BMJ

Confidential: For Review Only

A matche d cohort study to evaluate the risk of fall and fracture with the initiation of a prostate-selective alpha antagonist

Journal: BMJ

Manuscript ID BMJ.2015.028205.R1

Article Type: Research

BMJ Journal: BMJ

Date Submitted by the Author: 22-Sep-2015

Complete List of Authors: Welk, Blayne; Western University, McArthur, Eric; Institute for Clinical Evaluative Sciences,, Fraser, Lisa-Ann; Western University, Medicine Hayward, Jade; Institute for Clinical Evaluative Sciences,, Dixon, Stephanie; Institute for Clinical Evaluative Sciences,, Hwang, Joseph; Case Western Reserve University School of Medicine, Ordon, Michael; University of Toronto, Surgery (Urology)

Keywords: BPH, Fall, Fracture, Alpha antgonist

https://mc.manuscriptcentral.com/bmj Page 1 of 44 BMJ

1 2 The risk of fall and fracture with the initiation of a prostate-selective alpha antagonist: a 3 population based cohort study 4 5 6 Blayne Welk MD MSc 1,2,3 , Eric McArthur MSc 2, Lisa-Ann Fraser MD MSc 4, Jade Hayward 2, 7 Stephanie Dixon MSc PhD 2,3 , Y. Joseph Hwang MSc 5, Michael Ordon MD MSc 6 8 Confidential: For Review Only 9 1 Department of Surgery, Western University, London, Ontario, Canada 10 11 2 Institute for Clinical Evaluative Sciences, London, Ontario, Canada 12 3 Department of Epidemiology and Biostatistics, Western University, London, Ontario, Canada 13 4 Department of Medicine, Western University, London, Ontario 14 5 MD Candidate, Case Western Reserve University School of Medicine, Cleveland, Ohio, USA 15 6 Division of Urology, Department of Surgery, University of Toronto, Toronto, Ontario, Canada 16 17 18 Corresponding Author 19 Blayne Welk, MD MSc 20 Assistant Professor, Department of Surgery and Epidemiology and Biostatistics (Urologist), 21 Western University 22 23 Room B4-667 24 St Joseph's Health Care 25 268 Grosvenor Street London ON N6A 4V2 26 Telephone: 519 646-6367 | Fax: 519 646-6037 27 [email protected] 28 29 30 Eric McArthur, Research Analyst, Institute for Clinical Evaluative Sciences 31 Lisa-Ann Fraser, Assistant Professor of Medicine (Endocrinology), Western University 32 Jade Hayward, Research Assistant, Institute for Clinical Evaluative Sciences 33 34 Stephanie Dixon, Epidemiologist, Institute for Clinical Evaluative Sciences 35 Joseph Hwang, Medical Student, Case Western Reserve University School of Medicine 36 Michael Ordon, Assistant Professor of Surgery (Urologist), University 37 38 Addresses: 39 40 Mr McArthur: [email protected] 41 LHSC – Victoria Hospital 42 ELL-101, 800 Commissioners Rd. E. 43 London, Ontario N6A 4G5 44 519-685-8500 ext. 56555 45 46 47 Dr Fraser: [email protected] 48 St Joseph's Health Care 49 268 Grosvenor St 50 London, Ontario N6A 4V2 51 52 519-646-6245 53 54 Ms Hayward: [email protected] 55 LHSC – Victoria Hospital 56 ELL-101, 800 Commissioners Rd. E. 57 58 London, Ontario N6A 4G5 59 519-685-8500 ext. 56555 60 1 https://mc.manuscriptcentral.com/bmj BMJ Page 2 of 44

1 2 3 Dr Dixon: [email protected] 4 5 LHSC – Victoria Hospital 6 ELL-101, 800 Commissioners Rd. E. 7 London, Ontario N6A 4G5 8 519-685-8500Confidential: ext. 56555 For Review Only 9 10 11 Mr Hwang: [email protected] 12 603-105 La Rose Ave 13 Etobicoke ON M9P 1A9 14 15 Dr Ordon: [email protected] 16 17 St Michaels Hospital 18 30 Bond St, Toronto, ON M5B 1W8 19 (416) 360-4000 20 21 The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of 22 23 all authors, a worldwide licence to the Publishers and its licensees in perpetuity, in all forms, 24 formats and media (whether known now or created in the future), to i) publish, reproduce, 25 distribute, display and store the Contribution, ii) translate the Contribution into other languages, 26 create adaptations, reprints, include within collections and create summaries, extracts and/or, 27 abstracts of the Contribution, iii) create any other derivative work(s) based on the Contribution, iv) 28 29 to exploit all subsidiary rights in the Contribution, v) the inclusion of electronic links from the 30 Contribution to third party material where-ever it may be located; and, vi) license any third party to 31 do any or all of the above. 32 33 34 Word count: 3043 35 Abstract 343 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 https://mc.manuscriptcentral.com/bmj Page 3 of 44 BMJ

1 2 Abstract 3 4 5 6 7 Importance: Oral alpha antagonists are commonly used in older men to treat urinary symptoms. 8 Confidential: For Review Only 9 One of the side effects of these medications is hypotension, the consequence of which may be 10 11 12 serious falls or fractures. 13 14 15 16 17 Objectives: To determine the risk of fall and fracture among men initiating prostate-specific alpha 18 19 antagonist medication. 20 21 22 23 24 Design: Population-based, retrospective, matched cohort study 25 26 27 28 29 Setting: Ontario, Canada from 2003 through 2013. 30 31 32 33 34 Participants: Population based sample of men ≥66 years of age that were (or were not) newly 35 36 started on an alpha antagonist. 37 38 39 40 41 Exposures: Prostate-specific alpha antagonists: tamsulosin, alfuzosin, and silodosin. 42 43 44 45 46 Main Outcome Measure: The primary outcome was a hospital emergency room visit or inpatient 47 48 admission for a fall or fracture. We hypothesized a priori that prostate-specific alpha antagonist 49 50 exposure would increase the risk of these events. 51 52 53 54 55 Results: We identified 147,084 men who initiated prostate-specific alpha antagonist therapy. 56 57 58 These men were matched 1:1 (using a propensity score model) to an equal number of men who 59 60 3 https://mc.manuscriptcentral.com/bmj BMJ Page 4 of 44

