treat one case of N2O-SACD on average every 9 days. Since 2021, we have run a quality improvement project to improve its diagnosis and treatment and now share practical advice on its recognition, investigation and treatment to help other clinicians seeing similar rises in cases. The flowchart in figure 1 summarises our pathway.

Figure 1

Pathway for patients with suspected N2O-SACD. In the Emergency Department, the focus is on immediate treatment with I.M. hydroxocobalamin, time-critical investigations, and patient education. Due to subsequent non-attendance, the Emergency Department is occasionally the only point of patient contact. In our practice, methylmalonic acid (MMA) is preferred compared to homocysteine, as it is more specific and does not need to be transported on ice. However, the choice of test will be determined by local preference, availability, and turnaround time. In ambulatory care, the focus is on continuing treatment and repeated assessment of progress, and guided by the N2O-SACD multi-disciplinary team (MDT) meeting or specialist input. *In our practice, we arrange neurology follow-up if the final diagnosis remains uncertain, there are outstanding investigations to interpret or if there is significant disability.

Challenges in N2O-SACD

  • Recognising the disease.

  • Choosing and interpreting investigations.

  • Starting intramuscular B12 injections promptly and deciding on the duration of treatment.

  • Maximising treatment adherence and follow-up attendance.

  • Coordinating care between the emergency department, ambulatory medical care and neurology services.

Recognition and diagnosis

N2O-SACD is commonly misdiagnosed and so inappropriately treated.15 Awareness and recognition of the disease are crucial to preventing long-term harm. This can be helped by considering the local availability of N2O, the clinical features of the case, enquiring carefully about N2O use and maintaining a high index of suspicion (as patients may not always disclose N2O use).


Most patients developing N2O-SACD are young (16–30 years). Around 75% are male,9 13 16 but women may be more susceptible to N2O-SACD at higher doses of N2O (>20 canisters per session).9 In different geographical regions, members of different communities are more likely to use N2O and hence to develop N2O-SACD. In East London, most patients presenting to the hospital with N2O-SACD are of Bangladeshi descent13; in the Netherlands, Moroccan–Dutch men appear to be more affected17; and in Australia, harmful use is highest among university students.18

The risk of developing SACD symptoms increases with greater N2O use: those using one canister per session have around a 1.9% risk of experiencing distal paraesthesia compared with 8.5% in people who use over 100 canisters per session.9 In general, people presenting to the hospital with N2O-SACD report heavy recreational use over months to years.16 19–21 In our clinical practice over the past 5 years, patients have reported using an average of 580 canisters per week (IQR 147.5–1012.5), which is in keeping with other reports.20 However, those with pre-existing subclinical B12 deficiency are much more susceptible at lower doses of N2O.6 22–24 In this scenario, one-off use in anaesthesia can cause N2O-SACD,6 and anecdotally some of our patients who have come to harm through recreational use reported using only 5–10 canisters on one occasion.

Clinical features

The clinical features of N2O-SACD closely resemble those of classical SACD. Symptoms can present over a varying time course, from several days to 6 months.16 25 Where there has been only a single exposure to N2O, the time delay to symptoms is 2–6 weeks.5 The initial presenting complaint is often distal paraesthesia involving the lower and/or upper limbs. If the patient ignores this or if N2O-SACD is not diagnosed at presentation, patients may then develop gait ataxia (worse in low lighting), falls or inability to walk independently.1 13 Figure 1 shows features to prompt consideration of an alternative diagnosis.

On direct questioning, a large majority of patients have distal paraesthesia, mainly in the feet but often also the hands.13 16 Many report unsteadiness on walking.24 Less common symptoms are weakness,16 19 Lhermitte’s phenomenon,1 bladder or bowel urgency or incontinence,1 19 20 impotence or other sexual dysfunction,1 segmental myoclonus1 13 or psychiatric complaints.16 Some patients have urinary retention on presentation, and clinicians should consider arranging a bladder scan.20 Note there may also be symptoms suggesting pulmonary embolism or deep vein thrombosis, as N2O-induced hyperhomocysteinaemia can cause a hypercoagulable state.26 Additionally, clinicians should ask patients about risk factors for low serum B12 levels, which would increase their vulnerability to N2O-SACD. These include dietary preferences, previous gastric or small bowel resections, or chronic intestinal conditions such as Crohn’s disease or coeliac disease. Several common drugs can also decrease serum B12 concentrations, including gastric acid suppressants, metformin, potassium supplements and the contraceptive pill.

