Abstract
Background:
Esophageal adenocarcinoma has been inversely associated with exposure to ultraviolet radiation. This could be due to vitamin D deficiency or hyperparathyroidism promoting gastroesophageal reflux disease (GERD) and Barrett’s esophagus.
Aim:
To determine the association between parathyroid hormone (PTH) and vitamin D deficiency with GERD symptoms, erosive esophagitis, and Barrett’s esophagus.
Methods:
We assayed banked serum for PTH and total 25-hydroxy vitamin D from a cross-sectional cohort. Logistic regression was performed to estimate the associations of vitamin D deficiency and hyperparathyroidism with GERD symptoms, erosive esophagitis, and Barrett’s esophagus.
Results:
Sera from 605 men were assayed, including 150 with GERD, 216 with erosive esophagitis, 145 with Barrett’s esophagus, and 174 normal subjects. Contrary to our hypothesis, we found a strong inverse association between Barrett’s esophagus and hyperparathyroidism (OR = 0.516; 95% CI = 0.265, 1.01), and a trend toward an inverse association with vitamin D deficiency. We found no association between vitamin D deficiency or hyperparathyroidism with GERD symptoms or erosive esophagitis.
Conclusions:
Contrary to our hypothesis, we found an inverse association between serum PTH and Barrett’s esophagus. Validation of the finding, and the mechanism of that association deserves further study.
Keywords: Barrett’s esophagus, gastroesophageal reflux, parathyroid hormone, vitamin D
Introduction
The etiology of esophageal adenocarcinoma is incompletely understood. A major risk factor is gastroesophageal reflux disease (GERD), but only a minority of patients with GERD develop Barrett’s esophagus or esophageal adenocarcinoma. Esophageal adenocarcinoma has been associated with decreased exposure to ultraviolet radiation, suggesting that vitamin D deficiency may be responsible.(1) Vitamin D deficiency is associated with idiopathic, but not diabetic gastroparesis, suggesting that vitamin D deficiency may be responsible for the dysmotility in idiopathic cases.(2) So vitamin D deficiency might promote GERD either through subclinical gastroparesis or esophageal dysmotility. In addition, vitamin D deficiency can lead to secondary hyperparathyroidism, and hyperparathyroidism has been shown to be associated with altered gene expression in adipose tissue which promotes an inflammatory response.(3) Alterations in circulating adipokine levels have been associated with erosive esophagitis and Barrett’s esophagus.(4, 5) Therefore, we aimed to compare vitamin D and parathyroid hormone (PTH) status between patients with GERD symptoms, erosive esophagitis, or Barrett’s esophagus or none of those conditions.
Materials and Methods
We conducted a series of case-control analyses within a previously developed cross-sectional cohort, The Newly Diagnosed Barrett’s Esophagus Study.(4, 6–8) From 2008 through 2012, we recruited men who had reported to either the University of Michigan Health System (UMHS) or to the Ann Arbor Veterans Affairs Health System (AAVAHS) for a colonoscopy for the purposes of colorectal cancer screening. Regardless of symptoms, patients were recruited to undergo a research upper endoscopy at the time of their colonoscopy. As part of the same study, we additionally recruited men shortly after their first diagnosis of Barrett’s esophagus. Patients were excluded from eligibility in the study if they were female; had a prior upper endoscopy (for the colorectal cancer screenees); were undergoing colonoscopy for diagnostic indications; had prevalent cancer, ascites, esophageal varices or coagulopathy; had a history of esophagectomy; were inpatient at the time of recruitment; were < 50 years of age or > 79; or were unable to comprehend and cooperate with the study. Under Institutional Review Board approval at both UMHS and AAVAHS, subjects provided informed consent to bank their specimens for future studies such as this one.
Subjects completed multiple questionnaires including data on GERD symptom frequency, use of PPIs and H2RAs, physical activity using the Morgenstern Physical Activity Questionnaire,(9) dietary and supplementary intake of vitamin D and calcium using the Block Brief 2000 Food Frequency Questionnaire,(10) tobacco use and race. Current medications were also abstracted from the patient’s electronic medical record and confirmed with the subjects. There was 99.97% concurrence between this method and the questionnaire for current use of PPIs. In addition, subjects were measured for height, weight, waist circumference, and hip circumference.
