Article Figures & Data
Figures
- Figure 1.
Study flow diagram.
- Figure 2.
PPIs or H2 RAs use and the combined risk of any fracture in a random-effects model meta-analysis of case-controls studies and cohort studies.
CI=confidence interval; H2RA = histamine 2 receptor antagonist; OR = odds ratio; PPI = proton pump inhibitor; RR = relative risk.
Tables
Study Country; Study Design or Setting Study Years Outcome and Ascertainment Agent and Comparison Adjusted OR or RR (95% CI) Adjustment Variables Exposed and Unexposed Case/Control, n BMI = body mass index; CHF = congestive heart failure; COPD = chronic obstructive pulmonary disease; CRF = chronic renal failure; DMARD=disease-modifying anti-rheumatic drug; GPRD = General Practice Research Database; HCC = hospital-based case-control; ICD=International Classification of Diseases; H2RA=histamine 2 receptor antagonist; HRT = hormone replacement therapy; MI = myocardial infarction; MrOS = Osteoporotic Fractures in Men study; na = not available; NCC = nested case-control; NSAID=nonsteroidal anti-inflammatory drug; OR = odds ratio; PCC = population-based case-control; PPI = proton pump inhibitor; RA = rheumatoid arthritis; RR=relative risk; SOF = Study of Osteoporotic Fractures. aCombined OR with 95% CI calculated from data set. Grisso et al,49 1997 USA; HCC; Kaiser Permanente, northern California; men aged ≥45 y 1991–1993 Outcome: hip fracture Ascertainment: radiologically confirmed diagnosis Agent: cimetidine Comparison: ever use vs never use 2.00 (1.0–4.4) Age category, zip code or telephone exchange, age as a continuous variable, and BMI Exposed: 39/26 Unexposed: 317/376 Vestergaard et al,39 2006 Denmark; PCC; whole population 2000 Outcome: hip fracture, spine fracture, forearm fracture, any fracture Ascertainment: insurance code using ICD system Agent: PPIs, H2RA, other antacids, antihistamines, and NSAIDs Comparison: last use <1 y ago vs nonuse PPIs Total: 1.18 (1.1–1.4) Hip: 1.45 (1.3–1.7) Spine: 1.60 (1.3–2.0) Forearm: 0.95 (0.8–1.1) H2RAs Total: 0.88 (0.8–0.95) Hip: 0.69 (0.6–0.8) Spine: 1.00 (0.7–1.4) Forearm: 0.90 (0.7–1.1) Alcoholism; working or not; Charlson index; ever use of antiepileptic drugs, anxiolytics or sedatives, antidepressants, neuroleptics, corticosteroids; number of bed-days in 1999; number of contacts to general practitioner or specialist in 1999; living with someone or alone; prior fracture; education level; and income in 1999 Exposed: PPIs: 14,557/29,784 H2RAs: 11,202/26,333 Unexposed: PPIs: 10,098/344,178 H2RAs: 113,453/347,629 Yang et al,22 2006 UK; NCC; GPRD; patients aged ≥50 y 1987–2003 Outcome: hip fracture Ascertainment: confirmed by general practitioners Agent: PPIs and H2RAs Comparison: user (>1 y) vs nonuser PPIs: 1.44 (1.3–1.6) H2RAs: 1.23 (1.1–1.4) Age, sex, BMI, medication use, and health condition Exposed: PPIs: 571/3,351 H2RAs: 732/4,478 Unexposed: PPIs: 12,985/132,035 H2RA: 12,824/130,908 Kaye and Jick,42 2008 UK; NCC; GPRD; patients aged 50–79 y 1995–2005 Outcome: hip fracture Ascertainment: computerized medical records (Oxford Medical Information System or Read codes) Agent: PPIs Comparison: any PPI prescription vs no prescription 0.90 (0.7–1.1) Age, sex, smoking, and BMI Exposed: 132/1,428 Unexposed: 966/9,495 Targownik et al,23 2008 Canada; PCC; patients aged ≥50 y 1996–2004 Outcome: fractures of hip, vertebra, and wrist Ascertainment: physician or hospital admission with a diagnosis of osteoporosis-related fractures (ICD system) Agent: PPIs Comparison: user vs nonuser Exposure ≥7 y Hip+vertebra+wrist: 1.92 (1.2–3.2) Hip+vertebra: 2.47 (1.2–5.1) Hip: 4.55 (1.7–12.3) Area of residence, income, comorbidity, home care services, medications that might have affected the risk of osteoporosis