Use of ²-Blockers and Risk of Fractures
Total Page：16
File Type：pdf, Size：1020Kb
ORIGINAL CONTRIBUTION Use of -Blockers and Risk of Fractures Raymond G. Schlienger, PhD, MPH Context Animal studies suggest that the -blocker propranolol increases bone for- Marius E. Kraenzlin, MD mation, but data on whether use of -blockers (with or without concomitant use of thiazide diuretics) is associated with reduced fracture risk in humans are limited. Susan S. Jick, DSc Objective To determine whether use of -blockers alone or in combination with thia- Christoph R. Meier, PhD, MSc zides is associated with a decreased risk of fracture in adults. Design, Setting, and Participants Case-control analysis using the UK General Prac- TUDIES HAVE SUGGESTED THAT tice Research Database (GPRD). The study included 30601 case patients aged 30 to the sympathetic nervous sys- 79 years with an incident fracture diagnosis between 1993 and 1999 and 120819 con- tem has a catabolic effect on trols, matched to cases on age, sex, calendar time, and general practice attended. bones.1-4 In vitro data show that Main Outcome Measures Odds ratios (ORs) of having a fracture in association Sadrenergic agonists stimulate bone re- with use of -blockers or a combination of -blockers with thiazides. sorption in organ culture of mouse cal- Results The most frequent fractures were of the hand/lower arm (n=12837 [42.0%]) variae.4 Chemical sympathectomy with and of the foot (n=4627 [15.1%]). Compared with patients who did not use either guanethidine, a sympathetic neuro- -blockers or thiazide diuretics, the OR for current use of -blockers only (Ն3 pre- toxic agent, impairs bone resorption by scriptions) was 0.77 (95% confidence interval [CI], 0.72-0.83); for current use of thia- inhibiting preosteoclast differentiation zides only (Ն3 prescriptions), 0.80 (95% CI, 0.74-0.86); and for combined current  and disturbing osteoclast activation in use of -blockers and thiazides, 0.71 (95% CI, 0.64-0.79). Data were adjusted for adult rats.1 Additionally, the -blocker smoking; body mass index; number of practice visits; and use of calcium channel block- ers, angiotensin-converting enzyme inhibitors, antipsychotics, antidepressants, stat- propranolol increased bone formation in ins, antiepileptics, benzodiazepines, corticosteroids, and estrogens. ovariectomized female rats.3 These re-  sults suggest that -blockers may over- Conclusions Our data suggest that current use of -blockers is associated with a reduced risk of fractures, taken alone as well as in combination with thiazide diuretics. come loss of bone mass in postmeno- 5 Many elderly patients with hypertension who are at risk of developing osteoporosis pausal women. Indeed, in a recent may potentially benefit from combined therapy with -blockers and thiazides. observational study including 569 frac- JAMA. 2004;292:1326-1332 www.jama.com ture case and 775 control patients, use of -blockers was associated with a higher bone mineral density at the hip tween -blocker use—with or with- data, medical diagnoses, details of hos- and forearm in women aged 50 years or out concomitant use of thiazides— pital stays, and deaths using a stan- older, and -blocker use was associated and fracture risk in men and women dard, anonymous form. The physi- with a 30% decrease in fracture risk.6 aged 30 to 79 years. cians generate prescriptions directly Thiazide diuretics are thought to be with the computer; this information is protective against bone loss by reduc- METHODS automatically transcribed into the com- ing urinary calcium excretion. Several Study Population and Data Source epidemiologic studies suggest that ex- Data were derived from the UK-based Author Affiliations: Basel Pharmacoepidemiology Unit, posure to thiazides is associated with General Practice Research Database Division of Clinical Pharmacology and Toxicology (Drs 7-11 Schlienger and Meier) and Division of Endocrinol- a reduced fracture risk. (GPRD), which has been described in ogy, Diabetes and Clinical Nutrition (Dr Kraenzlin), Uni- 12,13 To our knowledge, no studies have detail elsewhere. More than 3 mil- versity Hospital, and Institute of Clinical Pharmacy, De- evaluated the association of -blocker lion people in the United Kingdom are partment of Pharmaceutical Sciences, University of Basel (Drs Schlienger and Meier), Basel, Switzerland; use and bone fractures in men and enrolled with selected general practi- and Boston Collaborative Drug Surveillance Pro- young women or the association of tioners who use office computers and gram, Boston University, School of Medicine, Lexing-  ton, Mass (Drs Jick and Meier). combined use of -blockers with thia- have agreed to provide data for re- Corresponding Author: Christoph R. Meier, PhD, MSc, zide diuretics with the risk of frac- search purposes. General practition- Basel Pharmacoepidemiology Unit, Division of Clini- cal Pharmacology and Toxicology, University Hospi- tures. In this large population-based ers have been trained to record medi- tal Basel, Hebelstrasse 2, CH-4031 Basel, Switzer- study we explored the association be- cal information including demographic land ([email protected]). 1326 JAMA, September 15, 2004—Vol 292, No. 11 (Reprinted) ©2004 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/24/2021 −BLOCKERS AND RISK OF FRACTURE puter record. A modification of the Ox- ism) as well as patients who used bis- ated mutually exclusive categories of us- ford Medical Information System clas- phosphonates (as an indicator of ers of -blockers only, thiazides only, sification (similar to the International osteoporosis or bone metastases) prior or a combination thereof. We ad- Classification of Diseases, Eighth Revi- to the index date were excluded from justed this analysis for use of other an- sion) is used to enter medical diag- the analysis. tihypertensive drugs as well as for a va- noses, and a coded drug dictionary riety of other drugs potentially affecting based on the UK Prescription Pricing Controls fracture risk. Users of both -blockers Authority dictionary is used for record- Controls were persons without a frac- and thiazides were individuals who ing prescriptions. The recorded infor- ture selected from the base popula- either used a fixed combination or had mation on drug exposure and diag- tion. We randomly selected up to 4 con- separate prescriptions for both -block- noses has been validated and proven to trols per case matched on age (±2 ers and thiazide diuretics. They were be of high quality.14,15 years), sex, general practice attended, classified as current users if both The GPRD has been the source of calendar time (by using the same in- -blockers and thiazides were pre- several observational studies, includ- dex date as for cases), and years of prior scribed within 60 days prior to the in- ing research on antihypertensive history in the GPRD (matched on the dex date. If they currently used only a drugs16,17 as well as fractures.18 Based year when the individual’s record in the -blocker and previously used thia- on previous record review by our- database began, ±1 year). Further- zides, or vice versa, they were catego- selves18 and by others,19 the validity of more, controls had to be alive on the rized in a separate group called “other fracture diagnoses in the GPRD is high, index date. The same exclusion crite- combinations.” with a confirmed proportion of at least ria were applied to controls as to case 90% after comparing computer- patients. Statistical Analysis recorded diagnoses with hospital dis- We conducted conditional logistic re- charge letters and/or questionnaire in- Exposure Assessment gression analyses using SAS version 8.1 formation provided by general We ascertained exposure to -blockers (SAS Institute Inc, Cary, NC). Risk es- practitioners. The study was ap- (ie, acebutolol, atenolol, betaxolol, bi- timates are presented as odds ratios proved by the Scientific and Ethical Ad- soprolol, carvedilol, celiprolol, labeta- (ORs) with 95% confidence intervals visory Group of the GPRD. lol, metoprolol, nadolol, nebivolol, ox- (CIs). P values are 2-sided and were prenolol, pindolol, propranolol, sotalol, considered statistically significant if less Case Definition and Validation timolol) and thiazide or thiazide-like di- than .05. We identified all patients with an inci- uretics (hydrochlorothiazide, chlortha- Odds ratios were adjusted for the po- dent diagnosis of a fracture recorded in lidone, bendrofluazide, benzthiazide, tential confounders age, sex, practice, the computerized medical record be- chlorothiazide, cyclopenthiazide, poly- calendar time, and years of recorded his- tween January 1993 and December thiazide, mefruside, xipamide, metola- tory in the GPRD prior to the index date 1999. More recent data were not in- zone, hydroflumethiazide, clopamide, (by matching), as well as smoking sta- cluded because of coding changes in the indapamide) for all cases and controls tus (none, current, past, unknown), database. Cases were identified with- prior to the index date. Exposure to body mass index (BMI; Ͻ25, 25-29.9, out any exposure information. Poten- study drugs as well as for a variety of Ն30 measured as weight in kilograms tial cases had to be aged 30 to 79 years other drugs was assessed as follows. Pa- divided by the square of height in at the date of the fracture (ie, index tients were classified as current users if meters), number of practice visits prior date), and they had to have been re- their last prescription for a study drug to the index date (Ͻ5, 5-9, 10-29, 30- corded in the database for