1 2 did not initiate alpha antagonist therapy. These groups were similar across 98 different 3 4 5 covariates including demographics, comorbid conditions, medication usage, healthcare 6 7 utilization and prior medical investigation. The relative proportion of alpha antagonist use was 8 Confidential: For Review Only 9 tamsulosin ( n=123,537, 84%), alfuzosin (n=20,827, 14%) and silodosin (n=2,720, 2%). We found 10 11 12 that men initiating a prostate-specific alpha antagonist had a significantly increased risk of both 13 14 falling (OR 1.14, 95% CI 1.07 to 1.21, absolute risk increase 0.17%, 95% CI 0.08 to 0.25%) and of 15 16 17 sustaining a fracture (OR 1.16, 95% CI 1.04 to 1.29, absolute risk increase 0.06%, 95% CI 0.02 to 18 19 0.11%). This increased risk was not observed in the time period prior to alpha antagonist 20 21 initiation. Secondary outcomes of hypotension and head trauma were also significantly increased 22 23 24 among men exposed to prostate-specific alpha antagonists (OR 1.80, 95% CI 1.59 to 2.03 and OR 25 26 1.15, 95% CI 1.04 to 1.27 respectively). 27 28 29 30 31 Conclusions and Relevance: Prostate-specific alpha antagonists are associated with a small, but 32 33 34 significantly increased risk of fall, fracture, and head trauma, likely as a result of medication- 35 36 induced hypotension. Caution should continue to be exercised with the initiation of these 37 38 medications. 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4 https://mc.manuscriptcentral.com/bmj Page 5 of 44 BMJ

1 2 Introduction 3 4 5 6 7 As men age, a significant number will develop bothersome lower urinary tract symptoms, such as 8 Confidential: For Review Only 9 urinary frequency, urgency, nocturia and a weak urinary stream. These symptoms are often 10 11 12 attributed to benign prostatic hyperplasia (BPH) and are frequently treated due to their negative 13 14 impact on quality of life[1]. The introduction of several alpha antagonist medications in the 15 16 17 1990’s fundamentally changed the treatment of BPH and lower urinary tract symptoms[2,3]. 18 19 These medications relax the prostatic smooth muscle, and improve urinary flow rates and 20 21 symptoms[1,2]. 22 23 24 25 26 Nonspecific (first generation) alpha antagonists, such as doxazosin and terazosin, act on all ααα1 27 28 29 receptor subtypes (ααα1A, B, D ), and are used for both the treatment of BPH and for hypertension. 30 31 32 This nonselective ααα1 receptor activity requires dose titration over 1-2 weeks to avoid 33 34 symptomatic hypotension from ααα1B antagonism. These nonspecific alpha antagonists have 35 36 37 generally fallen into disfavor as antihypertensives since the ALLHAT trial[4]. In this randomized, 38 39 double-blind trial, traditional antihypertensive medicines (including nonspecific alpha 40 41 42 antagonists) were compared with newer classes of antihypertensives; the doxazosin arm was 43 44 stopped early due to a higher rate of cardiac dysfunction . Similarly, among men with BPH 45 46 47 symptoms these nonspecific alpha antagonists have largely been replaced by prostate-specific 48 49 (second generation) alpha antagonists[5]. These newer medications do not need dose titration 50 51 and are solely indicated for male urinary symptoms from BPH. They include alfuzosin (which 52 53 54 exhibits clinical uroselectivity), and tamsulosin and silodosin (which exhibit true ααα1A receptor 55 56 57 selectivity, which is the primary alpha receptor in the prostate)[6-9]. Their selectivity for the 58 59 prostate hypothetically reduces the risk of hypotension due to reduced ααα1B antagonism. 60 5 https://mc.manuscriptcentral.com/bmj BMJ Page 6 of 44

1 2 However, as a class, all alpha antagonists have been associated with side effects such as 3 4 5 hypotension and dizziness[6,10,11]. For this reason, alpha antagonists in general have been 6 7 identified as a class of medication to discontinue in a patient who falls[12]. However, this 8 Confidential: For Review Only 9 recommendation is primarily based on perceptions of the hypotension risk associated with 10 11 12 nonspecific alpha antagonists, and the potentially lower risk, prostate-specific medications have 13 14 not been widely studied. In addition, previous research has failed to demonstrate a consistent 15 16 17 conclusion regarding the risk of falls and fractures with alpha antagonists used for BPH 18 19 symptoms. Studies have both demonstrated[11,13,14] and failed to demonstrate[15,16] this risk. 20 21 The effect of BPH treatment on falls and fractures has been identified as an important research 22 23 24 question[1,14] because of the direct morbidity associated with falls in the elderly (such as 25 26 fracture and head injury), the frequent need for nursing home admission after a fall, and the 27 28 29 substantial direct medical costs[12]. The objective of this study was to use administrative data to 30 31 assess the 90-day risk of fall or fracture among men initiating a prostate-specific alpha 32 33 34 antagonist. 35 36 37 38 Methods 39 40 41 42 43 Design and setting 44 45 46 47 48 We conducted a retrospective, matched-cohort study using datasets from the province of Ontario 49 50 (Canada) that were linked using unique, encoded identifiers and analyzed at the Institute for 51 52 53 Clinical Evaluative Sciences. Ontario has a population of over 13 million people with universal 54 55 access to a single public health care system, and universal drug coverage for those over 65 years 56 57 58 of age. This study was approved (2015 0906 137 000) by the research ethics board at 59 60 6 https://mc.manuscriptcentral.com/bmj Page 7 of 44 BMJ