On examination, the physical findings are generally symmetrical. Common signs are distal loss of vibration and joint-position sense,1 16 27 and a positive Romberg’s sign or ataxic gait.1 13 16 21 Pseudoathetosis may occur through loss of joint-position sense.11 There may be hyporeflexia or (less commonly) hyper-reflexia at the knee or ankle, and this pattern may reflect the relative contributions of myelopathy and neuropathy to the clinical picture.1 16 19 In more severe cases, the plantar responses may be extensor.19 21 25 Muscle weakness is usually relatively mild but can become severe in advanced cases.1 21

Quantification of N2O use

Patients often do not spontaneously report N2O use. They may also not connect N2O use with their neurological symptoms, although those living in areas where use is common do increasingly recognise its possibility of harm.28 It is important to enquire specifically and to record the quantity, duration, pattern and type of N2O container used. Most patients in our experience purchase N2O online, where it is readily available to the general public and is marketed for use in whipping cream. Commonly, patients use 8 g canisters (known as ‘whippits’ or ‘chargers’ in the UK and ‘nangs’ in Australia), but cylinder use—which make it easier to consume large amounts of N2O—is becoming more common. Occasionally, patients report using medical-grade N2O, which is not available for sale to the general public but may have been stolen from healthcare facilities and sold.29 We recommend quantifying N2O use in ‘canister equivalents’, with one cylinder approximately equal to 75 canisters (table 1). In general, N2O abusers use containers of the gas to inflate balloons, which are then inhaled. Using a mask or releasing N2O into enclosed spaces is less common and particularly dangerous, as there is a risk of death by asphyxiation without preceding dyspnoea.30

Table 1

Forms of N2O used for recreational use and their current online cost

Choosing and interpreting investigations

Investigations in suspected N2O-SACD may support the clinical diagnosis, but there should be no delay in starting treatment promptly (with intramuscular B12 injections) while undertaking results. We take blood on all patients and arrange MRI of the cervical and thoracic cord on most patients, except for those with minor symptoms and a clear diagnosis. Nerve conduction studies and electromyography are optional tests to undertake if the history or examination are inconsistent, or if improvement is suboptimal with treatment.

Blood testing

Active vitamin B12 is essential for the enzymatic conversion of homocysteine to methionine and of methylmalonic co-enzyme A to succinyl co-enzyme A. Thus, vitamin B12 deficiency or B12 inactivation tends to increase plasma homocysteine and methylmalonic acid (MMA). Tests for assessing B12 status include serum B12, homocysteine, MMA and holotranscobalamin (table 2).

Table 2

Biochemical tests for N2O-SACD

On first suspecting N2O-SACD, we take blood for serum B12 and MMA. The laboratory processes the serum B12 concentration routinely; the MMA is processed in all patients except those in whom the B12 concentration is below the reference range, where it would not change the diagnosis or management. Homocysteine is an alternative to MMA, but it must be transported to the laboratory on ice, which in practice places a barrier to its use. Serum B12 has relatively low sensitivity (~20% to 50%) for SACD, but MMA and homocysteine are both significantly raised in over 80% of reported cases.13 19 21 31 However, both homocysteine and MMA normalise with B12 treatment and with abstinence from N2O, and it is likely that MMA and homocysteine testing may be normal if delayed, although the exact time frame is unclear.32 Ideally therefore, all blood samples for biochemical tests should be drawn before starting treatment with B12 injections, but treatment should not be delayed for testing to occur.

In addition to these diagnostic blood tests, we test full blood count, urea and electrolytes, liver function tests and thyroid function tests in all patients as part of the routine medical workup (figure 1). In N2O-SACD, the mean corpuscular volume and haemoglobin are usually normal,13 16 21 in contrast to traditional SACD.33 Folate should be measured in all patients to identify codeficiencies. We offer HIV and syphilis testing to all patients, as both can cause treatable myelopathy that can present similarly to N2O-SACD. In addition, serum copper and zinc concentrations can be tested if the history of N2O use is not classical or if copper deficiency myeloneuropathy is a possibility; some patients anecdotally have multiple deficiencies simultaneously.23 In addition to the above tests, clinicians should consider the possibility of coexisting B12 malabsorption, which may have been unmasked by N2O use. This is particularly pertinent in those patients with an initial B12 concentration below the reference range or N2O-SACD after only modest N2O use. In these cases, clinicians should test serum anti-intrinsic factor and antigastric parietal cell antibodies to investigate for pernicious anaemia, and consider testing for coeliac disease and markers of generalised malabsorption (eg, ferritin and vitamin D).34


MR scans of the cervical and thoracic cord may both confirm the diagnosis and exclude alternative causes. In 50%–100% of patients with N2O-SACD, MRI shows dorsal column T2 hyperintensity in the cervical spinal cord with C3–C5 most commonly affected (figure 2B).13 16 21 Axial slices are particularly useful to see the classical ‘inverted V’ or ‘rabbit ears’ sign (figure 2A).35

Figure 2

Classical MRI findings in nitrous oxide-induced subacute combined degeneration of the cord. The T2 axial image (A) shows hyperintense signal in the dorsal columns of the cervical spinal cord, creating an ‘inverted V’ sign. The sagittal image (B) shows the commonly found pattern of hyperintensity extending longitudinally over several cervical spinal segments.

