a Department of Health Statistics and b Information Center, Second Military

Transcription

1 794 Original article Proton pump inhibitors therapy and risk of hip fracture: a systematic review and meta-analysis Xiaofei Ye a, *,HongLiu b, *, Cheng Wu a, *, Yingyi Qin a, Jiajie Zang a, Qingbin Gao a, Xinji Zhang a and Jia He a Background and aims Previous studies have reported inconsistent findings that proton pump inhibitors (PPIs) therapy might increase the risk of hip fracture. We investigated the association between PPIs therapy and hip fracture by a systematic review and meta-analysis. Methods We systematically searched PubMed, EMBASE, and the Cochrane Library. We included studies assessing the effects of PPIs on hip fracture. Data from the studies about odds ratio and 95% confidence interval were gathered and summarized. Results Seven studies met the inclusion criteria. PPIs therapy was associated with a statistically significant increase of hip fracture risk (pooled odds ratio = 1.24; 95% confidence interval: ; P < ) under a random model. Meanwhile, we found that the effect of PPIs on hip fracture differs in different duration groups. Conclusion These results indicate that PPIs therapy might have the potential risk of hip fracture. Different effects on hip fracture in the subgroup analysis do not support a causal relationship between PPIs and hip fracture. Whether the risk exists warrants further investigation. Eur J Gastroenterol Hepatol 23: c 2011 Wolters Kluwer Health Lippincott Williams & Wilkins. European Journal of Gastroenterology & Hepatology 2011, 23: Keywords: hip fracture, meta-analysis, proton pump inhibitors, systematic review a Department of Health Statistics and b Information Center, Second Military Medical University, Shanghai, China Correspondence to Jia He, PhD, Department of Health Statistics, Second Military Medical University, Shanghai , China Tel/fax: ; fax: ; hejia63@yahoo.com * Xiaofei Ye, Hong Liu, and Cheng Wu contributed equally to the writing of this article Received 7 March 2011 Accepted 8 May 2011 Introduction Proton pump inhibitors (PPIs) are widely prescribed medications with an excellent short-term safety profile in clinical practice since their introduction in the late 1980s. They are used for the treatment of various disorders, ranging from dyspepsia to gastroesophageal reflux disease, with a mechanism to inhibit the gastric acid secretion [1]. Countless individuals have benefited from them on a continuous or long-term basis. In Finland, 22% of the nursing home residents were administrated PPIs regularly [2]. Similar to any other medication, PPIs have unexpected potential adverse consequences and benefits. It has been reported that PPIs may lead to headache, nausea, diarrhea, abdominal pain, community-acquired pneumonia, acute interstitial nephritis, and anaphylactic reactions [3 7]. Recently, several studies from a variety of data sources have suggested that there might be a relationship between PPIs and fracture, especially hip fracture. The US Food and Drug Administration has also alerted consumers and healthcare providers regarding the potential increased risk of hip, wrist, and spine fractures associated with high doses or long-term use of PPIs [8]. Fracture is a common injury around the world. It is estimated that the number of people with fractures worldwide was approximately 56 million in 2000, and approximately 9 million new osteoporotic fractures occur each year [9]. The global number of hip fracture is estimated to be 7 21 million by 2050 [10]. Morbidity and mortality are associated with fracture. Mortality associated with hip fracture is approximately 5 10% after 1 month and 20% during the first year [10 12]. High prevalence of fracture leads to high health expenditures, lost productivity, and lower quality of life of the victims. Hence, it would be of substantial importance to identify the potential risk factors of fracture. An acidic environment in the gastrointestinal tract facilitates the release of ionized calcium from insoluble calcium salts [13]. The presence of gastric acid facilitates the absorption of dietary calcium. It has been reported that the use of PPIs decreases calcium absorption in the stomach, which increases the risk for hip fracture. Conversely, PPIs may also reduce bone resorption through proton pump inhibition of osteoclastic cells, which may decrease the risk of hip fracture [14]. Limited animal and human studies have shown that PPIs therapy may decrease insoluble calcium absorption or bone density [15,16]. A number of studies have been conducted to assess the effect of PPIs on hip fracture [17 19]. However, studies investigating the association between PPIs and hip fracture have reported conflicting results. Some studies suggested there is an association between them, whereas others argued that PPIs had no effect on hip fracture. To X c 2011 Wolters Kluwer Health Lippincott Williams & Wilkins DOI: /MEG.0b013e328348a56a