1 2 Sunnybrook Hospital (Toronto, Ontario), and individual patient consent was not required. No 3 4 5 patients were involved in setting the research question or the outcome measures, nor were they 6 7 involved in the design and implementation of the study. There are no plans to involve patients in 8 Confidential: For Review Only 9 dissemination. Study reporting follows the STROBE guidelines (eTable 1 in the supplement). 10 11 12 13 14 Data sources 15 16 17 18 19 The following administrative data sources have previously demonstrated validity and reliability 20 21 and were used for this study: 1) Registered Persons Database (containing vital statistics)[17], 2) 22 23 24 Ontario drug benefit database (containing all prescription drug use for seniors)[18], 3) Canadian 25 26 Institute for Health Information Discharge Abstract Database (identifying all inpatient 27 28 29 admissions and procedures)[19], 4) National Ambulatory Care Reporting System (identifying all 30 31 emergency room encounters)[20], and 5) Ontario Health Insurance Plan (containing all physician 32 33 34 billing codes for patient assessment or treatment)[21]. These databases contain patient level data 35 36 in linkable files without any direct patient identifiers, and are >99.5% complete for all study 37 38 variables. 39 40 41 42 43 Patient population 44 45 46 47 48 We established a cohort of men 66 years or older who filled their first outpatient prescription for 49 50 tamsulosin, alfuzosin, or silodosin between June 1, 2003 and December 31, 2013 (exposed men). 51 52 53 The date the prescription was filled served as the index date for the patient, and the beginning of 54 55 the 90-day observation window for our outcome. All remaining men in the province of Ontario 56 57 58 (unexposed men) were assigned a random index date based on the distribution of index dates for 59 60 7 https://mc.manuscriptcentral.com/bmj BMJ Page 8 of 44

1 2 men with alpha antagonist use. We excluded men who had used a nonspecific or prostate-specific 3 4 5 alpha antagonist in the 3 months prior to the index date, and those who were in hospital or had 6 7 an emergency room visit in the 2 days prior to the index date (eFigure 1 in the supplement). Men 8 Confidential: For Review Only 9 from the exposed cohort were then matched to men from the unexposed cohort based on age 10 11 12 (within ±2 years), residential status (community dwelling versus long-term care residents), prior 13 14 fracture status, 5-alpha reductase inhibitor usage, and the logit of a propensity score. We 15 16 17 assessed 98 different covariates representing comorbid medical conditions, medication usage, 18 19 and healthcare utilization to assure the similarity of our exposed and unexposed cohorts (coding 20 21 definitions are included in eTable 2a-c in the supplement). A 5 year look back window from the 22 23 24 index date was used for all specific medical comorbidities, and a 1 year look back window was 25 26 used for prior falls and fractures. Prior medications were assessed based on any prescription 27 28 29 within the 6 months prior to the index date, and prior hospital/physician utilization and medical 30 31 investigations and procedures were assessed based on the previous year. Of the 98 covariates, 42 32 33 34 were selected for inclusion in the propensity score (listed in eTable 3 in the supplement) based 35 36 on a standard difference of >7%[22] or a potentially important relationship with our study 37 38 outcomes. 39 40 41 42 43 Study outcomes 44 45 46 47 48 We pre-specified falls and fractures as our primary outcomes. Our secondary outcomes included 49 50 major osteoporotic fractures, hip fractures, hypotension, and head trauma. The data sources and 51 52 53 validated diagnostic codes that were used, as well as their measurement characteristics are 54 55 outlined in eTable 4 in the supplement. 56 57 58 59 60 8 https://mc.manuscriptcentral.com/bmj Page 9 of 44 BMJ

1 2 We observed patients for our study outcomes for 90 days after the index date. This time period 3 4 5 was chosen as adverse events are highest in the first few weeks after starting an alpha 6 7 antagonist[6,7], long term patient adherence to alpha blocker therapy is modest[23], and this 8 Confidential: For Review Only 9 time period minimizes crossover between the groups. 10 11 12 13 14 Statistical analysis 15 16 17 18 19 Baseline characteristics were compared using standardized differences (SD), which is better at 20 21 identifying meaningful differences between very large groups than traditional hypothesis tests; a 22 23 24 SD >10% was considered an important difference[22]. To create our two cohorts, we used greedy 25 26 matching (1:1) with a caliper of ±0.2 standard deviations of the logit of the propensity score[24]. 27 28 29 30 31 Our primary analysis was carried out using PROC LOGISTIC (SAS 9.4, SAS institute, Cary, NC, 32 33 34 USA). Conditional logistic regression was used to account for matching. Conditional odds ratios 35 36 (OR) and 95% confidence intervals are reported, as well as absolute risk. Secondary outcomes 37 38 were assessed using similar methodology. Subgroup analysis was carried out among men 39 40 41 stratified by age (younger or older than 75 years of age), type of alpha antagonist, residential 42 43 status, and 5-alpha reductase use; the significance of these subgroups was assessed using tests 44 45 46 for interactions. We considered two-tailed p-values <0.05 statistically significant. 47 48 49 50 Men with lower urinary tract symptoms from BPH have an increased risk of falls due to their 51 52 53 urinary symptoms[25]. Therefore, to further assess the temporal relationship between alpha 54 55 antagonist initiation and our outcomes, we wished to demonstrate that men in our exposed 56 57 58 cohort did not have an increased risk prior to the index date (which if observed would suggest 59 60 9 https://mc.manuscriptcentral.com/bmj BMJ Page 10 of 44

1 2 that persistence of baseline lower urinary tract symptoms among our exposed cohort may 3 4 5 account for an observed increase in our primary outcomes). We reapplied the exclusion criteria 6 7 to our cohorts on the index date minus 180 days, and eligible matched pairs were retained and 8 Confidential: For Review Only 9 used to re-calculate the 90-day primary outcomes based on this earlier index date. 10 11 12 13 14 Results 15 16 17 18 19 Cohort selection is presented in eFigure 1 of the supplement. We retained a total of 147,084 20 21 matched pairs between our exposed and unexposed cohorts. Complete baseline characteristics of 22 23 24 the unmatched and matched cohorts are presented in eTable 5 in the supplement. The median 25 26 age was 75 (IQR 70-80) years, and groups were well balanced, with no standardized difference 27 28 29 >6% among our 98 measured covariates. These covariates covered demographics (such as region 30 31 of residence and socioeconomic status), comorbid conditions (such as ocular disease, diabetes, 32 33 34 Parkinson’s disease, stroke and seizure), John Hopkins ACG comorbidity adjustment, medication 35 36 usage (including antihypertensives, antipsychotics, antidepressants and anticonvulsants), 37 38 healthcare utilization, and prior medical investigation. The most frequently used prostate- 39 40 41 specific alpha antagonist in our study was tamsulosin (n=123,537, 84%), followed by alfuzosin 42 43 (n=20,827, 14%) and silodosin (n=2,720, 2%). Dosage and prescription details are listed in 44 45 46 eTable 6 in the supplement. These medications were prescribed predominantly by a primary care 47 48 physician (54%) or a urologist (27%). 49 50 51 52 53 The primary outcomes of 90-day hospital visit for fall or fracture, as determined from hospital 54 55 and emergency room admissions, are listed in table 1. Men initiating a prostate-specific alpha 56 57 58 antagonist had a significantly increased risk of both presenting to hospital after a fall (OR 1.14, 59 60 10 https://mc.manuscriptcentral.com/bmj Page 11 of 44 BMJ