Given the currently high prevalence of N2O use, clinicians should be aware that N2O users may have an alternative cause for a spinal cord syndrome, including inflammatory processes, cervical cord compression and hereditary syndromes. MRI is particularly useful to look for these other causes, since each has a distinct MRI appearance (described elsewhere).36 However, N2O-SACD cannot be distinguished solely by imaging, as several other conditions exactly share its MRI findings; these include myelopathy caused by HIV infection, copper deficiency, vitamin E deficiency and methotrexate toxicity.37 MRI also cannot differentiate traditional SACD from N2O-SACD, although N2O-SACD may have less involvement of the thoracic cord, wider lesions on sagittal MRI, fewer spinal segments with lesions and a higher rate of the inverted V sign.19

We request MRI of the cervical cord on most patients with suspected N2O-SACD. The exception is for patients with a very clear history of recently heavy N2O use and mild symptoms that resolve with B12 therapy and who have stopped N2O; for such cases, imaging may not be necessary. We do not recommend routine brain MRI, unless suspecting an alternative disease process. Brain imaging in case series has either been universally normal,16 or it has shown T2/fluid attenuated inversion recovery (FLAIR) hyperintense white matter lesions in the frontal, periventricular and centrum semiovale regions.21

Further investigations

In general, patients with N2O-SACD do not need further investigation. In patients undergoing peripheral neurophysiology studies, the most common finding is a mixed axonal and demyelinating sensorimotor neuropathy, with the axonal component being more common than the demyelinating.16 21 38

Treatment and organisation of B12 injections

Patient education and support

Patient education is the most important aspect of N2O-SACD management. All patients should be warned that N2O has caused their symptoms and that treatment with B12 will not work if they continue to use N2O, since this would inactivate the B12 supplements. However, relapse following return to N2O use is relatively common,19 and so we encourage all patients to contact their local drug and alcohol service to support their abstinence.

Standard treatment regimen

As soon as N2O-SACD is suspected in the emergency department, patients should start alternate daily intramuscular injections of 1 mg hydroxocobalamin without delay, continued for at least 2 weeks.39 40 This treatment regimen is derived from the standard B12 course for traditional SACD.39 In practice, we give the first injection in the emergency department, and subsequent injections as an outpatient on Mondays, Wednesdays and Fridays (figure 1). It is often difficult to organise vitamin B12 injections in primary care, especially at short notice, with real risk of treatment delay or incomplete treatment. In rare cases where a patient needs admission due to loss of functional independence, alternate day injections can start as an inpatient and switch to outpatient after discharge.

When to stop injections

It is often difficult to decide when to stop injections. Standard practice in traditional SACD is to continue alternate day injections until there is no further neurological improvement. We base N2O-SACD treatment similarly on clinical progress. We therefore perform a standardised neurological examination at baseline and at 2 weeks, to track clinical improvement accurately (figure 1). Ideally, this includes an objective measure such as a 10 m walk test.

If the patient’s condition has improved and then plateaued, we consider stopping the acute course of injections. However, if there is ongoing improvement at 2 weeks, we continue for a further 2 weeks (six more injections) and re-review each fortnight with the same standardised neurological examination until improvement has plateaued, at which point we stop the acute course of injections. In patients where N2O is probably the sole cause of SACD and the serum B12 concentration was normal at first presentation, we do not continue B12 supplementation every 3 months long-term. Where coexisting B12 deficiency has probably contributed to a patient’s vulnerability to N2O-SACD, they should continue long-term supplementation with either B12 injections every 3–6 months or oral supplementation, depending on whether the deficiency is dietary or non-dietary.34

If there is no clinical improvement at all at 2 weeks, we rediscuss the diagnosis and scrutinise the patient’s reported abstinence from N2O. This is relatively rare, but it is reasonable to continue the injections for up to 8 weeks, as occasionally there is a delayed clinical improvement. Serum folate should be checked and replaced if low. Methionine or folic acid may also be possible treatment adjuncts41 42; however, there is no clinical evidence to support their routine use, and we do not do this in practice.