Endoscopists were directed by research coordinators to abide by standard guidelines for documenting Barrett’s esophagus (using the Prague criteria) and erosive esophagitis (using the Los Angeles Classification). Digital photographs were obtained in each patient, and the presence of Barrett’s esophagus was confirmed by the principal investigator (JHR).
Current PPI use was defined as use of at least 3 doses per week at the time of enrollment. Use of histamine receptor type-2 antagonists (H2RA) was defined similarly. GERD was defined as symptoms of heartburn or regurgitation occurring at least weekly either in a patient who does not use acid-reducing medications, or the patient reported to typically have those symptoms at least weekly before he began using such medications. Erosive esophagitis was defined as at least L.A. Classification A in severity. Barrett’s esophagus was defined as endoscopically visible columnar mucosa extending from the top of gastric folds cephalad with histologically confirmed specialized intestinal metaplasia. Normal controls were defined as the absence of all 3 of the above conditions.
Blood samples were drawn from each subject, and aliquots of serum were stored at −80 degrees centigrade. Serum samples were selected for assay in all available subjects with Barrett’s esophagus, erosive esophagitis, or at least weekly GERD symptoms (categories not mutually exclusive). In addition, samples were assayed from 174 randomly selected subjects without symptomatic GERD (or less than weekly), erosive esophagitis, or Barrett’s esophagus. Serum samples were assayed in duplicate by ELISAs for 25-hydroxy vitamin D, and for PTH (American Laboratory Products, Salem, NH). Intra-assay and inter-assay coefficients of variation were ≤ 9% for vitamin D and PTH. Because we found an unexpected inverse association between Barrett’s esophagus and PTH, we also assayed serum in duplicate for PTH related peptide (PTHrP) using an ELISA (LifeSpan Biosciences, Seattle, WA); intra-assay and inter-assay coefficients of variation were < 5%. Vitamin D and PTH were dichotomized at their normal values (> 20 ng/mL and < 65 pg/mL, respectively), and PTHrP was categorized into tertiles based on the observed concentrations in the 300 randomly selected colorectal cancer screenees including GERD symptoms, erosive esophagitis, and Barrett’s esophagus found in the proportions as the overall cohort of 822 men.
We fitted logistic regression models, adjusting for potential confounders, to estimate the association of each analyte with 1) Barrett’s esophagus compared to no Barrett’s esophagus, 2) erosive esophagitis compared to no erosive esophagitis or Barrett’s esophagus (since it is believed that those patients once had erosive esophagitis prior to development of Barrett’s esophagus), and 3) GERD symptoms compared to no GERD symptoms, erosive esophagitis, or Barrett’s esophagus (since patients with erosive esophagitis and Barrett’s esophagus, even without symptoms of GERD, are believed to all have pathologic reflux). Because PPI use has been associated with fragility bone fractures, (11, 12) PPI use might confound the relationship between vitamin D deficiency or hyperparathyroidism with GERD and its sequella. Therefore, we also fitted similar models to estimate the association of each analyte with PPI use.
Serum vitamin D concentration varies with seasonal exposure to ultraviolet radiation. We addressed this by fitting a regression model of vitamin D among the 300 randomly selected colorectal cancer screenees using sinusoidal functions of the day of the year the blood was drawn. Then for each subject, we calculated the residual difference between predicted and observed vitamin D, categorized the residual into tertiles, and used that residual instead of observed vitamin D in the main logistic regression models described above. We employed a similar strategy for PTH since we also observed seasonal variation in PTH concentration. No such seasonal variation was observed in PTHrP concentration.
Results
A total of 1,308 colorectal cancer screenees were recruited, of which 1,202 were eligible; 851 were enrolled, and 822 completed the upper endoscopy. 155 subjects (19%) had GERD symptoms on an at least a weekly basis, of whom 150 had serum available for assay. 222 had erosive esophagitis (27%), of whom 216 had serum available for assay. Barrett’s esophagus confirmed by histology was found in 70 subjects (9%). In addition, 80 men were enrolled who had been recently diagnosed with Barrett’s esophagus for the first time, providing a total of 150 subjects with Barrett’s esophagus, of whom 145 had serum available for assay. In addition, 174 subjects without Barrett’s esophagus, erosive esophagitis, or GERD symptoms were randomly selected for assay. In total, the sera from 605 subjects were assayed (categories of Barrett’s esophagus, erosive esophagitis, and symptomatic GERD were not mutually exclusive). 152 subjects were current users of PPI. Substantial proportions of subjects with GERD symptoms were not using a PPI (41%). Similarly, 87% of subjects with erosive esophagitis, and 87% of subjects with Barrett’s esophagus were not using a PPI. Vitamin D deficiency was identified in 133 subjects (25%), and hyperparathyroidism in 140 subjects (23%).
