Case Study 1 — Nine Years of Metformin: How a Deficiency Hides Behind a Diagnosis

A clinical case. Walt Prosser is an illustrative composite; the metformin–B12 interaction, the diagnostic trap, and the clinical guidance are not.

Setup

Walt Prosser, 68. Type 2 diabetes for eleven years, on metformin for nine. Also takes omeprazole for reflux, which he has been on for about four years and which nobody in this story ever mentioned because it isn't a diabetes drug.

The presenting complaint was mentioned on the way out of the room.

Pins and needles in both feet. Worse at night. Gradual onset over eighteen months, maybe two years. He had mentioned it once, to a nurse at a diabetes review, and been told it was diabetic neuropathy — which is a completely reasonable thing to be told, because it is by far the most common cause of exactly that presentation in exactly that patient.

It was also, in his case, probably wrong.

The three stacked mechanisms

Here's what was actually happening, and the reason it's worth a case study is that three independent mechanisms were pointing the same direction and each one alone would have been easy to dismiss.

1. Metformin

Metformin interferes with the calcium-dependent uptake of the B12–intrinsic factor complex at the ileum (Chapter 3 §3.4). The effect is dose- and duration-dependent, and Walt had nine years of exposure at a therapeutic dose.

This is well enough established that the American Diabetes Association's Standards of Care recommends periodic B12 measurement in metformin-treated patients, particularly those with anemia or peripheral neuropathy.

Walt had peripheral neuropathy and had never been tested.

2. Long-term acid suppression

B12 in food is bound to protein and must be liberated by gastric acid and pepsin before intrinsic factor can bind it. Proton pump inhibitors substantially reduce gastric acid, and long-term use is associated with reduced B12 status.

Four years of omeprazole. Nobody had connected it to the diabetes drug, because they sit in different mental categories.

3. Age

Gastric acid production declines with age independently of medication, and atrophic gastritis — common in older adults — impairs B12 liberation further.

Three routes, stacked, in one man. And the crucial feature: each one alone might not have caused clinical deficiency. Together they did.

Why the multivitamin made it worse

This is the part that reframes the whole story.

Walt had been taking a daily multivitamin for years — one of his nine bottles, taken faithfully with his morning coffee. It contained about 6 µg of B12, which is a perfectly sensible amount and comfortably above the RDA.

And it was almost entirely irrelevant to his situation.

When absorption is the bottleneck, a physiological dose doesn't get through. The therapeutic approach to malabsorptive B12 deficiency uses either high-dose oral B12 — where a small fraction is absorbed by passive diffusion independent of intrinsic factor — or intramuscular injection, bypassing the gut entirely.

Six micrograms does neither.

💡 Aha moment. The multivitamin didn't just fail to help. It actively obstructed the diagnosis.

Walt knew he took a multivitamin containing B12. So did his family. If anyone had raised the possibility of a vitamin deficiency, the answer was available and reassuring: he takes a multivitamin.

The appearance of coverage is worse than no coverage, because no coverage prompts a question and apparent coverage closes it. This is Chapter 13 §13.3's threshold in its most consequential form: intake was never the variable, and the intervention addressed intake.

The diagnostic trap

I want to be fair to whoever told him it was diabetic neuropathy, because the reasoning was sound.

Base rates favour it. Peripheral neuropathy in a long-standing type 2 diabetic is diabetic neuropathy until proven otherwise, most of the time. Roughly a substantial minority of people with diabetes develop it. It presents exactly like this: distal, symmetrical, worse at night, gradual.

The problem is that a plausible diagnosis stops the search. Once "diabetic neuropathy" was written down, the symptom had an owner. It was no longer an unexplained finding requiring investigation; it was an expected complication requiring management.

And B12 deficiency and diabetic neuropathy are clinically similar enough that you cannot reliably distinguish them at the bedside. Both are distal and symmetrical. Both are worse at night. The distinguishing features — proprioceptive loss, sometimes brisk reflexes in B12 deficiency, cognitive or mood changes — are subtle and inconsistent.

The only way to tell is to test. Which costs very little and takes a week.

⚠️ This is the clinically important point of the whole chapter. In a person with diabetes on long-term metformin who develops peripheral neuropathy, B12 deficiency is not an exotic alternative diagnosis — it is a common, cheap-to-exclude, and potentially reversible cause that sits underneath the obvious one. And the consequence of missing it is not trivial: B12-related neurological damage can become irreversible if prolonged.

