This site is intended for health professionals only

Overdiagnosis: B12 deficiency

Overdiagnosis: B12 deficiency

As part of our series on conditions that may be overdiagnosed, GP partner and trainer Dr David Coleman considers why and when we might be placing too much stock on vitamin B12 deficiency


Vitamin B12 deficiency is a relatively common condition with a range of causes and innumerable symptoms. Its prevalence peaks in the elderly population. The variable and often vague presentation, combined with the limitations of laboratory testing, often create a diagnostic challenge. The most common cause of vitamin B12 deficiency is the autoimmune disorder pernicious anaemia1, which affects the gastric mucosa and impacts the absorption of the vitamin.

B12 deficiency has become a contentious area of practice, with patients commonly reporting that they feel let down over their management3. This feeling was magnified during the early stages of the Covid pandemic when many practices paused injection clinics and converted patients to oral medication. A number of patient-focused campaign groups present a narrative of under-diagnosis and poor treatment. Meanwhile, many clinicians feel B12 deficiency is often overdiagnosed. My perspective is that our approach in general practice is inconsistent and that diagnoses will be missed and inappropriately made.

Earlier this year, NICE published its first-ever draft guideline on the diagnosis and management of vitamin B12 deficiency in over-16s. The guidance recommends testing for people with at least one sign or symptom and one or more risk factors for the condition.

Actual prevalence

Approximately 6% of the British population is thought to be affected2, but there is a surprisingly lack of clarity regarding the condition’s diagnosis and management.

Common features

Patients may present with neurological symptoms (peripheral neuropathy, visual disturbance, cognitive features), haematological problems (anaemia, macrocytosis) or gastrointestinal issues suggestive of malabsorption. They may also describe non-specific features, such as fatigue or be entirely asymptomatic.

Vitamin B12 testing is recommended for patients with suggestive symptoms, but not routinely for non-specific symptoms unless initial tests revealed macrocytosis, for example.

Diagnostic criteria vary around the country based on local laboratory ranges and available tests. Most regions recommend different approaches based on the presence or absence of objective parameters/symptoms. Where I practise in Doncaster4, if the patient has neurological symptoms or anaemia and a B12 level of <187ng/L we would diagnose vitamin B12 deficiency and treat accordingly with IM injections. Without clear symptoms, a level of 150-187ng/L would trigger a repeat test and an oral trial of treatment only if the borderline low result persisted.

Reasons for overdiagnosis

A recent UK study examining diagnostic testing rates in general practice between 2000 and 2015 highlighted a significant linear (around 17% annually) increase in vitamin B12 requests5. Other studies have shown that laboratory tests for vitamin B12 are mainly used for patients with non-specific symptoms rather than the recommended population6.

There are both patient and clinician factors for the increase in inappropriate testing. Many doctors find themselves practising defensively in an age of increasing litigation, labouring under the belief that more is better, or at least safer. Increasing workforce skill-mix may be a factor, too; non-medical clinicians may not have received the same depth of training in biochemistry, which could impact requesting habits. The intense workload, pressure on GP supervisor time, and heightened focus on passing exams may be impacting the ability of GP trainees to develop judicious skills in this area.

The wellness boom, magazine articles urging readers to boost their vitamin B12, and stories of celebrities paying for private vitamin injections7, coupled with the narrative of doctors ‘fobbing off’ patients and minimising the condition, may create unrealistic expectations regarding testing.

The final factor is the limitations of the tests themselves. Serum vitamin B12 assay has a sensitivity of around 97% at levels below 200 ng/L, but specificity is lower8. Intrinsic antibody testing is highly specific but has a low sensitivity of 50-60%9. No test is perfect, but when we test the wrong cohort of patients our imperfect tests become even less helpful.

Making improvements and supporting patients

I would recommend a two-pronged approach to improving diagnostic accuracy of diagnosis in practice:

  1. A practice-wide educational meeting to make sure all clinicians are aware of the latest local guidelines for testing and diagnosis. It may be that the process of testing for pernicious anaemia is being overlooked and that interval tests are not being done for borderline cases with minimal/non-specific symptoms. This will reduce inappropriate diagnoses and improve accuracy in the future.
  2. Review all patients with an active diagnosis of vitamin B12 to identify if the diagnostic guidelines were followed. If not, consider pausing treatment and retesting. This will require an element of transparency and patient consent. There are resources to support such conversations, such as those found on the Choosing Wisely website10.

System-level changes like user prompts when requesting bloods, or even making commonly over-requested blood tests harder to find, can also reduce unnecessary testing.

Vitamin B12 deficiency is both easily missed – sometimes with significant harm – and overdiagnosed. This places clinicians in a challenging and unenviable position. The key to overcoming this is a consistent approach and careful consideration regarding who to test. As we work at greater scale and with a wider range of clinicians, educational sessions and regular audit of our practice are key to maintaining this consistency.