As expected, vitamin D deficiency was associated with hyperparathyroidism (odds ratio [OR] = 1.76; 95% confidence interval [CI] = 1.15, 2.70). Dietary intake of vitamin D trended toward a non-significant inverse association with vitamin D deficiency (3rd tertile vs. 1st tertile OR = 0.737; 95% CI = 0.432, 1.24), but use of supplemental vitamin D was strongly inversely associated with vitamin D deficiency (OR = 0.202; 95% CI = 0.124, 0.33). Supplemental calcium use (≥ 500 mg/day) was strongly inversely associated with hyperparathyroidism compared to no supplemental calcium (OR = 0.291; 95% CI = 0.111, 0.765). Supplemental vitamin D use was also inversely associated with hyperparathyroidism (OR = 0.527; 95% CI = 0.344, 0.808).
Contrary to our hypothesis, there was a non-significant trend toward an inverse association between Barrett’s esophagus and vitamin D deficiency (fully adjusted OR = 0.555; 95% CI = 0.269, 1.15) (Table 1). Likewise, we found a significant inverse association between Barrett’s esophagus and hyperparathyroidism (OR = 0.588; 95% CI = 0.362, 0.956) (Table 1). The estimate of the effect was slightly stronger, but crossed the null value in the fully adjusted model (OR = 0.516; 95% CI = 0.265, 1.01). Because of this finding, we subsequently assayed serum for PTHrP, hypothesizing that if Barrett’s esophagus secreted PTHrP, it might lead to a negative feedback on serum PTH concentration. However, we found no evidence of an association between Barrett’s esophagus and serum PTHrP (Table 1).
Table 1.
Logistic Regression of Barrett’s Esophagus Compared to No Barrett’s Esophagus
| Predictor | No BE / BE | Unadjusted OR1 | Adjusted OR2 | Adjusted OR3 |
|---|---|---|---|---|
| Normal Vitamin D (≥ 20 ng/mL) Deficient Vitamin D (< 20 ng/mL) |
355 / 117 105 / 28 |
1 (reference) 0.809 (0.508, 1.29) |
1 (reference) 0.739 (0.401, 1.36) |
1 (reference) 0.555 (0.269, 1.15) |
| Normal PTH (≤ 65 pg/mL) Hyperparathyroid (> 65 pg/mL) |
344 / 121 116 / 24 |
1 (reference) 0.588 (0.362, 0.956) |
1 (reference) 0.427 (0.232, 0.786) |
1 (reference) 0.516 (0.265, 1.01) |
| 1st Tertile PTHrP (< 69 pg/mL) 2nd Tertile PTHrP (69 – 110 pg/mL) 3rd Tertile PTHrP (> 110 pg/mL) |
132 / 48 146 / 45 140 / 44 |
1 (reference) 0.848 (0.530, 1.36) 0.864 (0.538, 1.39) |
1 (reference) 0.981 (0.540, 1.78) 0.754 (0.407, 1.40) |
1 (reference) 0.916 (0.474, 1.77) 0.975 (0.490, 1.94) |
Model 1:unadjusted for any potential confounders.
Model 2:adjusted for age, waist-to-hip ratio, smoking status, race, physical activity, and current PPI use.
Model 3:additionally adjusted for dietary calcium, dietary vitamin D, supplemental calcium, and supplemental vitamin D.
We did not find evidence of an association between erosive esophagitis and vitamin D deficiency, hyperparathyroidism, or serum PTHrP concentration (Table 2). Similarly, we did not find any evidence of an association between GERD symptoms and vitamin D deficiency, hyperparathyroidism, or serum PTHrP concentration (Table 3). For all outcomes, we performed sensitivity analyses using a threshold of vitamin D deficiency of 30 ng/mL, finding similar results (data not shown).
Table 2.
Logistic Regression of Erosive Esophagitis Compared to No Erosive Esophagitis and No Barrett’s Esophagus
| Predictor | No EE / EE | Unadjusted OR1 | Adjusted OR2 | Adjusted OR3 |
|---|---|---|---|---|
| Normal Vitamin D (≥ 20 ng/mL) Deficient Vitamin D (< 20 ng/mL) |
202 / 168 65 / 48 |
1 (reference) 0.888 (0.580, 1.36) |
1 (reference) 0.954 (0.584, 1.56) |
1 (reference) 0.761 (0.422, 1.37) |
| Normal PTH (≤ 65 pg/mL) Hyperparathyroid (> 65 pg/mL) |
199 / 164 68 / 52 |
1 (reference) 0.936 (0.614, 1.43) |
1 (reference) 0.989 (0.614, 1.59) |
1 (reference) 0.919 (0.536, 1.58) |
| 1st Tertile PTHrP (< 69 pg/mL) 2nd Tertile PTHrP (69 – 110 pg/mL) 3rd Tertile PTHrP (> 110 pg/mL) |
110 / 32 125 / 28 119 / 25 |
1 (reference) 1.06 (0.667, 1.67) 1.12 (0.704, 1.79) |
1 (reference) 1.09 (0.655, 1.82) 1.08 (0.645, 1.80) |
1 (reference) 1.33 (0.749, 2.37) 1.13 (0.626, 2.03) |
Model 1:Unadjusted for any potential confounders.
Model 2:Adjusted for age, waist-to-hip ratio, smoking status, race, physical activity, and current PPI use.
Model 3:Additionally adjusted for dietary calcium, dietary vitamin D, supplemental calcium, and supplemental vitamin D.
Table 3.
Logistic Regression of At Least Weekly GERD Symptoms Compared to No GERD or < Weekly, No Erosive Esophagitis, and No Barrett’s Esophagus
| Predictor | No GERD / GERD | Unadjusted OR1 | Adjusted OR2 | Adjusted OR3 |
|---|---|---|---|---|
| Normal Vitamin D (≥ 20 ng/mL) Deficient Vitamin D (< 20 ng/mL) |
135 / 107 39 / 43 |
1 (reference) 1.09 (0.677, 1.75) |
1 (reference) 1.05 (0.523, 2.10) |
1 (reference) 0.858 (0.357, 2.06) |
| Normal PTH (≤ 65 pg/mL) Hyperparathyroid (> 65 pg/mL) |
135 / 111 39 / 39 |
1 (reference) 1.22 (0.730, 2.03) |
1 (reference) 1.07 (0.529, 2.17) |
1 (reference) 0.897 (0.399, 2.02) |
| 1st Tertile PTHrP (< 69 pg/mL) 2nd Tertile PTHrP (69 – 110 pg/mL) 3rd Tertile PTHrP (> 110 pg/mL) |
51 / 47 54 / 42 51 / 46 |
1 (reference) 0.844 (0.479, 1.49) 0.979 (0.558, 1.72) |
1 (reference) 0.919 (0.442, 1.91) 1.11 (0.547, 2.42) |
1 (reference) 1.04 (0.446, 2.44) 1.34 (0.587, 3.06) |
Model 1:Unadjusted for any potential confounders.
Model 2:Adjusted for age, waist-to-hip ratio, smoking status, race, physical activity, and current PPI use.
Model 3:Additionally adjusted for dietary calcium, dietary vitamin D, supplemental calcium, and supplemental vitamin D.
Accounting for seasonal variation in vitamin D and PTH, we found that PTH concentration greater than that predicted by calendar date was inversely associated with Barrett’s esophagus (3rd tertile vs. 1st tertile unadjusted OR = 0.618, 95% CI = 0.389, 0.980; fully adjusted OR = 0.705, 95% CI = 0.366, 1.36) (Table 4). We found non-significant trends toward associations of erosive esophagitis and GERD symptoms with observed vitamin D concentrations that were greater than predicted by calendar date (Tables 5 and 6, respectively).
Table 4.