1 2 95% CI 1.07 to 1.21) and sustaining a new fracture (1.16, 95% CI 1.04 to 1.29). These two 3 4 5 outcomes appear to be inter-related, as 80% of men in who sustained a fracture were also 6 7 identified as having a fall. Our secondary outcomes included specific fracture types. In general, 8 Confidential: For Review Only 9 there were few major osteoporotic fractures or hip fractures during our 90-day time frame, and 10 11 12 there was no significantly increased risk for these specific fractures among men starting 13 14 prostate-specific alpha antagonists. Other secondary outcomes included hospital admission or 15 16 17 emergency room visit for hypotension or head trauma, both of which were significantly increased 18 19 among men exposed to alpha antagonists (OR 1.80, 95% CI 1.59 to 2.03 and OR 1.15, 95% CI 1.04 20 21 to 1.27 respectively). 22 23 24 25 26 We used the most common primary outcome, hospital or emergency room visit for a fall, to 27 28 29 examine several subgroups within our study population (Table 2). Men older than 75 years of 30 31 age, those in long-term care, and those with prior fractures all had a higher baseline level of falls, 32 33 34 however these factors did not significantly modify the effect of alpha antagonists on falls (p=0.52, 35 36 p=0.25 and p=0.39 respectively for interaction). We found that prescribing tamsulosin 0.4mg, 37 38 (the most commonly prescribed prostate-specific alpha antagonist in our cohort), at a higher (off 39 40 41 label) dose of 0.8mg did not significantly modify the risk of a fall (p=0.49). 5-alpha reductase 42 43 inhibitors are the only other class of medications that are indicated specifically for the treatment 44 45 46 of BPH, and similar to alpha antagonists, they are efficacious at managing lower urinary tract 47 48 symptoms[1]. The presence of a prescription for one of these medications in the 180 days prior 49 50 to starting a prostate alpha antagonist suggests an existing diagnosis of BPH. There was a 51 52 53 significant interaction (p=0.02) among men with prior 5-alpha reductase inhibitor treatment and 54 55 new alpha antagonist usage, with those who had evidence of 5-alpha reductase inhibitor use 56 57 58 having a non-significant trend towards a lower risk of falls (OR 0.88, 95% CI 0.70 to 1.10). 59 60 11 https://mc.manuscriptcentral.com/bmj BMJ Page 12 of 44

1 2 3 4 5 To assess the temporal relationship between alpha antagonist initiation and our primary 6 7 outcomes, we repeated our analysis among eligible matched men 180 days prior to the first 8 Confidential: For Review Only 9 initiation of alpha antagonists. We demonstrated that men who were to be exposed to a prostate- 10 11 12 specific alpha antagonists in the future did not have a significantly higher risk of fall or fracture 13 14 (OR 0.92 95% CI 0.86 to 0.99 and OR 0.86, 95% CI 0.76 to 0.97 respectively) at this earlier time 15 16 17 point compared to men in the unexposed cohort (eTable 7 in the supplement). 18 19 20 21 Discussion 22 23 24 25 26 Principle findings 27 28 29 We observed a small, but significant increase in the risk of falls and fractures among older men 30 31 initiating prostate-specific alpha antagonist medications. These men were also at an increased 32 33 34 risk of hospital admission, or emergency room assessment, for a traumatic head injury, another 35 36 potential serious sequela of a fall. The majority of individuals who sustained a fracture or head 37 38 injury had a concomitant fall, suggesting that the mechanism for these potentially serious injuries 39 40 41 among men initiating alpha antagonist therapy is likely related to syncope or postural 42 43 hypotension, which are known side effects of the alpha antagonist class of medications[9]. This 44 45 46 conclusion is further strengthened by our observation that men starting prostate-specific alpha 47 48 antagonists were significantly more likely to require hospital admission or emergency room 49 50 assessment for hypotension. These outcomes are extremely relevant in an elderly population 51 52 53 which has an estimated 50% mortality rate during the first year after a fall requiring 54 55 hospitalization[26], and an increased mortality risk after fractures[27]. In the elderly, falls are 56 57 58 the most common cause of traumatic brain injury, and these older patients have greater cognitive 59 60 12 https://mc.manuscriptcentral.com/bmj Page 13 of 44 BMJ

1 2 decline and increased mortality compared to younger patients[28]. Finally, although the 3 4 5 magnitude of the risk of fall and fracture is low, the potential for life threatening events or death 6 7 from adverse events due to these medications has been reported to drug regulatory agencies[29]. 8 Confidential: For Review Only 9 Some specific results within our study deserve comment. A history of 5-alpha reductase use in 10 11 12 the 6 months prior to alpha antagonist initiation was a significant effect modifier in our study 13 14 (p=0.02), and trended towards a decrease risk of falls (OR 0.88, 95% CI 0.70 to 1.10). This may be 15 16 17 due to successful treatment of specific urinary tract symptoms (such as nocturia and 18 19 incontinence) known to cause falls[25] or due to manipulation of testosterone levels[30]. In our 20 21 analysis of falls and fractures in the 6 months prior to alpha antagonist initiation, men who were 22 23 24 eventually started on this medication actually had a significantly lower risk of falls and fractures 25 26 compared to the matched cohort who did use alpha antagonists. This may be due to physicians 27 28 29 selectively using alpha antagonists only among men subjectively judged to have a low risk of fall 30 31 or fracture. Nevertheless, the fact that the initiation of a prostate-specific alpha antagonist was 32 33 34 temporally related to a reversal in the direction of this risk further supports our conclusions. 35 36 37 38 Strengths and limitations 39 40 41 To our knowledge this is the first cohort study of prostate-specific alpha antagonists that 42 43 measures both the adverse event of falls and the potentially serious sequela such as fracture and 44 45 46 head injury. Using an extensive library of almost 100 administrative data covariates and 47 48 matching and propensity score techniques we were able to create a large cohort of men with 49 50 similar health characteristics. Compared to randomized trials, large observational studies are 51 52 53 ideally suited for the assessment of rare, but serious adverse drug reactions, and include patients 54 55 generally excluded from randomized trials (thus better assessing real world risk). Weaknesses of 56 57 58 our research include the potential for unmeasured confounders related to increased falls in the 59 60 13 https://mc.manuscriptcentral.com/bmj BMJ Page 14 of 44