What we did

Tested. B12 was low. Methylmalonic acid — which rises when B12 is functionally inadequate at tissue level and is useful when serum B12 is borderline — supported it.

Treated, through his physician, with high-dose oral B12 rather than the multivitamin. (Injection is an alternative; high-dose oral is often effective even in malabsorption because a small fraction crosses by passive diffusion, and it avoids repeat appointments.)

Reviewed everything else, which is where Chapter 11's Case Study 1 picks up: five of the nine supplements stopped, two kept, one referred back to his prescriber, and about $1,850 a year returned.

Added B12 monitoring to his ongoing diabetes reviews.

And added the list. Every supplement, with doses, in his phone and on his record — which is how the biotin problem in Chapter 16 was found.

What happened

Partial improvement. The tingling reduced over several months and did not fully resolve.

I want to be honest about that, because a case study that ends in complete recovery would be a better story and a worse lesson.

Two reasons the recovery was incomplete:

He probably has both. Eleven years of diabetes with imperfect glycemic control is enough to cause genuine diabetic neuropathy, and the B12 deficiency was likely superimposed on it. The original diagnosis wasn't wrong — it was incomplete, which is a different and more common failure than being wrong.

And time matters. Eighteen months to two years of symptomatic B12 deficiency is long enough for some neurological change to become permanent. Had this been caught at six months, the outcome would probably have been better — and the thing that would have caught it was a blood test that costs very little, in a patient who met an explicit guideline criterion for having it.

Analysis

1. Three mechanisms stacked, and each alone was dismissible. Metformin, PPI, age. Individually, none would necessarily produce clinical deficiency. The clinical skill is noticing accumulation, and it requires someone to be looking at the whole medication list rather than one drug at a time.

2. The multivitamin was worse than nothing. Not because it caused harm, but because it supplied a reassuring answer to a question that should have stayed open. This generalizes well beyond B12.

3. A plausible diagnosis is a search-terminator. The most dangerous kind of wrong answer is one that fits. "Diabetic neuropathy" was consistent with everything, required no further work, and had a management pathway. Correct-sounding explanations end investigations, and this is not a failure of individual clinicians so much as a property of how diagnosis works under time pressure.

4. The guideline already said to test. This isn't a case where the knowledge was missing. The ADA recommendation exists and is well known. The gap was between guidance and practice, which is where most preventable harm in medicine lives and which no amount of new research addresses.

5. And a limit on my own story. ⚠️ I did not diagnose or treat this. I noticed a pattern in a dietary consultation and referred it. That's the appropriate scope for a dietitian, and a book that implied otherwise would be teaching something harmful. What a nutrition professional can do is notice, ask, and route — and in this case, noticing took about ninety seconds and mattered more than anything else in the appointment.

Discussion Questions

1. Three mechanisms stacked. What system change would make accumulation more visible — a flag in the record, a pharmacist review, something else? What would each cost?

2. The multivitamin "supplied a reassuring answer to a question that should have stayed open." Name two other places in this book where a supplement or product has done the same thing.

3. The original diagnosis was incomplete rather than wrong. How should that change how a clinician documents a diagnosis? Would "diabetic neuropathy — B12 not yet excluded" have helped?

4. The ADA guidance already recommended testing. Why is the gap between guidance and practice so durable? What would actually close it for this specific recommendation?

5. Recovery was partial because of delay. How should that affect the threshold for testing in asymptomatic metformin-treated patients? Argue for and against routine periodic screening.

Your Turn

If you, or anyone you feed, takes metformin, a proton pump inhibitor, or is over 65 — this exercise is not hypothetical.

1. List the medications. All of them, including over-the-counter and long-term ones people forget — PPIs and antacids especially.
2. Check for stacking. Does more than one item on that list affect the same nutrient? Metformin and PPIs both affect B12. Several diuretics affect potassium and magnesium (Chapter 14).
3. Check the supplements against the actual problem. If there's a supplement in the cabinet for something, ask: does it address intake, absorption, or requirement — and which one is the actual bottleneck?
4. Ask about B12 at the next appointment if any of the §13.5 risk factors apply. Specifically: "I'm on metformin — should my B12 be checked?" Nine words.
5. Put the list in your phone.

Readers of drafts of this chapter have come back with two confirmed low-B12 results in relatives. It is not a rare finding. It is a rarely-looked-for one, and looking costs a sentence.