Which common medications deplete vitamins and minerals?

Why drugs change nutrient status

Acid-reducing drugs cut stomach acid, which can impair absorption of B12 and iron^3,5; other drugs are linked to low levels by routes that are less clear. The effect can build slowly, over months or years, and for most of these interactions there are too few high-quality intervention trials to say what to do about them¹⁰.

The strongest links

• Metformin and vitamin B12. In the Diabetes Prevention Program Outcomes Study, low B12 was more common on metformin than placebo at 5 years (4.3 versus 2.3 percent), and each year of use raised the odds of deficiency by 13 percent¹. In a 4.3-year Dutch trial, B12 fell 19 percent more on metformin than placebo². See vitamin B12.
• Acid-reducing drugs. Two or more years of a PPI went with 65 percent higher odds of B12 deficiency (25 percent for H2 blockers)³, with low magnesium (risk ratio 1.43)⁴ and with iron deficiency (odds ratio 2.49, falling after the drug was stopped)⁵. See acid reflux and magnesium.
• Diuretics. In adults 55 and older, thiazides were linked to low sodium, potassium and magnesium, and loop diuretics to low potassium⁶.

Weaker or mixed links

• Statins and CoQ10. CoQ10 supplements eased statin muscle symptoms in a 12-trial meta-analysis⁷ but not in an earlier 6-trial one⁸; both pooled small trials. For what large trials say about statin muscle symptoms, see what lowers cholesterol.
• The pill and B vitamins. Current data do not show lower folate; low-dose pills may lower vitamin B6; B12 markers were not affected⁹.

Where a supplement is standard care

Folic acid alongside methotrexate is an example where supplementing works: in a 79-person trial in rheumatoid arthritis it lowered toxicity without reducing the drug's effect¹¹. This is prescriber-directed, not a do-it-yourself fix.

What to do

If you have taken one of these medicines for years, or have numbness, tingling or anemia, ask for the relevant blood tests. A PPI should be reviewed regularly to confirm you still need it, and the decision to stop should rest on that, not on side-effect worries alone¹². See common medications and their hidden costs and safely reducing or stopping a medication.

Metformin and B12

The Diabetes Prevention Program assigned 2,155 people to placebo (n=1,082) or metformin 850 mg twice daily (n=1,073) for 3.2 years, then offered the metformin group 9 more years of open-label metformin. Low B12 (at or below 203 pg/mL) was more common on metformin at 5 years (4.3 versus 2.3 percent; P=.02) but not at 13 years (7.4 versus 5.4 percent; P=.12). Low or borderline-low B12 (at or below 298 pg/mL) was more common at both points (19.1 versus 9.5 percent, and 20.3 versus 15.6 percent). Anemia was more common on metformin but did not track B12 status, and neuropathy was more common in metformin users with low B12¹. In the Dutch trial, 390 insulin-treated people with type 2 diabetes took metformin 850 mg or placebo three times daily for 4.3 years. B12 was 19 percent lower on metformin (95% CI 14 to 24), the absolute risk of deficiency (below 150 pmol/L) was 7.2 percentage points higher (number needed to harm 13.8 over 4.3 years), and homocysteine was higher in deficient patients².

Acid-reducing drugs

A Kaiser Permanente case-control study compared 25,956 patients with new B12 deficiency with 184,199 controls. A supply of PPIs for two or more years carried an odds ratio of 1.65 (95% CI 1.58 to 1.73) and H2 blockers 1.25 (1.17 to 1.34); more than 1.5 pills a day was stronger (1.95) than under 0.75 (1.63)³. A meta-analysis of nine observational studies (109,798 patients) found a pooled risk ratio of 1.43 (1.08 to 1.88) for low magnesium in PPI users⁴. In a case-control study of 77,046 people with new iron deficiency and 389,314 controls, two or more years of PPIs had an odds ratio of 2.49 (2.35 to 2.64) and H2 blockers 1.58 (1.46 to 1.71). The association was stronger at higher doses (4.27 for more than 1.5 pills a day for 10 years or more) and weaker after stopping⁵.

Diuretics

In the Rotterdam Study of 5,179 people aged 55 and older, 15.0 percent had at least one electrolyte disorder. Thiazides were independently associated with low sodium, potassium and magnesium; loop diuretics with high sodium and low potassium; potassium-sparing diuretics with low sodium. Low sodium and low magnesium were linked to higher mortality⁶.

Statins, CoQ10, the pill and methotrexate

The 2018 meta-analysis of 12 trials (575 patients) found CoQ10 reduced muscle pain (weighted mean difference -1.60), weakness, cramps and tiredness, with no change in creatine kinase⁷. The 2015 meta-analysis of 6 trials (302 patients) found no significant effect on muscle pain (standardized mean difference -0.53, 95% CI -1.33 to 0.28) or creatine kinase, and called for larger trials⁸. For the pill, a review concluded that current data do not support lower folate status, that low-dose pills may lower vitamin B6, and that functional B12 indicators are not significantly affected, with definitive answers awaiting more work⁹. In the methotrexate trial, 79 people with rheumatoid arthritis took placebo or 5 mg or 27.5 mg of folic acid weekly. Folic acid left efficacy unchanged and lowered toxicity scores (P at or below .001)¹¹.

Where the evidence conflicts

Most acid-reducer and diuretic data are observational, so differences between people who take these drugs and people who do not cannot be ruled out^3,4,5,6. The two CoQ10 meta-analyses disagree^7,8. A low lab value is not always illness: in the metformin study, anemia did not track B12 status¹. A review of commonly prescribed drugs concluded that most interactions need better trials, that widespread supplementation is not recommended, and that at-risk patients should reach recommended intakes¹⁰. The American Gastroenterological Association advises regular review of every PPI prescription, a trial of de-prescribing when there is no definite indication, a warning about rebound acid symptoms, and basing the decision to stop on lack of indication rather than adverse-event worries¹².

Bottom line

Confidence is moderate (3 of 5). Metformin lowers B12 in randomized trials^1,2, and acid reducers and diuretics are linked to low B12, iron, magnesium and potassium in large observational studies^3,4,5,6. Statin-CoQ10 and pill-B vitamin links are weak or mixed^7,8,9. The practical step is a blood test and a medicine review, not a self-started supplement or a stopped prescription.