1 2 elderly, such as physical deconditioning, mobility impairment, and situational risk factors[12]. 3 4 5 However, it is unlikely such variables would change significantly over the 6 months period we 6 7 used to confirm a temporal increase in fall and fracture risk with alpha antagonist initiation. The 8 Confidential: For Review Only 9 administrative data coding elements we used to identify our primary outcomes had limited 10 11 12 sensitivity. Therefore, while our odds ratios are accurate (as there should not be differential 13 14 misclassification of our outcomes between our exposed and unexposed men), we have likely 15 16 17 underestimated the absolute risk of some of our outcomes. Furthermore, only an estimated 5% of 18 19 falls present to the hospital, so it is likely that we underestimated the true number of less serious 20 21 falls in this population[26]. Our primary exposure was defined as a patient filling a prescription, 22 23 24 however we do not have any measure of medication compliance and found only about 50% of 25 26 men refill their alpha antagonist medication after the first month. This should however bias the 27 28 29 odds ratio towards the null. Finally, our study only included men 66 years of age or older, and 30 31 predominately included tamsulosin. The relatively small number of alfuzosin and silodosin users, 32 33 34 and lack of younger men limits the generalizability of our results to these groups, and we may not 35 36 have had enough statistical power to demonstrate small but meaningful differences in our 37 38 outcomes between these medications. 39 40 41 42 43 Comparisons with other studies 44 45 46 Some previous studies in this area have shown similar results. A population-based cohort study 47 48 demonstrated an increased risk of hypotension requiring hospital admission among men treated 49 50 initially with tamsulosin compared to 5-alpha reductase inhibitors, particularly in the first 4 51 52 53 weeks of treatment[10]. Similarly, a post-marketing observational study of tamsulosin identified 54 55 a 0.7% risk of hypotension, with both dizziness and hypotension commonly reported as reasons 56 57 58 for discontinuing therapy[31]. A case-control study examining the effect of 5-alpha reductase 59 60 14 https://mc.manuscriptcentral.com/bmj Page 15 of 44 BMJ

1 2 inhibitors demonstrated the odds of a hip fracture were two times higher with a recent alpha 3 4 5 antagonist prescription[14]. Two smaller studies also support our conclusions, however they 6 7 included a mix of nonspecific and prostate-specific alpha antagonists, which limits their 8 Confidential: For Review Only 9 generalizability[11,13]. Studies that have failed to show an effect of alpha antagonists on falls or 10 11 12 fractures, include a smaller case-control study of patients using modified-release doxazosin[15], 13 14 and a case-control study[16] which defined alpha antagonist exposure based on the presence of 15 16 17 any prior prescription (not current usage). While placebo-controlled, randomized trials of 18 19 tamsulosin have suggested there are no clinically relevant changes to blood pressure, dizziness 20 21 as an adverse event was generally higher among tamsulosin patients compared to placebo 22 23 24 patients[32,33]. 25 26 27 28 29 Practice implications 30 31 Product monographs for most alpha antagonists have warnings regarding severe hypotension 32 33 34 and syncope, however previous research has suggested that prostate-specific alpha antagonists, 35 36 have a lower risk of hypotension related events compared to nonselective alpha antagonists[9]. 37 38 The current study supports the continued need to counsel men about the risk of hypotension 39 40 41 related adverse events even with prostate-specific alpha antagonist therapy, such as tamsulosin, 42 43 which has several-fold higher affinity for ααα1A receptor in the prostate compared to ααα1B 44 45 46 receptors in the vascular system[7]. Measures to reduce the risk of hypotension-related events 47 48 include taking alpha antagonists in the evening, with food, and avoiding driving or potentially 49 50 51 hazardous activities during the initiation of therapy. 52 53 54 55 Future directions 56 57 58 59 60 15 https://mc.manuscriptcentral.com/bmj BMJ Page 16 of 44

1 2 The impact of 5-alpha reductase inhibitors on fractures, and their potential role in modulating 3 4 5 the risk of falls among men using dual therapy (with an alpha antagonist) deserves further 6 7 analysis. Future studies with larger samples of men using the newest, prostate-specific 8 Confidential: For Review Only 9 antagonists (such as silodosin) should be undertaken to see if increased prostate selectivity 10 11 12 (compared to tamsulosin) mitigates the risk of fall and fracture. 13 14 15 16 17 Conclusions 18 19 20 21 Prostate-specific alpha antagonists, which are commonly used to treat lower urinary tract 22 23 24 symptoms among older men, are associated with a small, but significant increase in the risk of 25 26 hospitalization or emergency room assessment for a fall, and increase in the risk of fracture by 27 28 29 16% (95% CI 4 to 29%) and head trauma by 15% (95% CI 4 to 27%) during the first 90 days of 30 31 use. Although the absolute magnitude of the risk increase is generally small, it is accentuated in 32 33 34 patient populations with a higher baseline risk of falls, and continued caution should be exercised 35 36 while initiating these medications. 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 16 https://mc.manuscriptcentral.com/bmj Page 17 of 44 BMJ