or fracture Exposed: 415/872 Unexposed: 12,963/39,176 Yu et al,40 2008 USA; cohort; men and women aged ≥65 y 1986–2007 Outcome: nonspine and hip fracture Ascertainment: review of radiographic reports or radiologist Agent: PPIs and H2RAs Comparison: user vs nonuser PPIs Nonspine: 1.3 (1.1–1.53)a Hip: 1.05 (0.8–1.5)a H2RAs Nonspine: 1.04 (0.9–1.2)a Hip: 1.26 (0.9–1.7)a Age, clinic, race, BMI, alcohol, exercise, corticosteroids, NSAIDs, calcium supplement, osteoporosis medication, self-reported health, initial total hip BMD, weight change, (SOF; caffeine, estrogen), (MrOS; smoking, history of stomach surgery) na Roux et al,41 2009 France; cohort; women aged 55–79 y 1999–2007 Outcome: vertebral fracture Ascertainment: 2 radiologists Agent: omeprazole Comparison: user vs nonuser 3.10 (1.1–8.4) BMI, history of clinical low-trauma fracture, steroids, thiazide, thyroid hormone, calcium, vitamin D, smoking, alcohol, self-reported falls and health poor to fair, baseline lumbar spine T score, spine and/or hip T score ≤2.5 Exposed: risk among exposed=14/61 Unexposed: risk among unexposed=128/1,150 Gray et al,51 2010 USA; cohort; Women’s Health Initiative; women aged 50–79 y 1993–2005 Outcome: hip, spine, forearm, or wrist fracture Ascertainment: review of radiology reports Agent: PPIs and H2RAs Comparison: user vs nonuser PPIs Total: 1.25 (1.2–1.4) Hip: 1.00 (0.7–1.4) Spine: 1.47 (1.2–1.8) Forearm or wrist: 1.26 (1.1–1.5) H2RAs Total: 1.08 (1.0–1.1) Hip: 1.07 (0.9–1.3) Spine: 1.02 (0.9–1.2) Forearm or wrist: 1.05 (0.9–1.2) Age, race/ethnicity, BMI, enrollment in clinical trial status, indicator for cohort, smoking, physical activity, self-reported health, having a parent who broke a hip after age 40 y, treated diabetes mellitus, history of fracture at 55 y or older, corticosteroid use, physical function score, history of MI or angina, asthma or emphysema, arthritis, stomach or duodenal ulcer, moderate or severe heart-burn, osteoporosis, number of psychoactive medications, use of HRT and bisphosphonates na Corley et al,50 2010 USA; NCC; Kaiser Permanente, northern California 1995–2007 Outcome: hip/femur fracture Ascertainment: electronic coding Agent: PPIs and H2RAs Comparison: user vs nonuser PPIs Hip: 1.30 (1.21–1.39) H2RAs Hip: 1.18 (1.08–1.29) Age, sex, smoking, alcohol abuse, diabetes, arthritis, kidney disease, ethnicity, medications Exposed: PPIs: 1,558/4,806 H2RAs: 875/3,061 Unexposed: PPIs: 32,194/125,665 H2RAs: 32,877/127,410 Chiu et al,53 2010 Taiwan; PCC; patients aged ≥50 y 2005–2006 Outcome: hip fracture Ascertainment: ICD code for National Health Insurance Agent: PPIs Comparison: user vs nonuser 2.11 (1.45–3.07) Age, sex, hypertension, diabetes, stroke, asthma, COPD, CHF, MI, dementia, depression, CRF, NSAID, corticosteroids, anticoagulants, diuretics, antipsychotics, thyroxine, HRT, statins, antihypertensive medications, sedatives, bisphosphonates na Pouwels et al,52 2011 Netherlands; PCC 1991–2002 Outcome: hip/femur fracture Ascertainment: ICD code Agent: PPIs and H2RAs Comparison: current use vs never use PPIs Hip: 1.20 (1.04–1.4) H2RAs Hip: 1.19 (1.00–1.4) Use of other antacids, average daily dose of oral corticosteroids, anxiolytics/hypnotics, short- or long-acting benzodiazepines, HRT, anticonvulsants, antipsychotics, antidepressants, β-blockers, antidiabetics, ≥2 NSAIDs, DMARDs, a history of digestive system disorders, anemia, mental disorders, cerebrovascular disease, CHF, endocrine disorders, RA, diabetes mellitus, COPD and inflammatory bowel disease. Furthermore, the PPI analysis was adjusted for the use of H2RAs and the H2RA analysis for the use of PPIs Exposed: PPIs: 305/773 H2RAs: 196/520 Unexposed: PPIs: 5,810/23,430 H2RAs: 5,624/22,545 - Table 2.
PPI or H2RA Use and the Risk of Fracture in Subgroup Meta-analyses Using a Random-Effects Model
Factor No. of Studies Summary OR/RR (95% CI) Heterogeneity I2, % CI=confidence interval; H2RA = histamine 2 receptor antagonist; na = not available; NOS = Newcastle-Ottawa Scale; OR = odds ratio; PPI = proton pump inhibitor; RR=relative risk. a Of the following variables: age, sex, body mass index, alcohol, smoking, medications that might have affected the risk of osteoporosis or fracture, physical activity or exercise, prior fracture or low-trauma fracture, calcium supplement or antiosteoporotic medications, hormone replacement therapy, history of stomach surgery or digestive system disorders, osteoporosis, family history of osteoporotic fracture. Type of study PPIs 10 1.29 (1.18–1.41) 69.8 Case-control 4 1.40 (1.13–1.75) 73.7 Nested case-control 3 1.23 (1.03–1.47) 85.8 Cohort 3 1.29 (1.13–1.47) 38.7 H2RAs 7 1.10 (0.99–1.23) 86.3 Case-control 3 1.11 (0.81–1.51) 85.6 Nested case-control 2 1.20 (1.13–1.28) 0 Cohort 2 1.08 (1.02–1.13) 0 Methodologic quality (NOS score) PPIs 10 1.29 (1.18–1.41) 69.8 High (>7 points) 5 1.32 (1.18–1.47) 63.7 Low (≤7 points) 5 1.25 (1.06–1.48) 78.7 H2RAs 7 1.10 (0.99–1.23) 86.3 High (>7 points) 3 1.13 (1.05–1.21) 40.3 Low (≤7 points) 4 1.09 (0.87–1.38) 90.6 No. of adjustment variablesa PPIs 10 1.29 (1.18–1.41) 69.8 ≥5 6 1.36 (1.24–1.49) 63.0 <5 4 1.16 (0.98–1.38) 66.5 H2RAs 7 1.10 (0.99–1.23) 86.3 ≥5 4 1.14 (1.06–1.22) 57.7 <5 3 1.11 (0.81–1.51) 85.6 Outcome PPIs Hip fracture 9 1.31 (1.11–1.54) 88.4 Vertebral fracture 3 1.56 (1.31–1.85) 6.3 Other fracture 3 1.15 (0.94–1.42) 78.2 H2RAs Hip fracture 7 1.11 (0.95–1.29) 81.2 Other fracture 3 1.01 (0.93–1.11) 0 Long-term use PPIs Any fracture 10 1.30 (1.15–1.48) 87.4 Hip fracture 9 1.34 (1.09–1.66) 87.4 H2RAs Any fracture 6 1.11 (0.95–1.30) 88.9 Hip fracture 6 1.14 (0.90–1.43) 87.2 Dose (hip fracture) PPIs High dose 4 1.53 (1.18–1.97) 81.6 Usual dose 4 1.42 (1.31–1.53) 32.6 H2RAs High dose 4 1.14 (0.84–1.54) 92.3 Usual dose 4 1.05 (0.89–1.26) 82.2 Sex PPIs Male (hip fracture) 4 1.24 (0.93–1.65) 89.9 Female (hip or vertebral fracture) 5 1.21 (0.98–1.50) 88.8 H2RAs Male (hip fracture) 2 1.60 (0.92–2.77) 0 Selection Comparability Exposure Study Adequate Definition of Cases Representativeness of Cases Selection of Controls Definition of Controls Control for Important Factor or Additional Factor Ascertain ment of Exposure (Blinding) Same Method of Ascertainment for Subjects Nonresponse Rate Total Scorea a Possible total scores range from 0 (lowest quality) to 9 (highest quality). Studies included in the meta-analysis were defined as high-quality if their total score was greater than 7. Case-control (n=8) Grisso et al,49 1997 1 0 0 0 2 0 1 0 4 Vestergaard et al,39 2006 0 1 1 0 2 0 1 1 6 Yang et al,22 2006 0 1 1 0 2 0 1 1 6 Kaye and Jick,42 2008 0 1 1 1 2 0 1 1 7 Targownik et al,23 2008 0 1 1 1 2 0 1 1 7 Corley et al,50 2010 1 1 1 1 2 1 1 1 9 Chiu et al,53 2010 1 1 1 1 2 1 1 1 9 Pouwels et al,52 2011 0 1 1 1 2 1 1 1 8 Cohort (n=3) Yu et al,40 2008 1 1 1 0 2 1 1 0 7 Roux et al,41 2009 1 1 1 1 2 1 1 0 8 Gray et al,51 2010 1 1 1 1 2 1 1 0 8
The Article in Brief
Use of Acid Suppressive Drugs and Risk of Fracture: A Meta-Analysis of Observational Studies
Chun-Sick Eom , and colleagues
Background Acid suppression medications, including proton pump inhibitors (PPIs) and H2-receptor antagonists (H2RAs), are among the most widely used drugs in the world. Previous research has reported inconsistent findings about an association between use of such drugs and fracture risk. This study analyzed existing research to investigate the association between the use of PPIs and H2RAs and the risk of fracture.
What This Study Found Long-term use of proton pump inhibitors, but not H2-receptor antagonists, is associated with increased risk of fracture. The meta-analysis of 11 studies finds that PPIs are associated with a 29 percent increase of fracture, including 31 percent increased risk of hip fracture and a 54 percent increased risk of vertebral fractures. Long-term H2RA use, by contrast, was not significantly associated with fracture risk.
Implications
- Given the widespread use of acid-suppressive drugs, these findings have important implications for public health.
- The authors conclude that clinicians should carefully consider their decision to prescribe PPIs for patients, especially those who already have an elevated risk of fracture because of age or other factors.