David Coleman is a GP partner and trainer in Doncaster, South Yorkshire


  1. Andres E, Serraj K. Optimal management of pernicious anemia. J Blood Med 2012;3:9103.
  2. Hunt A, Harrington D, Robinson S. Vitamin B12 deficiency. BMJ 2014;349:g5226.
  3. Tyler N, Hodkinson A, Ahlam N, et al. Patient safety, self-injection, and B12 deficiency: a UK cross-sectional survey. British Journal of General Practice 2022;72(725):e891-e898.
  4. Clinical Guideline: Investigation and Management of Anaemia within DBTH
  5. O’Sullivan JW, Stevens S, Hobbs FDR, et al. Temporal trends in use of tests in UK primary care, 2000-15: retrospective analysis of 250 million tests. BMJ 2018;363:k4666.
  6. Koch H, van Bokhoven MA, ter Riet G, et al. What makes general practitioners order blood tests for patients with unexplained complaints? A cross-sectional study. European Journal of General Practice 2009;15(1):22-28.
  7. The Guardian, 2019: Celebrities help the £500 vitamin jab go mainstream
  8. Mohamed M. Pernicious anaemia BMJ 2020;369:m1319.
  9. Sukumar N, Saravanan P. Investigating vitamin B12 deficiency. BMJ 2019;365:l1865.
  10. Choosing Wisely


Visit Pulse Reference for details on 140 symptoms, including easily searchable symptoms and categories, offering you a free platform to check symptoms and receive potential diagnoses during consultations.


Please note, only GPs are permitted to add comments to articles

David Church 7 September, 2023 1:06 pm

People with the neurological symptoms, OR being in risk populations for malabsorption, (elderly, known malabsorption, GIT surgery, alcohol excess), we would check FBC, plus LFT, and other items as perhaps indicated.
If Neurological symptoms OR {risk population PLUS (Macrocytosis OR anaemia)} , I would request B12 and folate levels, alongside checking that we had already some idea of dietary sufficiency, and possibly VitD (though we are encouraged to supplement rather than test for VitD deficiency, in any risk groups).
If everything else is normal, and B12 is low or low-normal, we would supplement or re-test, respectively (considering any symptoms, and correcting any dietary factors.
Initially, we would supplement orally and send antibody request.
If they have antibodies or a re-test is lower, then they would get lifelong injections. I would never repeat the B12 test, but many do so annually, and get extremely high results.
If they respond to oral supplements, I would re-test after 1 year: if levels higher, continue supplements for life – at which point, there is the debate between expensive tablets or ‘risky’ injections. That debate needs more input.
If they do not respond to oral supplements, (or antibodies were positive), then they have pernicious anaemia.
This avoids the famous, but famously unavailable, ‘Schilling test’.
I have notice a lot of test requests do not follow, and I will now go see if the new guideline covers the remaining questions, and gets us all doing the same. i suspect the reason we were not is related to the difficulties around when the schilling test became unavailable. Hence the ‘Modified Geraint Test’, described above.

SUBHASH BHATT 7 September, 2023 1:28 pm

Fatigue is interesting symptom of all diseases . Ask patient “do you fill tired”. Not many say no.
Did not need guide lines for bread and butter medicine.

Trefor roscoe 8 September, 2023 7:22 pm

Excellent article. I did an audit over 20 years ago with a trainer. We found over a third of those on B12 injections had did not fulfill the proper diagnostic criteria.

Back then we had the capacity to bring these patients in and get them off injections

Some Bloke 8 September, 2023 7:24 pm

The most common cause of “b12 deficiency” is not pernicious anaemia. Thankfully prevalence of illness is really low, around 0.1% in general, higher in the elderly. So unless your practice population is incredibly old, you should not have more than ten- twenty of people who need regular injections per ten thousand patients. Add few more with massive gut or stomach resections. A growing number of obesity gut surgery- unclear how to interpret that.
The most common cause of “deficiency” is biodegradability of the stuff once outside the body. If you have patient with unexpected macrocytic anaemia and call the lab to ask to add on b12, they will laugh and inform you that the sample has to be very fresh. I bet that prevalence of so called deficiency, or whatever it is that we are medicating, directly correlate with distance from practice to the lab or time interval between venepuncture and processing in the labs.
I rarely start new “deficiency” cases on injections. There’s good evidence that oral supps will work for nearly all, including pernicious anaemia, cases. 1 mg a day is your annual requirement 365 days a year. Enough will get into blood stream by osmosis. Much more tricky to work out who benefits more from a particular form of vitamin – hydroxy-, cyano-, methyl -… All science is: trial and error only.
Beware strong loony b12 advocacy groups. Do not anger that lot. Aggressive bunch akin to human hating vegan animal rights maniacs.

Reply moderated
Dave Haddock 9 September, 2023 8:30 pm

Would a similar article on the vitamin D fetish would be helpful?

Some Bloke 10 September, 2023 5:38 pm

David, excellent idea. May I also suggest reviewing “diagnostic process” for “anxiety and depression” (crawling under cover fully expecting a s.h.t,storm of righteous outrage to head my way).

Richard Harper 12 September, 2023 9:48 am

Not only overdiagnosis, but overtreatment. The loading dose regimen is clearly a haematology directed regimen, with no consideration of GP resources.
NICE is advising 6 loading doses over 2 weeks based on expert advice. That’s 6 nurse appointments and the time, travel and environmental costs of patients attending these appointments.
I have never heard of a single patient who came to harm from having a single loading dose of B12. Maybe if severe megaloblastic anaemia, but haven’t seen one of those for years.