1 2 What this paper adds 3 4 5 What is already known on this subject: 6 7 • Prostate-specific alpha antagonists are commonly used to treat urinary symptoms in older 8 Confidential: For Review Only 9 men. 10 11 12 • One of the most serious side effects from these medications is hypotension, which may 13 14 lead to a serious fall, fracture or head injury. 15 16 17 • Previous observational studies have not focused solely on the prostate-specific alpha 18 19 antagonists (which are the predominate type of alpha antagonist used for male urinary 20 21 22 symptoms), and have had conflicting conclusions regarding the risk of fall or fracture with 23 24 the initiation of these medications. 25 26 27 28 29 What this study adds: 30 31 • The initiation of prostate-specific alpha antagonists leads to a significant increase in the 32 33 34 risk of a man presenting to a hospital emergency room, or being admitted to hospital, 35 36 because of a fall, fracture and head injury. 37 38 39 • Although the absolute magnitude of this risk increase is small, given the potential 40 41 seriousness of these events in the elderly, men should be counseled on ways to decrease 42 43 this risk (such as taking medication before bed or with food, and avoiding potentially 44 45 46 hazardous activities when starting therapy). 47 48 49 50 51 52 53 54 55 56 57 58 59 60 17 https://mc.manuscriptcentral.com/bmj BMJ Page 18 of 44

1 2 Acknowledgements 3 4 5 Funding/support: This project was conducted at the Institute for Clinical Evaluative Sciences 6 7 (ICES) Western Site through the Kidney, Dialysis, and Transplantation (KDT) research program. 8 Confidential: For Review Only 9 The ICES is funded by an annual grant from the Ontario Ministry of Health and Long-term Care. 10 11 12 ICES Western is funded by an operating grant from the Academic Medical Organization of 13 14 Southwestern Ontario. 15 16 17 18 19 Disclaimer: The opinions, results, and conclusions are those of the authors, and no endorsement 20 21 by the Institute for Clinical Evaluative Sciences, Ontario Ministry of Health and Long-Term Care, 22 23 24 or Academic Medical Organization of Southwestern Ontario is intended or should be inferred. 25 26 27 28 29 Conflict of interest: We have read and understood BMJ policy on declaration of interests. All 30 31 authors have completed the ICMJE uniform disclosure form and declare: no additional support 32 33 34 from any organization for the submitted work; Blayne Welk has received research grants from 35 36 Astellas, and Lisa-Ann Fraser does consultancy for Amgen; there are no other relationships or 37 38 activities that could appear to have influenced the submitted work. Eric McArthur, Y. Joseph 39 40 41 Hwang, Stephanie Dixon, and Michael Ordon have no disclosures. 42 43 44 45 46 Role of the Sponsor: The sponsors had no role in the design and conduct of the study; collection, 47 48 management, analysis, and interpretation of the data; preparation, review, or approval of the 49 50 manuscript; and decision to submit the manuscript for publication. 51 52 53 54 55 Contributors: Blayne Welk initiated the collaborative project, co-designed and edited the data 56 57 58 collection protocol, wrote the statistical analysis plan, and drafted and revised the paper. He is 59 60 18 https://mc.manuscriptcentral.com/bmj Page 19 of 44 BMJ

1 2 the guarantor. Y. Joseph Hwang co-designed the data collection plan, contributed to the analysis 3 4 5 of the results, and revised the paper. Lisa-Ann Fraser, Stephanie Dixon, and Michael Ordon edited 6 7 the data collection protocol, contributed to the analysis of the data, and edited the manuscript. 8 Confidential: For Review Only 9 Eric McArthur edited the data collection protocol, performed the data programming and 10 11 12 statistical analysis, and revised the manuscript. Jade Hayward edited the data collection protocol, 13 14 assisted with the data programming and analysis, and revised the manuscript. 15 16 17 18 19 Affirmation: The lead author (the manuscript’s guarantor) affirms that the manuscript is an 20 21 honest, accurate, and transparent account of the study being reported; that no important aspects 22 23 24 of the study have been omitted; and that any discrepancies from the study as planned (and, if 25 26 relevant, registered) have been explained. 27 28 29 30 31 Data sharing: no additional data available due to governmental privacy regulations. 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 19 https://mc.manuscriptcentral.com/bmj BMJ Page 20 of 44

1 2 Table 1. 90 day risk for hospitalization or emergency room assessment for fall or fracture among 3 men starting prostate-specific alpha antagonist therapy. Secondary outcomes included specific 4 5 fracture subgroups (major osteoporotic, hip), and hospitalization or emergency room assessment 6 for hypotension or head trauma. 7 8 Confidential: For Review Only 9 Odds Ratio (95% Change in Absolute Risk 10 Events n (%) CI) % (95% CI) 11 12 Alpha antagonist Non-alpha users antagonist users 13 (Exposed cohort) (Unexposed cohort) 14 15 n=147,084 n=147,084 16 Primary outcome 17 Fall 2129 (1.45%) 1881 (1.28%) 1.14 (1.07 to 1.21) +0.17 (+0.08 to +0.25) 18 Fracture 699 (0.48%) 605 (0.41%) 1.16 (1.04 to 1.29) +0.06 (+0.02 to +0.11) 19 20 Secondary outcomes 21 Major osteoporotic fracture* 312 (0.21%) 298 (0.20%) 1.05 (0.89 to 1.23) +0.01 (-0.02 to +0.04) 22 Hip fracture 159 (0.11%) 163 (0.11%) 0.98 (0.78 to 1.21) 0.00 (-0.02 to +0.03) 23 24 Hypotension 706 (0.48%) 394 (0.27%) 1.80 (1.59 to 2.03) +0.21 (+0.17 to +0.26) 25 Head trauma 888 (0.60%) 773 (0.53%) 1.15 (1.04 to 1.27) +0.08 (+0.02 to +0.13) 26 *Defined as hip, forearm, humerus or pelvic fracture. 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 20 https://mc.manuscriptcentral.com/bmj Page 21 of 44 BMJ

1 2 Table 2. Subgroup analysis of 90 day risk for hospitalization or emergency room assessment for 3 fall among men starting prostate-specific alpha antagonist therapy. 4 5 6 7 8 Confidential: For ReviewOdds Ratio (95% InteractionOnly Change in Absolute 9 Events/Number at Risk (%) CI) p value Risk % (95% CI) 10 11 Alpha antagonist Non-alpha antagonist 12 users users 13 (Exposed cohort) (Unexposed cohort) 14 Age 15 66-74.9 years 603/72106 (0.84%) 515/72106 (0.71%) 1.17 (1.04 to 1.31) 0.52 +0.12 (+0.03 to +0.21) 16 ≥75 years 1526/74978 (2.04%) 1366/74978 (1.82%) 1.12 (1.04 to 1.21) +0.21 (+0.07 to +0.35) 17 18 Type of alpha antagonist 19 Tamsulosin 1810/123537 (1.47%) 1625/123537 (1.32%) 1.12 (1.04 to 1.19) 0.44 +0.15 (+0.06 to +0.24) 20 Alfuzosin 279/20827 (1.34%) 226/20827 (1.09%) 1.24 (1.04 to 1.48) +0.25 (+0.04 to +0.46) 21 22 Silodosin 40/2720 (1.47%) 30/2720 (1.10%) 1.35 (0.83 to 2.18) +0.37 (-0.23 to +0.96) 23 Residential status 24 Community dwelling 1969/143016 (1.38%) 1757/143016 (1.23%) 1.12 (1.05 to 1.20) 0.25 +0.15 (+0.07 to +0.23) 25 Long-term care 160/4068 (3.93%) 124/4068 (3.05%) 1.30 (1.03-1.65) +0.88 (+0.09 to +1.69) 26 27 Prior fracture status 28 Fracture in past year 57/1294 (4.40%) 59/1294 (4.56%) 0.96 (0.66 to 1.40) 0.39 -0.15 (-1.77 to +1.46) 29 No facture in past year 2072/145790 (1.42%) 1822/145790 (1.25%) 1.14 (1.07 to 1.22) +0.17 (+0.09 to +0.25) 30 31 5-alpha reductase use 32 Present in past 180 days 147/10826 (1.36%) 167/10826 (1.54%) 0.88 (0.70 to 1.10) 0.02 -0.18 (-0.51 to +0.13) 33 Absent in past 180 days 1982/136258 (1.45%) 1714/136258 (1.26%) 1.16 (1.09 to 1.24) +0.20 (+0.11 to +0.28) 34 Tamsulosin Dose 35 36 0.4mg (standard dose) 1622/113989 (1.42%) 1488/113989 (1.31%) 1.09 (1.02 to 1.17) 0.49 +0.12 (+0.02 to +0.21) 37 0.8mg (off label dose) 105/6775 (1.55%) 87/6775 (1.28%) 1.21 (0.91 to 1.62) +0.27 (-0.13 to +0.67) 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 21 https://mc.manuscriptcentral.com/bmj BMJ Page 22 of 44

1 2 References 3 4 5 1 Hollingsworth JM, Wilt TJ. Lower urinary tract symptoms in men. BMJ 2014; 349 :g4474. 6 doi:10.1136/bmj.g4474 7 8 2 YuanConfidential: J, Liu Y, Yang Z, et al. The efficacy For and safety Review of alpha-1 blockers forOnly benign prostatic 9 hyperplasia: an overview of 15 systematic reviews. Curr Med Res Opin 2013; 29 :279–87. 10 doi:10.1185/03007995.2013.766594 11 12 13 3 McVary KT, Roehrborn CG, Avins AL, et al. Update on AUA Guideline on the Management of 14 Benign Prostatic Hyperplasia. Journal of Urology 2011; 185 :1793–803. 15 doi:10.1016/j.juro.2011.01.074 16 17 4 Group ACR. Major cardiovascular events in hypertensive patients randomized to doxazosin 18 19 vs chlorthalidone: the antihypertensive and lipid-lowering treatment to prevent heart attack 20 trial (ALLHAT). JAMA 2000; 283 :1967–75. doi:10.1001/jama.283.15.1967 21 22 5 Filson CP, Hollingsworth JM, CLEMENS JQ, et al. The efficacy and safety of combined therapy 23 with α-blockers and anticholinergics for men with benign prostatic hyperplasia: a meta- 24 analysis. J Urol 2013; 190 :2153–60. doi:10.1016/j.juro.2013.05.058 25 26 27 6 Dunn CJ, Matheson A, Faulds DM. Tamsulosin: a review of its pharmacology and therapeutic 28 efficacy in the management of lower urinary tract symptoms. Drugs Aging 2002; 19 :135–61. 29 30 7 Elhilali MM. Alfuzosin: an alpha1-receptor blocker for the treatment of lower urinary tract 31 symptoms associated with benign prostatic hyperplasia. Expert Opin Pharmacother 32 2006; 7:583–96. doi:10.1517/14656566.7.5.583 33 34 35 8 Osman NI, Chapple CR, Cruz F, et al. Silodosin : a new subtype selective alpha-1 antagonist 36 for the treatment of lower urinary tract symptoms in patients with benign prostatic 37 hyperplasia. Expert Opin Pharmacother 2012; 13 :2085–96. 38 doi:10.1517/14656566.2012.714368 39 40 41 9 Oelke M, Gericke A, Michel MC. Cardiovascular and ocular safety of α1-adrenoceptor 42 antagonists in the treatment of male lower urinary tract symptoms. Expert Opin Drug Saf 43 2014; 13 :1187–97. doi:10.1517/14740338.2014.936376 44 45 10 Bird ST, Delaney JAC, Brophy JM, et al. Tamsulosin treatment for benign prostatic 46 hyperplasia and risk of severe hypotension in men aged 40-85 years in the United States: 47 48 risk window analyses using between and within patient methodology. BMJ 2013; 347 :f6320. 49 doi:10.1136/bmj.f6320 50 51 11 Chrischilles E, Rubenstein L, Chao J, Kreder KJ. Initiation of nonselective alpha 1-antagonist 52 therapy and occurrence of hypotension-related adverse events among men with benign 53 prostatic hyperplasia: a retrospective cohort study. Clinical therapeutics 2001; 23 (5):727- 54 55 743. 56 57 12 Tinetti ME, Kumar C. The patient who falls: "It's always a trade-off". JAMA: The Journal of the 58 American Medical Association 2010; 303 :258–66. doi:10.1001/jama.2009.2024 59 60 22 https://mc.manuscriptcentral.com/bmj Page 23 of 44 BMJ

