Does Metformin Deplete Magnesium? What 15+ Years of Research Actually Shows.
If you take Metformin for type 2 diabetes, the research suggests your magnesium is probably running lower than it would be otherwise. And your routine blood test is almost certainly not catching the full picture. The studies have existed for over fifteen years. They are published in major peer-reviewed journals. A possible cellular cause has been identified. The topic just hasn't reached most doctor's offices yet.
This article walks through what the research shows. Why standard testing has limits. And what low magnesium tends to look like in the body. It is not medical advice. We are not telling anyone what to take or what dose to take. Those are conversations for a person and their doctor. The goal is to translate the research into plain language. So you can have a more informed version of that conversation.
What the research actually shows
The link between Metformin and lower magnesium is not new. The first major community study was published in 2013 in PLOS ONE. It looked at type 2 diabetics in Australia. Patients on Metformin had clearly lower magnesium levels than diabetics managed by diet alone. A higher percentage of them met the line for low magnesium.1
This was not a one-off finding. A 2020 study in Canadian Journal of Physiology and Pharmacology proposed a cellular reason. Metformin lowers the activity of TRPM6, a channel the body uses to absorb magnesium from food and hold onto it in the kidneys. Short-term use raised TRPM6 activity. Long-term use lowered it.2
In plain language: the longer Metformin is taken, the less efficiently the body holds onto magnesium.
The picture is mostly consistent across the research, though not unanimous. A 2019 study in Scientific Reports found that diabetic mice had lower magnesium. But Metformin treatment did not change the low magnesium in that animal model. That counter-finding is worth noting.3 The directional pattern still holds across most human studies. The exact mechanism is still being worked out.
A 2025 review in Kidney Medicine, written by researchers at Massachusetts General Hospital, summed up where the field stands. Long-term Metformin use is now discussed in kidney medicine literature as a possible cause of low magnesium in some type 2 diabetes patients. Two main mechanisms are proposed: lowered TRPM6 activity and losses through the gut.4
The short version of fifteen years of research: Metformin use is linked to lower magnesium in most studies that have looked. A possible cellular cause has been proposed. The effect appears to grow with longer treatment. The topic is now part of kidney and hormone medicine literature.
Why standard testing has limits
The test most doctors order for magnesium is called serum magnesium. It measures the magnesium in the liquid part of the blood. It is cheap, widely available, and part of routine bloodwork.
The limit of this test is anatomical. According to a 2010 review in Magnesium Research, less than 1% of total body magnesium is in the bloodstream. The rest is inside cells, bones, and soft tissues. That is where magnesium does most of its work.5
The body protects the blood number. When magnesium runs low, the body pulls it from inside the cells to keep blood levels normal. The blood test can read normal long after cellular stores have been declining.
A normal blood magnesium result does not mean magnesium status is adequate. It means the body has been doing the work to keep the blood number normal — by drawing from somewhere else.
A 2018 review in Open Heart discussed what the authors call subclinical magnesium deficiency. Cellular magnesium is low, but blood magnesium reads normal. The authors argued this state may affect a meaningful portion of the general population. Higher rates show up in people with diabetes, those on certain medications, and those under chronic stress.6
More accurate tests exist. RBC magnesium, ionized magnesium, and magnesium loading tests are all options. But they are not part of standard bloodwork. Whether they make sense in any individual case is a clinical judgment, not something to decide from an article.
What low magnesium tends to look like
Magnesium is involved in over 600 reactions in the body, according to a 2015 review in Physiological Reviews.7 When magnesium runs low, several body systems can show it.
Sleep
Magnesium is involved in the activity of GABA. GABA is a brain chemical the body uses to calm down and fall asleep.8 Low magnesium has been linked in observational studies to worse sleep quality.
Muscle function
Magnesium plays a role in muscle relaxation after a contraction. It works against calcium at the point where nerves meet muscle.7 Low magnesium is linked to cramps and muscle twitches in some groups.
Energy
ATP is the molecule cells use to store and transfer energy. In the body, ATP is mostly bound to magnesium.7 Magnesium is needed for the reactions that produce and use ATP.
Nerves and stress
Magnesium is involved in the part of the nervous system that controls stress response and mood.7
None of these signs is unique to magnesium. Each can have other causes. The presence of any of them is not a diagnosis. And the standard blood test alone may not be enough to assess status.
About modern diets and intake
According to a 2012 review in Nutrition Reviews, a large portion of the U.S. population does not meet the recommended daily intake for magnesium through diet alone.9 The 2015 Physiological Reviews paper notes that magnesium content in fruits and vegetables has dropped by an estimated 20 to 30% over the past several decades. That decline is linked in part to changes in farming practice and soil composition.7
The Recommended Dietary Allowance for adults is 310 to 320 mg per day for women and 400 to 420 mg per day for men, depending on age. Foods relatively dense in magnesium include pumpkin seeds, almonds, spinach, black beans, dark chocolate, and avocado.
