Proton Pump Inhibitors and Severe Hypomagnesaemia

Tim Cundy; Jonathan Mackay


Curr Opin Gastroenterol. 2011;27(2):180-185. 

In This Article

Possible Mechanisms

In short-term studies (of just a few days), the intestinal absorption of magnesium is not apparently affected by PPI,[18,19] but the observation that renal handling of magnesium is normal suggests that the problem lies with the gut. This makes PPI-induced hypomagnesaemia distinctive from other drug-associated causes, which are characterized by increased urinary losses. We do not yet know whether the problem is due to impaired intestinal transport of magnesium (active or passive) or increased losses into the gut – so what follows is necessarily speculative.

High-dose oral supplementation with magnesium salts (typically providing 30–50 mmol per day) can raise plasma magnesium sufficiently to relieve symptoms while patients still take PPI, although they do not restore normal concentrations. One interpretation of this observation has been that there is impairment of the active transport mechanism, but the problem can be overcome by loading the passive system. Two possibilities have been raised. First, that the function of the TRM6/7 channels might be impaired by the lack of protons[14,20] or that the affected patients are heterozygous for mutations in the TRPM6 gene.[10,12,14,16,17•] However, this hypothesis remains unproven. In one patient, the gene sequence for SLC12A3 (which codes for one of the proteins involved in renal magnesium transport) was normal,[14] but TRPM6 sequence analysis has not been reported.

Another possibility is that the passive paracellular transport mechanism is much less efficient. About 7% of ingested magnesium is normally absorbed by this process, but if this proportion were reduced significantly (or there was substantial outward leakage) then a large increment in magnesium intake would be needed to maintain net absorption. Can PPIs affect paracellular transport of magnesium? This has not been established, but Mullin et al. have demonstrated that esomeprazole can increase the inward intestinal permeability to sucrose, suggesting that the function of the tight junction barrier that controls paracellular transport can be affected by PPIs.[21] Another suggestion is that acidity from the stomach helps keep mineral ions in solution, facilitating absorption, so that PPI-induced hypochlorhydria could impair the availability of ionic magnesium.[9,22] It is not known whether patients with PPI-induced hypomagnesaemia have unusually pronounced achlorhydria.

A further possibility is that PPIs increase faecal magnesium loss. This could be from the small intestine or from the colon. The rapid recurrence of hypomagnesaemia after substantial intravenous loading[9,10,14,17•] suggests the possibility of intestinal leakage. Diarrhoea can be a side-effect of PPIs, particularly in the elderly, and there is a possible link between PPI use and microscopic colitis,[23] but the normal colonic biopsies in PPI-associated hypomagnesaemia make the latter unlikely.

Omeprazole undergoes a chiral shift in vivo, which converts the inactive R-enantiomer to the active S-enantiomer, doubling the concentration of the active form. This chiral shift is accomplished by the CYP2C19 isoenzyme of cytochrome P450. Different genotypes can cause variable rates of metabolism, but Hoorn et al.[15] found no single CYP2C19 genotype in patients with hypomagnesaemia. PPI-induced hypomagnesemia is seen with esomeprazole, which does not require a chiral shift, so differing rates of drug metabolism cannot explain the phenomenon.