1 2 13 SOUVEREIN PC, Van Staa TP, EGBERTS ACG, et al. Use of alpha-blockers and the risk of 3 hip/femur fractures. J Intern Med 2003; 254 :548–54. 4 5 6 14 Jacobsen SJ, Cheetham TC, Haque R, et al. Association between 5-alpha reductase inhibition 7 and risk of hip fracture. JAMA: The Journal of the American Medical Association 8 2008;Confidential:300 :1660–4. doi:10.1001/jama.300.14.1660 For Review Only 9 10 15 Hall GC, McMahon AD. Comparative study of modified release alpha-blocker exposure in 11 elderly patients with fractures. Pharmacoepidemiol Drug Saf 2007; 16 :901–7. 12 13 doi:10.1002/pds.1402 14 15 16 Vestergaard P, Rejnmark L, Mosekilde L. Risk of fractures associated with treatment for 16 benign prostate hyperplasia in men. Osteoporos Int 2011; 22 :731–7. doi:10.1007/s00198- 17 010-1320-4 18 19 20 17 Jha P, Deboer D, Sykora K, et al. Characteristics and mortality outcomes of thrombolysis trial 21 participants and nonparticipants: a population-based comparison. J Am Coll Cardiol 22 1996; 27 :1335–42. 23 24 18 Levy AR, O'Brien BJ, Sellors C, et al. Coding accuracy of administrative drug claims in the 25 Ontario Drug Benefit database. Can J Clin Pharmacol 2003; 10 :67–71. 26 27 28 19 WilliamsJ, YoungW. AppendixI: a summary of studies on the quality of health care 29 administrative databases in Canada. In: Goel V, Williams JI, Anderson GM, et al, eds. Patterns 30 of Health Care in Ontario: the ICES Practice Atlas. 2nd ed. Ottawa, ON: Canadian Medical 31 Association; 2000:339-347. 32 33 20 Gibson D, Richards H, Chapman A. The national ambulatory care reporting system: Factors 34 35 that affect the quality of its emergency data. International Journal of Information Quality 36 2008; 2:97–114. 37 38 21 Raina P, Torrance-Rynard V, Wong M, et al. Agreement between self-reported and routinely 39 collected health-care utilization data among seniors. Health Serv Res 2002; 37 :751–74. 40 41 42 22 Austin PC. Using the standardized difference to compare the prevalence of a binary variable 43 between two groups in observational research. Communications in Statistics - Simulation and 44 Computation 2009; 38 :1228–34. 45 46 23 Filson CP, Wei JT, Hollingsworth JM. Trends in medical management of men with lower 47 urinary tract symptoms suggestive of benign prostatic hyperplasia. Urology 2013; 82 :1386– 48 49 92. doi:10.1016/j.urology.2013.07.062 50 51 24 Austin PC. Statistical criteria for selecting the optimal number of untreated subjects matched 52 to each treated subject when using many-to-one matching on the propensity score. American 53 Journal of Epidemiology 2010; 172 :1092–7. doi:10.1093/aje/kwq224 54 55 56 25 Parsons JK, Mougey J, Lambert L, et al. Lower urinary tract symptoms increase the risk of 57 falls in older men. BJU Int 2009; 104 :63–8. doi:10.1111/j.1464-410X.2008.08317.x 58 59 26 Rubenstein LZ. Falls in older people: epidemiology, risk factors and strategies for 60 23 https://mc.manuscriptcentral.com/bmj BMJ Page 24 of 44

1 2 prevention. Age Ageing 2006; 35 Suppl 2 :ii37–ii41. doi:10.1093/ageing/afl084 3 4 27 Bliuc D, Nguyen ND, Milch VE, et al. Mortality risk associated with low-trauma osteoporotic 5 6 fracture and subsequent fracture in men and women. JAMA: The Journal of the American 7 Medical Association 2009; 301 :513–21. doi:10.1001/jama.2009.50 8 Confidential: For Review Only 9 28 Thompson HJ, McCormick WC, Kagan SH. Traumatic brain injury in older adults: 10 epidemiology, outcomes, and future implications. J Am Geriatr Soc 2006; 54 :1590–5. 11 doi:10.1111/j.1532-5415.2006.00894.x 12 13 14 29 Tamsulosin FDA/HC side effect reports by outcome: Syncope. https://www.rxisk.org/ 15 (accessed 22 Sept 2015). 16 17 30 Orwoll E, Lambert LC, Marshall LM, et al. Endogenous testosterone levels, physical 18 performance, and fall risk in older men. Arch Intern Med 2006; 166 :2124–31. 19 20 doi:10.1001/archinte.166.19.2124 21 22 31 Mann RD, Biswas P, Freemantle S, et al. The pharmacovigilance of tamsulosin: event data on 23 12484 patients. BJU Int 2000; 85 :446–50. 24 25 32 Lepor H. Phase III multicenter placebo-controlled study of tamsulosin in benign prostatic 26 hyperplasia. Tamsulosin Investigator Group. Urology 1998; 51 :892–900. 27 28 29 33 Narayan P, Tewari A. A second phase III multicenter placebo controlled study of 2 dosages of 30 modified release tamsulosin in patients with symptoms of benign prostatic hyperplasia. 31 United States 93-01 Study Group. Journal of Urology 1998; 160 :1701–6. doi:10.1016/S0022- 32 5347(01)62386-3 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 24 https://mc.manuscriptcentral.com/bmj Page 25 of 44 BMJ

1 2 Supplemental Online Content 3 4 5 6 7 8 eTable Confidential:1. STROBE checklist for cohort studies For Review Only 9 eTable 2a. Administrative data definitions used to define medical comorbidities 10 11 eTable 2b. Specific drugs included within each medication class 12 eTable 2c. Administrative data definitions used to define medical investigations and procedures. 13 eTable 4. Coding definition for primary and secondary outcomes 14 15 eTable 5. Complete baselines for unmatched and matched cohorts. 16 eTable 6. Study alpha antagonists and their doses and prescription details. 17 eTable 7. Primary outcomes among exposed and unexposed patients in the 180 days prior to 18 alpha antagonist initiation or index date. 19 20 21 eFigure 1. Flow diagram of cohort selection 22 Supplemental References 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 1 https://mc.manuscriptcentral.com/bmj BMJ Page 26 of 44

1 2 eTable 1. STROBE checklist for cohort studies 3

4

5 Item Page 6 No. Recommendation No. 7 Title and abstract 1 (a) Indicate the study’s design with a Abstract, