Opposing Drug Effects on Magnesium and Micronutrient Status in Type 2 Diabetes and their Relation to Cardiovascular Risk: Proton Pump Inhibitors, SGLT2 Inhibitors, Diuretics and Metformin: A scoping review

People with type 2 diabetes (T2DM) are exposed to an unusually mineral-active pharmacopoeia. Proton pump inhibitors (PPIs), sodium-glucose cotransporter 2 (SGLT2) inhibitors, diuretics and metformin all act on mineral or micronutrient handling, and three of the four act on magnesium. Each has been studied as a single exposure, so the net effect of the combination an individual patient experiences has not been characterised. This review maps evidence across all four classes and uses an exploratory scoring framework to characterise the human evidence supporting the proposed PPI-mineral-tissue pathway. Scoping review reported per PRISMA-ScR; five databases were searched to 31 March 2026 and 65 sources of evidence were charted. No meta-analysis was performed and no pooled estimate was generated, because the literature lacks a common effect measure. Each link in the proposed pathway was instead scored against a pre-specified, author-developed heuristic requiring separate human evidence at the exposure-biomarker, biomarker-tissue and tissue-outcome steps. That heuristic is exploratory and is not a validated evidence-grading system. The four classes act on the same targets in opposing directions. SGLT2 inhibitors raise serum magnesium by approximately 0.07 mmol/L (95% CI 0.06–0.08) as a class effect and have been used therapeutically in refractory hypomagnesaemia; loop and thiazide diuretics deplete it; PPI-associated hypomagnesaemia is heterogeneous (pooled RR 1.44, 95% CI 1.13–1.76) and concentrated in diuretic co-exposed patients, in whom the interaction is formally significant. Because hypomagnesaemia both follows from and worsens insulin resistance, these exposures act on a loop rather than a static deficit. PPIs also raise metformin exposure by 15–16% through organic cation transporter inhibition, and PPIs and metformin independently reduce vitamin B12 absorption. Add-on PPI therapy lowers glycated haemoglobin by 0.36%, while selenium supplementation increased incident T2DM. No mineral scores at the highest level across all three steps for a PPI-attributable cardiac effect, and an original exploratory triangulation analysis indicates that, under several unverified assumptions including log-linearity and transportability between genetic and pharmacological effects, the magnesium pathway could account for approximately 2–10% of the observed cardiovascular association. Drug-induced mineral disturbance in T2DM cannot be attributed to any single agent, because three commonly co-prescribed classes act on magnesium in opposing directions in the same patients. One restricted claim survives appraisal: in diuretic-treated patients, PPI co-exposure may lower magnesium enough to matter, and SGLT2 inhibitor co-exposure may offset it. Both halves of that claim are testable in cohorts that already exist. On present estimates the magnesium pathway nonetheless appears biologically real but quantitatively small, and unlikely on its own to account for the reported cardiovascular associations. Supplementation is not supported and, for selenium, is contraindicated.