About supplement forms
Not all magnesium supplements are absorbed equally. A 2001 study in Magnesium Research by Firoz and Graber compared four common forms — magnesium oxide, magnesium chloride, magnesium lactate, and magnesium aspartate. They reported that magnesium oxide showed clearly lower absorption than the other three.10 A 2003 study in the same journal compared citrate, oxide, and amino acid chelate. It also found oxide absorbed worse than the other forms.11
The number on a supplement label that shows how much magnesium the product actually delivers is called elemental magnesium. It is listed in the supplement facts panel. It can be very different from the total compound weight on the front of the bottle. Magnesium glycinate, for example, is about 14% elemental magnesium by weight. So a 1,000 mg dose of the glycinate compound delivers about 140 mg of elemental magnesium.
For anyone reading a label: the compound weight on the front isn't what the body uses. The elemental number on the back is.
What to do with all this
This article is a research summary, not clinical guidance. The studies cited describe patterns and mechanisms seen in groups and in lab settings. They don't establish what any specific person should do.
For anyone on long-term medication and curious about whether any of this applies to them, the next step is a conversation with a doctor. Useful questions include: What does my recent bloodwork show? Are there other tests that would be useful? Are there interactions between my medications and nutrient absorption I should know about?
The research is the research. What to do with it is a clinical question.
The honest summary
The Metformin–magnesium research has been building since the early 2010s. The mechanism is biologically plausible, and a possible pathway has been identified. The directional link holds across most human studies, though not all animal models. The topic now appears in kidney and hormone medicine literature.
The reason it isn't yet a routine part of doctor visits is structural, not scientific. Research moves slowly into clinical practice. Routine testing was not designed to detect the kind of cellular changes the research describes.
The information has been there the whole time. It just hasn't been translated.
References
- Peters KE, Chubb SAP, Davis WA, Davis TME. The relationship between hypomagnesemia, metformin therapy and cardiovascular disease complicating type 2 diabetes: the Fremantle Diabetes Study. PLOS ONE. 2013;8(9):e74355. https://pmc.ncbi.nlm.nih.gov/articles/PMC3760872/
- Bouras H, Roig SR, Kurstjens S, Tack CJJ, Kebieche M, de Baaij JHF, Hoenderop JGJ. Metformin regulates TRPM6, a potential explanation for magnesium imbalance in type 2 diabetes patients. Canadian Journal of Physiology and Pharmacology. 2020 Jun;98(6):400-411. https://pubmed.ncbi.nlm.nih.gov/32017603/
- Kurstjens S, Bouras H, Overmars-Bos C, Kebieche M, Bindels RJM, Hoenderop JGJ, de Baaij JHF. Diabetes-induced hypomagnesemia is not modulated by metformin treatment in mice. Scientific Reports. 2019 Feb 11;9(1):1770. https://www.nature.com/articles/s41598-018-38351-3
- Xu EJ, Steele DJR, Fenves AZ. Hypomagnesemia With Metformin Use in Diabetes Mellitus: A Case and Narrative Review. Kidney Medicine. 2025 May 16;7(7):101030. https://pmc.ncbi.nlm.nih.gov/articles/PMC12221734/
- Elin RJ. Assessment of magnesium status for diagnosis and therapy. Magnesium Research. 2010;23(4):S194–S198.
- DiNicolantonio JJ, O'Keefe JH, Wilson W. Subclinical magnesium deficiency: a principal driver of cardiovascular disease and a public health crisis. Open Heart. 2018;5(1):e000668. https://openheart.bmj.com/content/5/1/e000668
- de Baaij JHF, Hoenderop JGJ, Bindels RJM. Magnesium in Man: Implications for Health and Disease. Physiological Reviews. 2015;95(1):1–46.
- Boyle NB, Lawton C, Dye L. The Effects of Magnesium Supplementation on Subjective Anxiety and Stress—A Systematic Review. Nutrients. 2017;9(5):429. https://pmc.ncbi.nlm.nih.gov/articles/PMC5452159/
- Rosanoff A, Weaver CM, Rude RK. Suboptimal magnesium status in the United States: are the health consequences underestimated? Nutrition Reviews. 2012;70(3):153–164.
- Firoz M, Graber M. Bioavailability of US commercial magnesium preparations. Magnesium Research. 2001;14(4):257–262.
- Walker AF, Marakis G, Christie S, Byng M. Mg citrate found more bioavailable than other Mg preparations in a randomised, double-blind study. Magnesium Research. 2003;16(3):183–191.