Metformin Side Effects: Trial Incidence Rates, GI Timing and B12 Monitoring

Almost every summary of metformin side effects gives the same undifferentiated list: nausea, diarrhoea, stomach upset, metallic taste, rare lactic acidosis. That list is accurate and close to useless, because it does not say how often any of it happens, how the numbers change between the two formulations, or which symptoms fade and which do not.

The registration trials that supported approval reported specific rates against placebo. Those numbers are the ones worth knowing, because they are what separates a common and self-limiting problem from a rare and serious one.

Immediate-release metformin: reported rates against placebo

In the US clinical trial that supported approval, 141 patients received metformin at up to 2,550 mg per day and 145 received placebo. Adverse reactions occurring in more than 5 percent of metformin patients, and more often than on placebo, were reported as follows.

Adverse reaction	Metformin (n=141)	Placebo (n=145)
Diarrhoea	53.2%	11.7%
Nausea and vomiting	25.5%	8.3%
Flatulence	12.1%	5.5%
Asthenia (weakness)	9.2%	5.5%
Indigestion	7.1%	4.1%
Abdominal discomfort	6.4%	4.8%
Headache	5.7%	4.8%

Two things stand out. Diarrhoea was reported by more than half of patients, which is why it dominates every real-world account of starting metformin. And it caused discontinuation in 6 percent of patients, meaning the large majority who experienced it carried on.

The gap between the metformin and placebo columns is also worth reading carefully. Headache at 5.7 percent versus 4.8 percent is barely a signal. Diarrhoea at 53.2 percent versus 11.7 percent is unmistakable. Grouping them in one bullet list, as most articles do, flattens a difference of an order of magnitude.

A further set of reactions was reported in 1 to 5 percent of metformin patients and more often than placebo: abnormal stools, hypoglycaemia, myalgia, lightheadedness, dyspnoea, nail disorder, rash, increased sweating, taste disorder, chest discomfort, chills, flu syndrome, flushing and palpitation. Note where hypoglycaemia sits. On metformin alone it is in the 1 to 5 percent band, not the common one, which is a defining difference from the sulfonylureas.

Extended-release metformin: the same drug, different numbers

In placebo-controlled trials of extended-release metformin, 781 patients received the drug and 195 received placebo.

Adverse reaction	Extended-release (n=781)	Placebo (n=195)
Diarrhoea	9.6%	2.6%
Nausea and vomiting	6.5%	1.5%

Diarrhoea led to discontinuation in 0.6 percent of extended-release patients, compared with 6 percent on immediate-release. That is a tenfold difference in the rate at which the side effect ended treatment.

These figures come from separate trials with different designs and populations, so they cannot be read as a head-to-head comparison. They do explain why switching formulation is the standard first response to gastrointestinal intolerance, and why Glycomet 500mg SR, Glycomet 850mg SR and Metsmall 500 SR exist alongside the plain tablets.

Reactions reported in 1 to 5 percent of extended-release patients were abdominal pain, constipation, abdominal distention, dyspepsia and heartburn, flatulence, dizziness, headache, upper respiratory infection and taste disturbance.

When the gastrointestinal effects arrive and when they leave

The label makes a clinically important observation that rarely gets quoted. Gastrointestinal symptoms are common during initiation of therapy, but once a patient is stable on a given dose, later gastrointestinal symptoms are unlikely to be caused by the drug and could indicate lactic acidosis or another serious condition.

That gives a usable rule. Nausea and loose stools in the first two weeks, or in the days after a dose increase, follow the expected pattern. The same symptoms appearing months into stable treatment do not, and deserve investigation rather than reassurance.

Practical measures that reduce early symptoms without changing the drug: take every dose with food rather than before or between meals, hold each titration step for the full week before stepping up, and start at a lower strength such as Glycomet 250mg if 500 mg twice daily proved intolerable. The full escalation schedule is set out in our metformin dosage guide.

Vitamin B12: the side effect that produces no symptoms early

In metformin trials of 29 weeks duration, a decrease to subnormal levels of previously normal serum vitamin B12 was observed in approximately 7 percent of patients.

This matters because B12 deficiency is silent for a long time and then presents as anaemia or peripheral neuropathy, and peripheral neuropathy in a person with diabetes is easily attributed to the diabetes itself. Periodic B12 measurement is recommended with long-term metformin use, and it moves from optional to necessary in anyone who develops anaemia or new neuropathic symptoms.

Lactic acidosis: rare, serious, and predictable

Metformin carries a boxed warning for lactic acidosis. The rate is genuinely low. The reported incidence is approximately 0.03 cases per 1,000 patient-years, with approximately 0.015 fatal cases per 1,000 patient-years. In more than 20,000 patient-years of exposure in clinical trials there were no reported cases at all. When it does occur it is fatal in roughly half of cases.

The onset is described in the label as subtle, with non-specific symptoms: malaise, myalgia, respiratory distress, increasing somnolence and abdominal discomfort. More marked acidosis can bring hypothermia, hypotension and resistant bradyarrhythmias. Laboratory findings include blood lactate above 5 mmol/L, anion gap acidosis without ketonuria or ketonaemia, a raised lactate to pyruvate ratio, and metformin plasma levels generally above 5 mcg/mL.

The risk factors are the useful part, because nearly every reported case involved one of them:

• Renal impairment, which is the dominant factor
• Age 65 years or older
• Radiological studies using iodinated contrast
• Surgery and other procedures
• Hypoxic states such as acute congestive heart failure
• Excessive alcohol intake, acute or chronic, which potentiates the effect of metformin on lactate metabolism
• Hepatic impairment, which limits the ability to clear lactate
• Certain concomitant drugs, including carbonic anhydrase inhibitors such as topiramate

Lactic acidosis is a medical emergency treated in hospital. Metformin is dialysable, with clearance of up to 170 mL/min under good haemodynamic conditions, so prompt haemodialysis is recommended to correct the acidosis and remove the accumulated drug.

One caution against over-reading a single number: fasting venous plasma lactate above the upper limit of normal but below 5 mmol/L does not by itself indicate impending lactic acidosis, and may reflect poorly controlled diabetes, obesity, vigorous exercise or sample handling.

Overdose

Metformin overdose has been reported with ingestions greater than 50 grams. Hypoglycaemia was reported in approximately 10 percent of cases, although no causal association with metformin was established. Massive overdose can produce metabolic acidosis with a high lactate, and is managed with supportive care and haemodialysis in a hospital setting.

Side effects that are commonly attributed to metformin but are not the drug

Hypoglycaemia is the clearest example. On metformin alone it sits in the 1 to 5 percent band, and metformin does not stimulate insulin release. When someone on metformin has repeated hypoglycaemic episodes, the usual explanation is a second agent, most often a sulfonylurea, either as a separate tablet or inside a combination product such as Glycomet-GP 1 or Amaryl M 1mg. The mechanistic difference between the two drug classes is set out in our comparison of metformin and glimepiride, and the sulfonylurea side is covered in the Amaryl and glimepiride guide.

Weight change is another. Metformin is broadly weight neutral or associated with modest weight loss, so weight gain during treatment usually points to something else in the regimen.

What actually warrants stopping

Most metformin side effects are dose-related, front-loaded and manageable by slowing titration or changing formulation. The situations that call for stopping and seeking medical attention are narrower: symptoms suggesting lactic acidosis, particularly new gastrointestinal upset after a long stable period; any acute illness with dehydration, vomiting or reduced fluid intake; and the defined pauses around contrast imaging and surgery.

Women taking metformin for polycystic ovary syndrome face the same side effect profile with some additional considerations, covered in our guide to metformin for PCOS. For how metformin compares with the other oral options on tolerability, see the type 2 diabetes treatment overview.

Medical disclaimer

This article is educational and does not replace advice from a qualified healthcare professional. Metformin requires a prescription in the United States, United Kingdom, Canada and Australia. Trial incidence rates come from studies run under specific conditions and cannot be directly compared with rates from trials of other drugs or applied to any individual. If you develop symptoms that concern you while taking metformin, contact your prescriber. Symptoms suggesting lactic acidosis need urgent medical assessment.

References

• FDA-approved prescribing information for Glucophage and Glucophage XR (metformin hydrochloride), Adverse Reactions, Table 1 and Table 2.
• DailyMed, Metformin Hydrochloride Tablets USP, Adverse Reactions, Warnings and Overdosage sections.
• FDA-approved labelling, boxed warning on metformin-associated lactic acidosis.
• Graham GG et al. Clinical pharmacokinetics of metformin. PubMed PMID: 21241070.
• Dell''Aglio DM et al. Massive metformin overdose. PubMed PMID: 29534338.