Consideration of oral supplementation

Intramuscular B12 supplementation is strongly preferable to oral supplements. However, a small minority of patients find injections unacceptable (eg, severe needle phobia). A potential alternative—but only in preference to no supplementation—is oral cyanocobalamin 1000–2000 mcg once a day.43

Maximising adherence and attendance

Patients with substance abuse issues are at higher risk of missing follow-up appointments and investigations44; our experience with N2O-SACD suggests that heavy recreational N2O users are no exception. However, several of our local interventions have improved follow-up attendance. First, a clear handover pathway between the emergency department, ambulatory care and neurology has helped to minimise the time between appointments and to clarify a standard point of contact for patients (figure 1). Second, standardising B12 injection days to Monday, Wednesday and Friday, at the same location, makes the injections schedule easy to remember. Third, we explicitly warn patients that two or more episodes of unannounced non-attendance may lead to discharge from the clinic. These interventions have improved attendance to appointments in our practice and reduced average time per B12 injection.

Coordination of care

Where the first point of patient contact is in the emergency department, the initial injection, investigations and education occur at that time (figure 1), as patients sometimes do not attend further appointments. When the first point of contact is at the general practice, the first injection and investigations usually occur at the first hospital visit. Such patients often do not require admission but do need an urgent course of regular intramuscular injections and specialist input. This is relatively unique and does not fit easily into existing care delivery structures. Different centres seeing patients with N2O-SACD naturally choose different pathways to deliver this care, which could be centred on neurology day units or on medical ambulatory care. In our practice, we treat patients in medical ambulatory care with close neurology input via a weekly N2O-SACD multidisciplinary team meeting (figure 1). This meeting forms the crux of the patient pathway and consists of a neurologist, ambulatory care consultant and senior ambulatory care nurse. We complete several important steps to aid the overall delivery of care:

  • We review cases and investigations and confirm the diagnosis.

  • We determine the level of further investigation required (eg, MRI, nerve conduction studies/electromyography, testing for alternative causes of B12 deficiency such as pernicious anaemia).

  • We decide how long to continue treatment.

  • We clinically code N2O -SACD to facilitate tracking of cases using the appropriate SNOMED codes: (subacute combined degeneration of the spinal cord due to use of N2O (disorder)–SCTID: 1105051000000102 and N2O misuse (finding)–SCTID: 1104931000000109).

  • We book outpatient neurology follow-up for patients who have completed a treatment course, particularly for those with ongoing symptoms or with an uncertain diagnosis. Patients with classical presentations that have completed the diagnostic workup and recovered may not need further outpatient follow-up.

Medical ambulatory care is often well set up to deliver more intensive outpatient treatment. With close neurology input, we can effectively manage N2O-SACD in this setting. Depending on available resources, other centres may wish to develop similar pathways, assign a dedicated consultant or create new settings to deliver care for this condition.

Key points

  • Serum B12 is often normal in N2O-SACD; functional B12 testing with methylmalonic acid or homocysteine can help to establish the diagnosis.

  • Intramuscular B12 injection is a low-risk, high-impact treatment and should be given as promptly as possible once N2O-SACD is suspected.

  • B12 injections should continue until improvement reaches a plateau; abstinence from N2O is crucial to recovery and allows B12 injections to work.

  • Successful treatment of this condition requires a clear and smooth set of connections between the emergency department, ambulatory care and outpatient neurology.

Further reading

  • Kunam VK, Velayudhan V, Chaudhry ZA, et al. Incomplete cord syndromes: Clinical and imaging review. RadioGraphics 2018;38:1201–22. doi:10.1148/rg.2018170178

  • Gao H, Li W, Ren J, et al. Clinical and MRI differences between patients with subacute combined degeneration of the spinal cord related vs unrelated to recreational nitrous oxide use: a retrospective study. Frontiers in Neurology 2021;12. (accessed 10 Sep 2022).

  • Oussalah A, Julien M, Levy J, et al. Global burden related to nitrous oxide exposure in medical and recreational settings: A systematic review and individual patient data meta-analysis. Journal of Clinical Medicine 2019;8:551. doi:10.3390/jcm8040551

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.

Ethics statements

Patient consent for publication

Ethics approval

Not applicable.



  • Contributors AJN, RMA, JW and AW conceived of the pathway and project. AJN led and supervised the project. RMA provided expert advice on biochemical testing. AJ, AP and LL collected recommendations for the pathway and conducted the Royal London Hospital audit. AP collated and wrote the manuscript. TH, NE, JBL and DN reviewed the manuscript, provided feedback on the pathway and gave expert opinions.

  • Funding The Preventive Neurology Unit is funded by Barts Charity.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed by Neil Anderson, Auckland, New Zealand.

Other content recommended for you