Nutritional Consequences of Long-term Acid Suppression; Are They Clinically Important?

David A. Johnson


Curr Opin Gastroenterol. 2016;32(2):136-140. 

In This Article


In 2011, the FDA issued a warning statement, based on a series of case reports, concerning the potential for low magnesium levels based among long-term PPI users ( That report noted cases of profound hypomagnesemia and suggested that in 25% of these cases, there was no response to magnesium repletion until the PPI was discontinued. Although the current mechanism for this increased risk is unknown, it is speculated that altered absorption occurs with long-term PPI use. Magnesium uptake within the intestine, however, is independent of a pH effect, but rather mediated by both active and passive reabsorption. Active transport is regulated by the transient receptor potential melastatin 6 and 7 channels, whereas passive reabsorption is concentration dependent and regulated by tight junction complexes.[14,15]

Conceptually, PPIs may alter the luminal pH, resulting in a reduced extrusion of protons into the intestine and leading to decreased activity of the transient receptor potential melastatin 6 and 7 channels.[16] The vast majority of magnesium absorption is via passive absorption so any alteration conceptually should have a de minimis effect.

Although the earlier reports supported the association of hypomagnesemia and PPIs,[17] there were significant confounding factors with the alleged association. According to the more recent studies, there have been a number of studies closely reviewing the risk of hypomagnesemia with PPI use across a number of settings. To look at the frequency of hypomagnesemia in low-risk patients, a case-control study (154 patients, 84 controls) was performed in outpatients at a gastroenterology clinic who were on PPI therapy for at least 6 months and without chronic kidney disease or diuretic use.[18] Results of this study showed that magnesium levels did not significantly differ between PPI users and nonusers (2.17 ± 0.20 mg/dl and 2.19 ± 0.15, respectively). Similarly, even among high-risk patients, in a cross-sectional study that was performed on 512 consecutive renal transplant recipients (20% of whom were on a PPI), by multivariable analysis, PPI use was not an independent predictor of hypomagnesemia.[19] The most recent report of possible related hypomagnesemia evaluated 590 participants who received long-term PPIs and met exclusion criteria for other magnesium loss, total exposure of 2293 PPI-years (average 5.7 years/patient), concluded that in the absence of known precipitating factors, chronic PPI use does not therefore appear to be associated with hypomagnesemia.[20] A recent systematic review and meta-analysis (nine studies, 115 455 patients) also supports this association (OR 1.775, 95% CI 1.077–2.924); however, they also noted that no definitive conclusion could be reached because of significant heterogeneity between studies (i2 = 98%, P < 0.001).[21]

Evidence Summary

There is considerable question as to how an alteration in pH would affect the dominant passive pathway for magnesium absorption. Outside of a possible idiosyncratic effect on a rare number of patients, clinicians should be aware for this potential interaction but not routinely monitor for this. In particular, this association appears to be most pronounced in patients on PPIs and concurrent diuretic therapy.