2 PPIs and hip fracture Ye et al. 795 address this uncertainty, we performed a systematic review and meta-analysis to determine whether the association between PPIs and hip fracture exists quantitatively. Methods We conducted this study following the checklist for metaanalyses of observational studies (Meta-analysis Of Observational Studies in Epidemiology) proposed by Stroup et al. [20]. Data collection Related studies in all languages were identified by searching multiple electronic databases including: PubMed (from 1976 to June 2010), EMBASE (from 1966 to June 2010), and the Cochrane Library (no date restriction). The search was carried out in June We restricted the search to studies on human participants, using the following terms: proton pump inhibitors, proton pump inhibitor, PPIs, omeprazole, lansoprazole, pantoprazole, rabeprazole, leminoprazole, esomeprazole, dexlansoprazole, and fracture. Study selection We assessed the quality of included studies by the Strengthening the Reporting of Observational Studies in Epidemiology checklist, taking 22 items as gauge. Only those studies of high quality could be included in a metaanalysis to reach a credible conclusion. To be included in this systematic review and metaanalysis, trials should have compared hip fracture between PPI users and non-ppi users. Cohort study, case control study, or similar designs were included. The reviewers extracted the data independently using standardized data collection forms. The following information was collected from the identified trials: publication information (title, first author s name, year), study characteristics (study design, study region, endpoint index, regimen details), characteristics of study population (sample size, age), and outcome for the whole study populations. The information about fracture, exposure ascertainment, and control variables was shown in Table 1. The literature search and data extraction were undertaken independently by two investigators (X.Y., Y.Q.) to determine whether they met the criteria for inclusion. When there was discrepancy between the two investigators, the two investigators discussed with another investigator (J.H.) to make a consensus. Statistical analysis Odds ratio () was used as a measurement for the relationship between PPIs therapy and the risk of hip fracture. Differences observed between groups were expressed as with 95% confidence interval (CI). For cohort studies or prospective studies, relative risk and hazard ratio were used as a surrogate measure of the corresponding. As the incidence of hip fracture is low, the approximates the relative risk and hazard ratio [27]. Individual and 95% CI were extracted or calculated initially, and then log () and its corresponding standard error were estimated for pooling. Statistical heterogeneity between trials was evaluated by w 2 test and I 2 statistic, and was considered to be present when the P value is less than 0.1. In the presence of statistical heterogeneity, a random-effect model was used for the analysis. In the absence of statistically significant heterogeneity, the by the fixed-effect model is calculated. Subgroup analyses were conducted to compare the effect in different subtypes, such as region, dosage, and duration. We assessed the presence of publication bias by examining funnel plot and by calculating the Egger coefficient. All statistical analyses were conducted with the freeware Review Manager (version 5.0 for Windows, Cochrane Collaboration, Oxford, UK, 2010) and STATA 11.0 (Stata Corporation, Lakeway, Texas, USA). Results Results of search strategy The literature search yielded 109 articles and all were reviewed in detail using the search strategy summarized in Fig. 1. Thirteen studies were reviewed as potentially relevant studies, of which six trials were excluded Table 1 Information about fracture, exposure ascertainment, and control variables Number Source Information about fracture Exposure ascertainment Control variables 1 ICD codes Automated pharmacy Year of birth, sex, geographical region 2 ICD codes Automated pharmacy Year of birth, sex, duration of membership, first year of membership, race 3 Adjudicated Patient interview Race, history of fracture, current and past smoking 4 De Vries et al. [23] ICD-9 Read oxmis codes Diabetes mellitus, rheumatoid arthritis, other diseases, smoking status, BMI, history of fracture 5 Targownik et al. [24] ICD-9 Automated pharmacy Age, sex, comorbidity 6 Yu et al. [25] adjudicated Patient interview Age, clinic, race, BMI, alcohol use, exercise, NSAID use, etc. 7 Vestergaard et al. [26] ICD-10 Automated pharmacy Age, sex, other drug information, comorbidity, social information ICD, International Classification of Diseases.