Logistic Regression of Residual Vitamin D and Residual PTH for Barrett’s Esophagus Compared to No Barrett’s Esophagus
| Predictor | No BE / BE | Unadjusted OR1 | Adjusted OR2 | Adjusted OR3 |
|---|---|---|---|---|
| 1st Tertile Residual Vitamin D (< −9.51 ng/mL) 2nd Tertile Residual Vitamin D (−9.51 – 0.45 ng/mL) 3rd Tertile Residual Vitamin D (> 0.45 ng/mL) |
157 / 47 158 / 47 145 / 51 |
1 (reference) 0.846 (0.536, 1.33) 0.851 (0.539, 1.34) |
1 (reference) 1.03 (0.574, 1.84) 1.15 (0.636, 2.06) |
1 (reference) 1.26 (0.650, 2.44) 1.24 (0.633, 2.43) |
| 1st Tertile Residual PTH (< −15.1 pg/mL) 2nd Tertile Residual PTH (−15.1 – 2.5 pg/mL) 3rd Tertile Residual PTH (> 2.5 pg/mL) |
157 / 60 142 / 47 161 / 38 |
1 (reference) 0.866 (0.555, 1.35) 0.618 (0.389, 0.980) |
1 (reference) 1.05 (0.594, 1.84) 0.511 (0.284, 0.920) |
1 (reference) 1.43 (0.752, 2.71) 0.705 (0.366, 1.36) |
Residual vitamin D was calculated by the difference between observed vitamin D and expected vitamin D based on a fitted sinusoidal function of vitamin D across calendar date. Residual PTH was calculated similarly.
Model 1:Unadjusted for any potential confounders.
Model 2:Adjusted for age, waist-to-hip ratio, smoking status, race, physical activity, and current PPI use.
Model 3:Additionally adjusted for dietary calcium, dietary vitamin D, supplemental calcium, and supplemental vitamin D.
Table 5.
Logistic Regression of Residual Vitamin D and Residual PTH for Erosive Esophagitis Compared to No Erosive Esophagitis and No Barrett’s Esophagus
| Predictor | No EE / EE | Unadjusted OR1 | Adjusted OR2 | Adjusted OR3 |
|---|---|---|---|---|
| 1st Tertile Residual Vitamin D (< −9.51 ng/mL) 2nd Tertile Residual Vitamin D (−9.51 – 0.45 ng/mL) 3rd Tertile Residual Vitamin D (> 0.45 ng/mL) |
95 / 71 90 / 75 82 / 70 |
1 (reference) 1.12 (0.722, 1.72) 1.14 (0.733, 1.78) |
1 (reference) 1.24 (0.762, 2.02) 1.54 (0.919, 2.57) |
1 (reference) 1.51 (0.851, 2.69) 1.76 (0.958, 3.22) |
| 1st Tertile Residual PTH (< −15.1 pg/mL) 2nd Tertile Residual PTH (−15.1 – 2.5 pg/mL) 3rd Tertile Residual PTH (> 2.5 pg/mL) |
94 / 71 82 / 69 91 / 76 |
1 (reference) 1.11 (0.714, 1.74) 1.11 (0.717, 1.71) |
1 (reference) 1.02 (0.626, 1.67) 1.14 (0.696, 1.86) |
1 (reference) 1.21 (0.692, 2.12) 1.08 (0.617, 1.87) |
Residual vitamin D was calculated by the difference between observed vitamin D and expected vitamin D based on a fitted sinusoidal function of vitamin D across calendar date. Residual PTH was calculated similarly.
Model 1:Unadjusted for any potential confounders.
Model 2:Adjusted for age, waist-to-hip ratio, smoking status, race, physical activity, and current PPI use.
Model 3:Additionally adjusted for dietary calcium, dietary vitamin D, supplemental calcium, and supplemental vitamin D.
Table 6.
Logistic Regression of Residual Vitamin D and Residual PTH for At Least Weekly GERD Symptoms Compared to No GERD or < Weekly, No Erosive Esophagitis, and No Barrett’s Esophagus
| Predictor | No GERD / GERD | Unadjusted OR1 | Adjusted OR2 | Adjusted OR3 |
|---|---|---|---|---|
| 1st Tertile Residual Vitamin D (< −9.51 ng/mL) 2nd Tertile Residual Vitamin D (−9.51 – 0.45 ng/mL) 3rd Tertile Residual Vitamin D (> 0.45 ng/mL) |
62 / 58 60 / 50 52 / 42 |
1 (reference) 0.891 (0.530, 1.50) 0.863 (0.502, 1.48) |
1 (reference) 0.990 (0.495, 1.98) 1.49 (0.729, 3.04) |
1 (reference) 1.26 (0.549, 2.89) 1.42 (0.587, 3.46) |
| 1st Tertile Residual PTH (< −15.1 pg/mL) 2nd Tertile Residual PTH (−15.1 – 2.5 pg/mL) 3rd Tertile Residual PTH (> 2.5 pg/mL) |
65 / 52 55 / 44 54 / 54 |
1 (reference) 0.962 (0.587, 1.58) 1.15 (0.717, 1.86) |
1 (reference) 1.20 (0.642, 2.24) 0.851 (0.448, 1.62) |
1 (reference) 1.76 (0.853, 3.62) 0.983 (0.464, 2.08) |
Residual vitamin D was calculated by the difference between observed vitamin D and expected vitamin D based on a fitted sinusoidal function of vitamin D across calendar date. Residual PTH was calculated similarly.
Model 1:Unadjusted for any potential confounders.
Model 2:Adjusted for age, waist-to-hip ratio, smoking status, race, physical activity, and current PPI use.
Model 3:Additionally adjusted for dietary calcium, dietary vitamin D, supplemental calcium, and supplemental vitamin D.
There was no association detected between PPI use and vitamin D deficiency or hyperparathyroidism (fully adjusted OR = 0.881; 95% CI = 0.518, 1.50) (Supplemental Tables 1 and 2, Supplemental Digital Content 1, http://links.lww.com/JCG/A452 respectively), including in analyses accounting for seasonal variation in vitamin D or PTH concentrations (Supplemental Table 3, Supplemental Digital Content 1, http://links.lww.com/JCG/A452).
Discussion
We conducted a series of case-control analyses, finding no evidence of vitamin D deficiency or hyperparathyroidism in promoting GERD, erosive esophagitis, or Barrett’s esophagus. In fact, we found some evidence for an inverse association between both vitamin D deficiency and hyperparathyroidism with Barrett’s esophagus.
Esophageal adenocarcinoma is believed to arise in the setting of chronic GERD, in particular among patients with erosive esophagitis that subsequently heal with Barrett’s esophagus. Esophageal adenocarcinoma has been associated with decreased exposure to ultraviolet radiation, suggesting that vitamin D deficiency may be responsible.(1) That study was a population based retrospective case-control study conducted in Australia, in which subjects reported the duration they had lived in each location during their lives. The locations were correlated to atmospheric data on ultraviolet radiation, so lifetime ultraviolet radiation exposure could be calculated. In contrast, a recent systematic review identified only a single study examining the association between serum vitamin D and esophageal adenocarcinoma, suggesting an increase in risk of esophageal adenocarcinoma with greater vitamin D concentration, although with a very imprecise estimate of effect (OR = 1.63, 95% CI = 0.25, 2.12).(13, 14) There were also non-significant associations between dietary intake of vitamin D and Barrett’s esophagus or esophageal adenocarcinoma, and inconsistent associations between particular polymorphisms in the vitamin D receptor and either Barrett’s esophagus or esophageal adenocarcinoma.(14) In our study, we found no evidence for vitamin D deficiency in promoting GERD, erosive esophagitis or Barrett’s esophagus. Rather, we found an inverse trend between vitamin D deficiency and Barrett’s esophagus. Likewise, we found evidence for an inverse association between PTH and Barrett’s esophagus. Ex vivo, PTH augments contractility in opossum esophageal circulatory smooth muscle fibers.(15) So perhaps individuals with lower concentrations of PTH have diminished esophageal peristalsis and clearance of refluxate, or diminished tone of the lower esophageal sphincter. But if so, we would expect that PTH would also be inversely associated with GERD symptoms and erosive esophagitis, neither of which were observed in our study. Our findings suggest that if ultraviolet radiation protects against esophageal adenocarcinoma via vitamin D synthesis, then it likely exerts its effect on the transition from Barrett’s esophagus to cancer. Research using cell cultures or animal studies should be pursued to understand the effects of vitamin D and PTH on the development of Barrett’s esophagus and the progression from Barrett’s esophagus to cancer.
Since PPI use has been associated with fragility bone fractures, we considered that PPI use might confound the associations between vitamin D deficiency or hyperparathyroidism with GERD and its sequellae. Interestingly, we found no evidence of associations between PPI use and hyperparathyroidism or vitamin D deficiency. Chronic use of PPIs has been implicated in increasing the risk of fragility bone fractures in a number of observational studies. But the mechanism of the observed associations is not known. It has been hypothesized that gastric hypochlorhydria induced by PPIs does not allow for dietary calcium to be ionized and hence leads to malabsorption to calcium. There is scant evidence to support that hypothesis; a randomized cross-over trial in 18 elderly women demonstrated that omeprazole 20mg daily interfered with absorption of radiolabelled calcium carbonate (9.1% absorbed while on placebo vs. 3.5% while on omeprazole, p = 0.003).(16) Whether that difference is clinically meaningful has not been clear. In a randomized trial of 13 healthy volunteers, omeprazole 40mg daily did not alter absorption of calcium from a standardized test meal (22% on omeprazole vs. 16% off omeprazole).(17) Large observational studies have demonstrated weak associations between PPI use and fragility bone fractures. (11, 12) But exposure to PPIs for less than 1 year has been associated with bone fractures;(11) such a short duration should not be explained by malabsorption of calcium. Furthermore, in a large longitudinal cohort, patients who used PPIs had similar progression of bone mineral density compared to those not using PPIs.(18) In states of calcium malabsorption, secondary hyperparathyroidism should develop. Yet, we found no evidence of an association between PPI use and hyperparathyroidism. Together, these findings suggest that PPIs do not cause clinically significant malabsorption of calcium.
Our study has a few important limitations. The study was observational, cross-sectional, restricted to men, and we do not have data on outcomes of esophageal adenocarcinoma. There could have been misclassification of supplemental calcium intake as the questionnaire for supplemental calcium queried regarding “calcium or TUMS” but not with other specific calcium antacids.
Our study also had a few notable strengths. We were able to directly assay vitamin D and PTH regardless of clinical indication in a large cohort. Our unique study design recruiting colorectal cancer screenees allowed for enrollment of substantial numbers of patients with GERD symptoms, erosive esophagitis, or Barrett’s esophagus who had not used PPIs, and we were able to adjust for potential confounding by PPI use and other potential confounders. Finally, we addressed the seasonal variation in vitamin D and PTH concentration by calculating the residual difference between predicted and observed concentrations, finding associations similar to the primary analyses.
In conclusion, we found evidence that Barrett’s esophagus is inversely associated with vitamin D deficiency and with hyperparathyroidism. Further research is needed to understand the mechanism of those associations with Barrett’s esophagus.
Supplementary Material
Acknowledgments
Funding: Funding was provided by a 2015 American College of Gastroenterology Clinical Research Award. In addition, JHR’s effort was funded by the U.S. Department of Veterans Affairs (I01 CX000899).
Abbreviations:
- AAVAHS
Ann Arbor Veterans Affairs Health System
- GERD
gastroesophageal reflux disease
- H2RA
histamine receptor type-2 antagonists
- PPI
proton pump inhibitor
- PTH
parathyroid hormone
- PTHrP
parathyroid hormone related peptide
- UMHS
University of Michigan Health System
Footnotes
Disclosures: None of the authors have potential conflicts of interest.
REFERENCES
- 1.Tran B, Lucas R, Kimlin M, Whiteman D, Neale R, Australian Cancer S. Association between ambient ultraviolet radiation and risk of esophageal cancer. Am J Gastroenterol 2012. December;107(12):1803–13. PubMed PMID: 23032986. English. [DOI] [PubMed] [Google Scholar]
- 2.Kedar A, Nikitina Y, Henry OR, Abell KB, Vedanarayanan V, Griswold ME, et al. Gastric dysmotility and low serum vitamin D levels in patients with gastroparesis. Horm Metab Res 2013. January;45(1):47–53. PubMed PMID: 22956309. English. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Christensen MHE, Dankel SN, Nordbø Y, Varhaug JE, Almås B, Lien EA, et al. Primary Hyperparathyroidism Influences the Expression of Inflammatory and Metabolic Genes in Adipose Tissue. PLoS ONE 2011;6(6):e20481. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rubenstein JH, Morgenstern H, McConell D, Scheiman JM, Schoenfeld P, Appelman H, et al. Associations of Diabetes Mellitus, Insulin, Leptin, and Ghrelin with Gastroesophageal Reflux and Barrett’s Esophagus. Gastroenterology 2013;145(6):1237–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Quigley EM, Jacobson BC, Lenglinger J, Rubenstein JH, El-Serag H, Cicala M, et al. Barrett’s esophagus: clinical features, obesity, and imaging. Ann N Y Acad Sci 2011. September;1232:36–52. PubMed PMID: 21950806. English. [DOI] [PubMed] [Google Scholar]
- 6.Rubenstein JH, Morgenstern H, Appelman H, Scheiman J, Schoenfeld P, McMahon LF Jr., et al. Prediction of Barrett’s esophagus among men. Am J Gastroenterol 2013. March;108(3):353–62. PubMed PMID: 23318485. English. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rubenstein JH, Inadomi JM, Scheiman J, Schoenfeld P, Appelman HD, Zhang M, et al. Association between Helicobacter pylori and Barrett’s Esophagus, Erosive Esophagitis, and Gastroesophageal Reflux Symptoms. Clinical Gastroenterology and Hepatology 2013: 10.1016/j.cgh.2013.08.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rubenstein JH, Morgenstern H, Chey WD, Murray J, Scheiman J, Schoenfeld P, et al. Protective role of gluteofemoral obesity in erosive oesophagitis and Barrett’s oesophagus. Gut 2014;63(2):230–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rubenstein JH, Morgenstern H, Kellenberg J, Kalish T, Donovan J, Inadomi J, et al. Validation of a new physical activity questionnaire for a sedentary population. Dig Dis Sci 2011. September;56(9):2678–87. PubMed PMID: 21409377. English. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Block G, Hartman AM, Naughton D. A reduced dietary questionnaire: development and validation. Epidemiology 1990. January;1(1):58–64. PubMed PMID: 2081241. [DOI] [PubMed] [Google Scholar]
- 11.Ngamruengphong S, Leontiadis GI, Radhi S, Dentino A, Nugent K. Proton pump inhibitors and risk of fracture: a systematic review and meta-analysis of observational studies. Am J Gastroenterol 2011. July;106(7):1209–18; quiz 19. PubMed PMID: 21483462. English. [DOI] [PubMed] [Google Scholar]
- 12.Khalili H, Huang ES, Jacobson BC, Camargo CA Jr., Feskanich D, Chan AT. Use of proton pump inhibitors and risk of hip fracture in relation to dietary and lifestyle factors: a prospective cohort study. Bmj 2012;344:e372 PubMed PMID: 22294756. Pubmed Central PMCID: PMC3269660. Epub 2012/02/02. eng. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Abnet CC, Chen Y, Chow W- H, Gao Y- T, Helzlsouer KJ, Le Marchand L, et al. Circulating 25-hydroxyvitamin D and risk of esophageal and gastric cancer: Cohort Consortium Vitamin D Pooling Project of Rarer Cancers. Am J Epidemiol 2010. July 1;172(1):94–106. PubMed PMID: 20562192. Pubmed Central PMCID: Source: NLM. PMC2892544. English. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Zgaga L, O’Sullivan F, Cantwell MM, Murray LJ, Thota PN, Coleman HG. Markers of Vitamin D Exposure and Esophageal Cancer Risk: A Systematic Review and Meta-analysis. Cancer Epidemiology, Biomarkers & Prevention 2016;25(6):877–86. PubMed PMID: 27030602. [DOI] [PubMed] [Google Scholar]
- 15.Tuma SN, Mukhopadhyay A. The effect of parathyroid hormone on the esophageal smooth muscle of the opossum. Am J Gastroenterol 1980;74(5):415–8. PubMed PMID: 7234818. [PubMed] [Google Scholar]
- 16.O’Connell MB, Madden DM, Murray AM, Heaney RP, Kerzner LJ. Effects of proton pump inhibitors on calcium carbonate absorption in women: a randomized crossover trial. Am J Med 2005. July;118(7):778–81. PubMed PMID: 15989913. English. [DOI] [PubMed] [Google Scholar]
- 17.Serfaty-Lacrosniere C, Wood RJ, Voytko D, Saltzman JR, Pedrosa M, Sepe TE, et al. Hypochlorhydria from short-term omeprazole treatment does not inhibit intestinal absorption of calcium, phosphorus, magnesium or zinc from food in humans. J Am Coll Nutr 1995. August;14(4):364–8. PubMed PMID: 8568113. English. [DOI] [PubMed] [Google Scholar]
- 18.Targownik LE, Lix LM, Leung S, Leslie WD. Proton-pump inhibitor use is not associated with osteoporosis or accelerated bone mineral density loss. Gastroenterology 2010. March;138(3):896–904. PubMed PMID: 19931262. English. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.