Proton pump inhibitors and risk of vitamin and mineral deficiency: evidence and clinical implications

Vitamin B₁₂

Vitamin B₁₂ (cobalamin) is an essential water-soluble nutrient acquired from animal-derived food sources including meats, fish, shellfish, poultry, eggs and dairy products. Lacto-ovo vegetarians are generally not considered to be at risk for deficiency, while true vegans may risk deficiency unless they consume supplements or vitamin B₁₂ fortified foods including cereals and soy-based products. Since the common human diet contains substantially more vitamin B₁₂ than is required, a large functional reserve with respect to vitamin B₁₂ absorption is assumed [Howden, 2000].

Vitamin B₁₂ absorption involves peptic enzymes to cleave dietary B₁₂ from dietary proteins. This is performed primarily by pepsin, which requires gastric acid to activate it from its pepsinogen precursor. Without gastric acid, vitamin B₁₂ would not be cleaved from dietary protein and would not be able to bind to R-proteins, which in turn protect vitamin B₁₂ from pancreatic digestion [Festen, 1991]. It has been hypothesized that since gastric acidity is required for vitamin B₁₂ absorption, acid suppression may lead to malabsorption and ultimately vitamin B₁₂ deficiency from atrophic gastritis and achlorhydria [Howden, 2000].

It has been estimated that vitamin B₁₂ deficiency affects up to 20% of the elderly, and has been linked to impaired gastrointestinal absorption syndromes and pernicious anemia [Andres et al. 2004]. Most clinical cases are undetected and are found incidentally, while more profound cases may present with neuropsychiatric and hematologic findings (e.g. macrocytic anemia) that may be a harbinger of more severe underlying disease. The reduction in upper small intestine gastric acid via PPI therapy may promote bacterial overgrowth allowing for increased bacterial consumption of cobalamin, but the clinical correlation of adverse effects on nutritional status has never been determined.

Studies to date that have evaluated the relationship between PPI use and vitamin B₁₂ deficiency are predominantly case reports and cross-sectional observational trials. The results from these trials have not yielded consistent data to create either therapeutic guidelines or to offer recommendations for routine dietary supplementation. In a systematic review by Sheen and Triadafilopoulos, the authors summarize experimental evidence suggesting that PPIs reduce the absorption of protein-bound vitamin B₁₂ while not completely inhibiting the process, yet PPI therapy does not statistically affect the absorption of unbound and bound vitamin B₁₂ [Sheen and Triadafilopoulos, 2011; Koop, 1992; Schenk et al. 1996].

The initial concern regarding the risk of vitamin B₁₂ deficiency due to long-term AST use is derived from a prospective trial in patients with Zollinger–Ellison syndrome (ZES). A trial of 131 patients treated with either omeprazole (n = 111; mean 4.5 years) or histamine H2 receptor antagonist (H2RAs; n = 20; mean 10 years), determined that serum vitamin B₁₂ levels were significantly lower in patients treated with omeprazole (p = 0.03), especially in subjects with omeprazole-induced sustained hyposecretion (p = 0.0014) or complete achlorhydria (p < 0.0001), while serum folate levels were not statistically affected [Tremanini et al. 1998]. In 68 patients with two vitamin B₁₂ assays obtained 5 years apart, serum levels were decreased by 30% in patients who were achlorhydric (p = 0.001). The duration of omeprazole treatment was found to be inversely correlated with serum vitamin B₁₂ levels (p = 0.013), but not with serum folate levels.

Trials from early last decade have demonstrated a more likely association with PPIs and vitamin B₁₂ deficiency. A case-control study in Medicaid patients evaluated 125 patients with vitamin B₁₂ deficiency compared with 500 control subjects [Force et al. 2003]. More than 18% of patients who received vitamin B₁₂ supplementation had been given AST with either PPIs or H2RAs over a 12-month period compared with 11% in the control group (odds ratio [OR] = 1.82; p = 0.025). Another case-control study in a US university-based geriatric primary care setting identified 53 patients with cobalamin deficiency and compared them with 212 controls with respect to past or current use of prescription AST [Valuck and Ruscin, 2004]. The relative risk of vitamin B₁₂ deficiency associated with AST for <12 months was 1.03 (95% confidence interval [CI] = 0.46–2.31), while that for ≥12 months was 4.46 (95% CI = 1.49–13.3). With an estimated baseline 5% risk of vitamin B₁₂ in the elderly in this trial, the number needed to harm (NNH) is 7; with an assumed baseline risk of 10% deficiency, the NNH is 4.

Ito and Jensen recently summarized the results of more recent trials examining vitamin B₁₂ deficiency risk due to PPI therapy in a systematic review [Ito and Jensen, 2010]. One case-control study involving 125 long-term (greater than 3 years) PPI users from general practices who were aged 65 years or older, compared with the same number of controls [den Elzen et al. 2008]. Vitamin B₁₂ deficiency status was determined by measurement of serum vitamin B₁₂ and homocysteine levels, as well as red blood cell mean corpuscular volume. This trial found no significant association between long-term PPI use and vitamin B₁₂ status between the two groups. A longitudinal study of 61 acid hypersecretors, 46 of whom had ZES and all of whom were taking long-term PPI therapy, were followed for up to 18 years to determine risk of vitamin B₁₂ malabsorption [Hirschowitz et al. 2008]. Of 61 patients, 10% had low serum vitamin B₁₂ levels, while further investigation discovered vitamin B₁₂ deficiency 31% patients despite normal serum vitamin B₁₂ levels. Lastly, a cross-sectional study evaluated the effects of chronic use of H2RAs (150 patients), PPIs (141 patients) or no AST (251 patients) in patients in nursing homes or community ambulatory care facilities aged 60 to 102 years over a period of 6 or more years [Dharmarajan et al. 2008]. Decreased serum vitamin B₁₂ levels were discovered in 20% of the nursing home patients and 29% of the community care patients at baseline. While PPI use was correlated with lower serum vitamin B₁₂ levels, H2RA use was not.

A recent study by Rozgony and colleagues prospectively examined trends in serum vitamin B₁₂ levels in 36 long-term care residents aged 60 to 89 years, 17 of whom were on chronic PPI therapy [Rozgony et al. 2010]. The authors found a statistically significant difference in mean serum vitamin B₁₂ and methylmalonic acid levels and frequency of vitamin B₁₂ deficiency at baseline between long-term PPI users compared with nonusers (75% versus 11%, p = 0.006). The PPI users were subsequently given 8 weeks of intranasal vitamin B₁₂ nasal spray dosed at 500 μg weekly. Follow-up evaluation concluded that serum vitamin B₁₂ levels then increased (p = 0.012), while the frequency of vitamin B₁₂ deficiency decreased (p = 0.004).

While some data suggest an increased risk of vitamin B₁₂ deficiency with PPI use, most current evidence is based upon small, poorly controlled and nonrandomized studies and case reports. Assumption of increased risk of vitamin B₁₂ deficiency in the elderly and malnourished may be likely to some degree, but there is no evidence to suggest additional supplementation or monitoring of serum cobalamin levels. Prospective trials, with patient recruitment at the commencement of PPI therapy, are needed to prove a direct cause-and-effect relationship of vitamin B₁₂ deficiency.

Vitamin C

A recent review by McColl highlights a detailed understanding of the effect of PPIs on vitamin C absorption [McColl, 2009]. Humans are unable to synthesize vitamin C and thus must rely upon obtaining adequate concentrations of the water-soluble vitamin from dietary intake. PPIs affect its bioavailability via lowering its concentration in gastric juices as well as the proportion of vitamin C in its active antioxidant form, ascorbic acid. The ascorbic acid that is in turn secreted by the gastric mucosa directly affects the concentration of nitrite and iron in gastric juices.

Mowat and colleagues investigated the effects of omeprazole 40 mg for 4 weeks on vitamin C concentrations in gastric juice in healthy individuals [Mowat et al. 1999]. Median intragastric pH increased from 1.4 prior to omeprazole therapy to 7.2 (range 3.5–8.5) while subjects were taking omeprazole, and vitamin C concentrations decreased from 5 μm/l pretreatment with omeprazole to 3 μm/l while subjects were taking omeprazole, reflecting a notable decrease in the biologically active form of ascorbic acid. The clinical significance relative to the decreases in vitamin C is derived from a theory that it may protect against conversion of nitrite to N-nitroso compounds by bacteria which could potentially colonize the achlorhydric stomach environment.

There is additional evidence that PPIs lower serum vitamin C concentrations, an effect more prominently noted in patients who were infected with Helicobacter pylori [Mowat et al. 1999; Henry et al. 2005]. Authors of these trials have postulated that the gastritis in the body of the stomach secondary to H. pylori infection taking a PPI (omeprazole in these studies) may increase vitamin C metabolism leading to systemic depletion. The clinical significance of this effect remains unknown, and vitamin C supplementation is not currently recommended in patients taking PPIs or in those being treated for H. pylori infection.

Calcium

A review by Insogna provides the framework upon which to understand the relationship between PPI therapy and calcium metabolism [Insogna, 2009]. It has been proposed that both stomach acid and the slightly acidic milieu of the proximal duodenum are necessary in order to dissociate ingested calcium from a food bolus rendering it available for absorption. Without this environment, elemental calcium would not be absorbed, potentially leading to compensatory physiologic changes including secondary hyperparathyroidism. While there is conflicting evidence with regard to the role of intragastric hydrochloric acid in calcium absorption, and PPIs are known to inhibit this mechanism, one study found that gastric acid secretion and gastric acidity do not normally play a role in the absorption of dietary calcium [Bo-Linn et al. 1984]. PPIs reduce resorption of calcium from bone, as osteoclasts also possess proton pumps, thus their activity is thought to be directly affected by PPIs [Farina and Gagliardi, 2002].

Many studies examining the impact of PPIs on intestinal calcium absorption are limited by the evaluation of patients with renal failure on hemodialysis or those with hypo- or achlorhydria, two chronic conditions known to adversely affect calcium metabolism [Insogna, 2009]. Several short-term trials with small numbers of subjects evaluating the effects of PPIs on intestinal calcium absorption found no significant increase in serum calcium during PPI therapy, consistent with decreased calcium absorption [Hardy et al. 1998; Graziani et al. 2002; O’Connell et al. 2005]. One particular study, in postmenopausal women who received calcium supplementation taken on an empty stomach, did not demonstrate decreased calcium resorption with concomitant PPI utilization [O’Connell et al. 2005].

To date, no long-term prospective, randomized, controlled trials exist to examine the potential increased risk of bone fracture associated with concomitant PPI use, as all published data is derived from retrospective case-controlled, cohort and cross-sectional studies. A series of trials implicating PPI therapy as leading to an increased risk of osteoporotic fracture has stimulated a critical analysis of the possible relationship, suggesting a causal relationship [Yang et al. 2006; Vestergaard et al. 2006; Targownik et al. 2008; Kaye and Jick, 2008; Gray et al. 2010]. Newer research offers conflicting data, implying that study subjects who had been on PPI therapy may be at a higher risk of osteoporotic fracture compared with an individual at average risk, yet the authors concluded that PPI use was not associated with a change in bone mineral density [Targownik et al. 2010].

In 2010, the US Food and Drug Administration released a warning revising the prescription and over-the-counter (OTC) labels for PPIs to include new safety information regarding a potential increased risk of fractures of the hip, wrist and spine with the use of these medications [FDA, 2010]. This warning was updated in 2011, as the FDA determined that an osteoporosis and fracture warning on the OTC PPI ‘Drug Facts’ label is not indicated, concluding that fracture risk with short-term, low-dose PPI therapy is unlikely [FDA, 2011]. Patients felt to be at highest risk for fractures had either received high doses of prescription PPIs (higher than OTC PPI doses) or had used a PPI for 1 year or more, or both.

Iron

It has been postulated that chronic PPI therapy results in clinically significant iron malabsorption due to gastric acid hyposecretion and the risk of achlorhydria. Dietary iron is present in food as either nonheme (66%) or heme iron (32%) and absorption of nonheme iron is markedly improved by gastric acid. Gastric acid assists food sources containing nonheme iron to dissociate and to solubilize the iron salts, allowing formation of complexes with sugars and amines facilitating absorption [Ito and Jensen, 2010].

While many patients with a history of vagotomy, gastric resection or atrophic gastritis have been shown to have iron deficiency anemia, a cohort of patients with ZES who received treatment with PPIs for over 10 years did not demonstrate clinically significant iron deficiency [Stewart et al. 1998]. Several case reports have suggested an association between PPI use and iron deficiency anemia, with subjects not responding to iron replacement while on concomitant PPI therapy but responding favorably when PPI therapy was stopped [Sharma et al. 2004; Hutchinson et al. 2007].

A retrospective cohort study of adult patients in an academic outpatient setting who received PPI therapy for at least 1 year sought to evaluate any potential change in hematologic indices against matched controls [Sarzynski et al. 2011]. Of the 98 patients who met inclusion criteria, PPI users were observed to have significant decreases in mean hemoglobin and hematocrit concentrations (p < 0.01 for both parameters) compared with matched controls. After adjustment for confounders including invasive evaluation, the odds ratio of decreasing hemoglobin by 1.0 g/dl while on chronic PPI therapy was 5.03 (95% CI = 1.71–14.78, p < 0.01), while the odds ratio of decreasing hematocrit by 3% was 5.46 (95% CI = 1.67–17.85, p < 0.01). Of the PPI users in this trial, 35% had no documented appropriate indication for such therapy, similar to results in previous ambulatory care setting trials [Heidelbaugh et al. 2010]. Despite these results, there is no current recommendation to monitor patients on chronic PPI therapy for iron deficiency anemia.

Magnesium

Hypomagnesemia secondary to chronic PPI therapy is now a well-documented yet still rare phenomenon, as to date there is no widely accepted mechanism to explain such an association. One researcher posited that it may occur in cases of ‘poor metabolizers’ of PPIs, but this has been disproven [Hoorn et al. 2010]. Hypomagnesemia has been documented with all PPIs that are biochemically substituted pyridylmethylsulphonyl benzimidazadole derivatives (in order of potency: rabeprazole, esomeprazole, omeprazole, lansoprazole and pantoprazole). While several trials have documented recurrence of hypomagnesemia when one PPI is substituted for another, it considered as a class effect [Cundy and Mackay, 2011].

Fewer than 30 cases of hypomagnesemia have been described in the literature associated with PPI therapy since 2006, with 61% having received PPI therapy for 5 or more years and 29% for at least 10 years [Sheen and Triadafilopoulos, 2011]. The median age at diagnosis was 70 (range 51–82) and deficiency was reported more frequently in women than men. Most patients with suspected PPI-induced hypomagnesemia presented with concomitant hypokalemia and hypocalcemia, as well as severe ataxia, paresthesias, seizures, confusion and gastrointestinal symptoms requiring hospitalization. In these patients, there was no evidence of magnesium malabsorption, renal wasting or histological abnormalities on small intestinal or colonic biopsies. The most severe symptoms have been associated with the lowest plasma magnesium and calcium concentrations [Cundy and Mackay, 2011].

In 2011, the FDA released a warning based upon several published case reports stating that PPIs may cause hypomagnesemia if taken for longer than a year [FDA, 2011]. In approximately 25% of cases reviewed, magnesium supplementation alone did not adequately increase serum magnesium levels and thus PPI therapy had to be discontinued. It is considered to be reasonable practice to screen patients with a history of cardiac arrhythmias or those on antiarrhythmic agents for low serum magnesium if they are on chronic PPI therapy.

Gau and colleagues conducted a cross-sectional study to determine the potential effects of PPI therapy relative to development of hypomagnesemia [Gau et al. 2012]. Data from 207 hospitalized adults on PPI therapy aged over 50 years who had measurement of serum magnesium revealed that both standard and high-dose PPI use was associated with an increased risk of hypomagnesemia (adjusted OR = 2.50, 95% CI = 1.43–4.36), suggesting a subclinical insufficiency or deficiency status in asymptomatic individuals.