3 796 European Journal of Gastroenterology & Hepatology 2011, Vol 23 No 9 Fig. 1 Potentially relevant citations identified and screened for retrieval (n=109) Abstracts and title excluded during first screening (n=96) Case control studies Cross-sectional studies Nested case control studies Prospective studies Retrospective cohort studies (n=3) (n=3) (n=3) Studies excluded (n= 6) Reasons for exclusion: PPI and osteoporosis, not fracture Not hip fracture Cross-sectional study Studies from same database (n=1) (n=1) Case control studies Nested case control studies Prospective studies Retrospective cohort studies (n=1) Studies with usable information By outcomes (n=7) Flow chart. PPI, proton pump inhibitor. according to the exclusion criteria. The remaining seven studies met the inclusion criteria [14,21 23,25,26,28]. The characteristics of the original studies were presented in Table 2. All these studies were conducted in North America and Europe. In a study conducted by Yu et al. [25], two separate databases, Osteoporotic Fractures in Men Study and the Study of Osteoporotic Fractures for women were studied. We took them as two independent studies and eight studies in total were included in this study. Five studies mentioned the dosage of PPIs and only five studies mentioned about the medication duration. Together, participants were included in this meta-analysis. In a study conducted by, patients were divided into several groups according to the exposure time. We just selected the adjusted from the shortest exposure time group with time greater than 1 year. In a study conducted by, data from the patients, who currently used PPIs, were extracted. Effect of proton pump inhibitors on hip fracture From the eight studies, meta-analysis showed a significantly increased risk of hip fracture for patients who received PPIs compared with non-ppis (pooled = 1.24; 95% CI: ; P < ) under a random model (w 2 = 12.56; I 2 = 44%; P = 0.08; Fig. 2). There was no evidence of publication bias in our study and the Egger coefficient was also calculated (Egger = 1.552; P = 0.064). Region All the included studies were conducted in Europe and North America. We conducted a subgroup analysis for these two regions. Three studies conducted in Europe revealed an association between PPIs therapy and hip fracture (pooled = 1.29; 95% CI: ; P < ) under a random model. Another five studies carried out in North America also demonstrated such association (pooled = 1.26; 95% CI: ; P < ) under a fixed model (Fig. 3). No significant

4 PPIs and hip fracture Ye et al. 797 Table 2 Characteristic of the included studies Source Study design Index Region Study participants Dosage Duration Pouwels et al. [14] Case control study Nested case control study Netherland United States 6763 cases of hip/femur fractures aged 18 years and older; controls cases of hip/femur fractures and controls Prospective study HR United States postmenopausal women years old De Vries et al. [29] Targownik et al. [28] Retrospective cohort study Retrospective cohort study RR <1 DDD; DDD; >1.75 DDD < 0.75 pills/ day; pills/day; Z 1.5 pills/ day UK GPRD PPI user < 1 DDD; DDD; >1.75 DDD Manitoba, Canada cases of osteoporosis-related fracture and controls; aged 50 years and older NA NA First 3 months 3 12 months months > 36 months > 7 years < 1 year 1 2 years 2 4 years 4 6 years 6 8 years 8 10 years > 10 years < 1 year 1 3 years > 3 years < 1 year 1 2 years 2 3 years > 3 years Z 1 year to Z 7 years at 1 year interval Yu et al. [25] Prospective study Relative hazard United States MrOS study: 5755 men; aged 65 years and older NA NA Relative hazard SOF study: 5339 women; aged 65 years and older NA NA Vestergaard et al. [26] Case control study Denmark fracture cases and controls < 0.25 DDD; DDD; Z 1DDD < 3 months DDD, defined daily dosages; GPRD, General Practice Research Database; HR, hazards ratio; MrOS, Osteoporotic Fractures in Men Study; NA, not available;, odds ratio; RR, relative risk; SOF, Study of Osteoporotic Fractures. Fig. 2 Study (95% CI) Weight (%) De Vries et al. [23] Yu _SOF et al. [25] Yu _MrOS et al. [25] Vestergaard et al. [26] 1.20 ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 18.0 Pooled Favors non-ppi users 1 2 Favors PPI users 1.24 ( ) Z = 5.42, P < Heterogeneity (χ 2 = 12.56, I 2 = 44%, P = 0.08) Total effect. CI, confidence interval;, odds ratio; PPI, proton pump inhibitor; SOF, Study of Osteoporotic Fractures. interactions were found in different regions (interaction P = 0.735). Duration Whether the effect of long duration use of PPIs on hip fracture existed was considered in several studies, but different durations were analyzed in different studies, which were summarized in Table 2. We performed a subgroup analysis according to three durations: short duration (< 1 year), medium duration (1 3 years), and long duration (> 3 years).weextractedthedatafromtheoriginalstudiesfirst and then combined the data together. Some studies did not

5 798 European Journal of Gastroenterology & Hepatology 2011, Vol 23 No 9 Fig. 3 Study (95% CI) Weight (%) Region Europe De Vries et al. [21] Vestergaard et al. [26] 1.20 ( ) ( ) ( ) 33.5 Duration <1 year North America Yu _SOF et al. [25] Yu _MrOS et al. [25] De Vries et al. [23] 1 3 years De Vries F [23] >3 years De Vries et al. [23] >6 years 1.29 ( ); P < (I 2 =61%, χ 2 =5.15, P = 0.08) 1.00 ( ) 1.30 ( ) 1.09 ( ) 1.16 ( ) 0.62 ( ) 1.26 ( ); P < (I 2 =45%, χ 2 =7.22, P = 0.12) 1.28 ( ) 1.00 ( ) 1.25 ( ) 1.31 ( ) 1.25 ( ); P < (I 2 =0%, χ 2 =1.02, P = 0.80) 1.18 ( ) 0.98 ( ) 1.22 ( ) 1.19 ( ); P = (I 2 =0%, χ 2 =0.68, P = 0.71) 1.09 ( ) 1.01 ( ) 1.17 ( ) 0.97 ( ) 1.11 ( ); P = 0.14 (I 2 =0%, χ 2 =1.04, P = 0.79) 1.38 ( ) 2.49 ( ) 1.40 ( ); P = (I 2 =70%, χ 2 =3.33, P = 0.07) Favors non-ppi users 1 2 Favors PPI users Subgroup analysis. CI, confidence interval;, odds ratio; PPI, proton pump inhibitor; SOF, Study of Osteoporotic Fractures. give an exact effect for our expected duration. For example, in the study conducted by Pouwel et al. [14], they just analyzed the duration of more than 3 months and 4 12 months. To get the duration of more than 1 year, we first combined these two durations and got the whole effect on hip fracture into a duration of less than 1 year. In addition, we performed a subgroup analysis of a long-term duration for a further 6 years.

6 PPIs and hip fracture Ye et al. 799 Our analysis indicated that there was association between PPIs and hip fracture in the short-duration group (pooled = 1.25; 95% CI: ; P < ) and mediumduration group (pooled = 1.19; 95% CI: ; P = 0.006). No association was observed in the longduration group (pooled = 1.11; 95% CI: ; P = 0.14). However, association was observed in the group of duration greater than 6 years (pooled = 1.40; 95% CI: ; P < ). Discussion In our study, eight studies met the inclusion criteria. We observed that PPIs therapy was associated with the risk of hip fracture (pooled = 1.24; 95% CI: ; P < ) under a random model. Different effects on hip fracture in the subgroup analysis did not support a causal relationship between PPIs and hip fracture. If there is a really causal association between PPIs and hip fracture, the effect might increase with the duration. Whether the risk exists warrants further investigation. The mechanism by which extended use of PPIs increases the risk of hip fracture remains unknown. In-vitro calcium carbonate dissolution and disintegration is dependent on low ph. O Connell et al. [16] found that during omeprazole (20 mg) therapy, significantly decreased absorption of calcium was taken by elderly women (Z 65 years) after an overnight fast and on an empty stomach. However, a study conducted by Serfaty- Lacrosniere et al. [30] demonstrated that moderate hypochlorhydria resulting from short-term omeprazole treatment does not substantially lead to calcium deficiencies due to mineral malabsorption. PPIs also inhibit the osteoclastic proton transport system, but the affinity for the osteoclast-type proton pump is much weaker than gastric pump [31]. Further studies about the mechanism are requested. Other factors may contribute to the incidence of hip fracture, a strategy of stratification of sex was performed in the study by. The risk of fracture was statistically significantly higher among current PPI users who were men ( = 1.57; 95% CI: ) compared with women ( = 1.12; 95% CI: ; [14]). In a recent study by Targownik et al. [24], they found that the association between PPI use and hip fracture is probably related to factors independent of osteoporosis. Calcium supplement tended to be another contributor to hip fracture. An increased hip fracture incidence was observed among male PPI users without calcium supplement, but not among male PPI users with calcium supplement [25]. In the process of study selection, we observed that the studies conducted by Kaye and Jick [32], Yang et al. [33], and De Vries et al. [23] came from the same database UK General Practice Research Database (GPRD). Perhaps, these three studies largely analyzed the same study population. We just chose the GPRD study with the largest and most recent version of the database. We also found that the use of International Classification of Diseases (ICD) codes and relying on pharmacy data to capture data might lead to information bias in the original studies. The way of adjudicated diagnosis and patient interview is better in reducing the information bias. We noticed that even the same database GPRD used by three separate studies, resulted in different conclusions. A negative association was observed in the Kaye and Jick [32] study and a positive association was obtained in studies conducted by Yang et al. [33] and De Vries et al. [23]. Selection bias could explain the difference [29]. Kaye just selected the patients without major risk factors into their study, which might suggest there was a possibility that other risk factors contribute to the hip fracture. In addition, the study by De Vries et al. [23] found a small positive association but the authors emphasized that there was inverse association between duration use of acid-suppressive medications and fracture risk, which does not support adverse effects on the bone. Our study has some limitations that deserve further discussion. First and foremost, this meta-analysis used data from epidemiological studies (case control study, cohort study, and similar designs), which were unable to control three major potential confounders: alcohol use (none of the studies, patients tend to lie about this also to their general practitioner), smoking status, and BMI. Second, we just extracted part of data from two original studies, as there were no data about the total effect, which might distort the intrinsic relationship. Third, the included studies were conducted only in Europe and North America, no evidence was found in other continents. Fourth, we only searched for the existing publications concerning PPIs and fractures; this would lead to inclusion of mere epidemiological studies of drug safety. The original randomized clinical trials reporting efficacy and adverse effects of PPIs should be explored and reported. However, we did not contact the drug companies or original researchers to get the unreported adverse events. Cohort studies, case control studies, and cross-sectional studies were included in our study to investigate the relationship between PPIs and hip fracture. In fact, the evidence supported by these designs was less strong than the randomized controlled trials. These studies show only association, and causal relationship is impossible to conclude from these studies. The conclusion drawn from these studies should be explained cautiously. Due to the limitation of meta-analysis, further studies are required to determine the clinical importance of our findings. Specifically designed studies are needed to investigate the effect of PPIs on hip fracture. Long-term, well-designed, matched cohort study or randomized controlled trials (RCTs) conducted to explicitly assess the incidence of osteoporotic fractures, would be the

7 800 European Journal of Gastroenterology & Hepatology 2011, Vol 23 No 9 preferred experimental approach [34]. RCT, such as those conducted in clinical trials, needs high expenditure and practitioners in this area. It requires several rigorous processes including participant enrolment, randomization, and so on. It is suggested that pharmaceutical companies manufacturing PPIs should take the responsibility to conduct such RCT. In conclusion, this meta-analysis shows that PPIs therapy perhaps has the potential risk of increasing hip fracture. Different effects on hip fracture in the subgroup analysis do not support the causal relationship between PPIs and hip fracture. Whether the risk exists warrants further investigation. Acknowledgements This study was conducted under two grants from the National Nature Science Foundation of China ( , ): a grant sponsored by Program of Shanghai Subject Chief Scientist (09XD ), a grant from the leading talents of science in Shanghai 2010 (022), and a grant from the innovative program for PhD in Second Military Medical University. We are indebted to the authors of primary studies. Author contributions: X.Y. and H.L. drafted and revised the manuscript, contributed to the study design, analysis, and interpretation of the data. C.W., J.Z., and X.Z. contributed to the data analysis and revised the manuscript. Y.Q. and Q.G. contributed to the revision of the manuscript. J.H. contributed to the interpretation of the results and the conception of the study. Conflicts of interest There are no conflicts of interest. References 1 Savarino V, Di Mario F, Scarpignato C. Proton pump inhibitors in GD: an overview of their pharmacology, efficacy and safety. Pharmacol Res 2009; 59: Teramura-Gronblad M, Hosia-Randell H, Muurinen S, Pitkala K. Use of proton-pump inhibitors and their associated risks among frail elderly nursing home residents. Scand J Prim Health Care 2010; 28: Laheij RJ, Sturkenboom MC, Hassing RJ, Dieleman J, Stricker BH, Jansen JB. Risk of community-acquired pneumonia and use of gastric acidsuppressive drugs. JAMA 2004; 292: Harmark L, Van der Wiel HE, de Groot MC, Van Grootheest AC. Proton pump inhibitor-induced acute interstitial nephritis. Br J Clin Pharmacol 2007; 64: Lobera T, Navarro B, Del Pozo MD, González I, Blasco A, Escudero R, et al. Nine cases of omeprazole allergy: cross-reactivity between proton pump inhibitors. J Investig Allergol Clin Immunol 2009; 19: Martin RM, Dunn NR, Freemantle S, Shakir S. The rates of common adverse events reported during treatment with proton pump inhibitors used in general practice in England: cohort studies. Br J Clin Pharmacol 2000; 50: Lodato F, Azzaroli F, Turco L, Mazzella N, Buonfiglioli F, Zoli M, Mazzella G. Adverse effects of proton pump inhibitors. Best Pract Res Clin Gastroenterol 2010; 24: Voelker R. Proton pump inhibitors linked to fracture risk. JAMA 2010; 304:29. 9 Johnell O, Kanis JA. An estimate of the worldwide prevalence and disability associated with osteoporotic fractures. Osteoporos Int 2006; 17: Parker M, Johansen A. Hip fracture. BMJ 2006; 333: Cumming RG, Nevitt MC, Cummings SR. Epidemiology of hip fractures. Epidemiol Rev 1997; 19: Leibson CL, Tosteson AN, Gabriel SE, Ransom JE, Melton LJ. Mortality, disability, and nursing home use for persons with and without hip fracture: a population-based study. J Am Geriatr Soc 2002; 50: Sheikh MS, Santa Ana CA, Nicar MJ, Schiller LR, Fordtran JS. Gastrointestinal absorption of calcium from milk and calcium salts. N Engl J Med 1987; 317: Pouwels S, Lalmohamed A, Souverein P, Cooper C, Veldt BJ, Leufkens HG, et al. Use of proton pump inhibitors and risk of hip/femur fracture: a population-based case-control study. Osteoporos Int 2011; 22: Chonan O, Takahashi R, Yasui H, Watanuki M. Effect of L-lactic acid on calcium absorption in rats fed omeprazole. J Nutr Sci Vitaminol (Tokyo) 1998; 44: 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; 118: Gendreau-Webb R. Bone up on proton pump inhibitors and fracture risk. Nursing 2007; 37: Laine L. Proton pump inhibitors and bone fractures? Am J Gastroenterol 2009; 104(Suppl 2):S21 S Moayyedi PCommittee CAGCA. Hip fracture and proton pump inhibitor therapy: position statement. Can J Gastroenterol 2008; 22: Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, etal. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000; 283: Corley DA, Kubo A, Zhao W, Quesenberry C. Proton pump inhibitors and histamine-2 receptor antagonists are associated with hip fractures among at-risk patients. Gastroenterology 2010; 139: Gray SL, LaCroix AZ, Larson J, Robbins J, Cauley JA, Manson JE, Chen Z. Proton pump inhibitor use, hip fracture, and change in bone mineral density in postmenopausal women: results from the Women s Health Initiative. Arch Intern Med 2010; 170: De Vries F, Cooper AL, Cockle SM, Van Staa TP, Cooper C. Fracture risk in patients receiving acid-suppressant medication alone and in combination with bisphosphonates. Osteoporos Int 2009; 20: 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; 138: Yu EW, Blackwell T, Ensrud KE, Hillier TA, Lane NE, Orwoll E, Bauer DC. Acid-suppressive medications and risk of bone loss and fracture in older adults. Calcif Tissue Int 2008; 83: Vestergaard P, Rejnmark L, Mosekilde L. Proton pump inhibitors, histamine H2 receptor antagonists, and other antacid medications and the risk of fracture. Calcif Tissue Int 2006; 79: Reynolds K, Lewis B, Nolen JD, Kinney GL, Sathya B, He J. Alcohol consumption and risk of stroke: a meta-analysis. JAMA 2003; 289: Targownik LE, Lix LM, Metge CJ, Prior HJ, Leung S, Leslie WD. Use of proton pump inhibitors and risk of osteoporosis-related fractures. CMAJ 2008; 179: De Vries F, Van Staa TP, Leufkens HG. Proton pump inhibitors, fracture risk and selection bias: three studies, same database, two answers. Osteoporos Int 2011; 22: Serfaty-Lacrosniere C, Wood RJ, Voytko D, Saltzman JR, Pedrosa M, Sepe TE, Russell RR. 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; 14: Zaidi M. Modularity of osteoclast behaviour and mode-specific inhibition of osteoclast function. Biosci Rep 1990; 10: Kaye JA, Jick H. Proton pump inhibitor use and risk of hip fractures in patients without major risk factors. Pharmacotherapy 2008; 28: Yang YX, Lewis JD, Epstein S, Metz DC. Long-term proton pump inhibitor therapy and risk of hip fracture. JAMA 2006; 296: Richards JB, Goltzman D. Proton pump inhibitors: balancing the benefits and potential fracture risks. CMAJ 2008; 179: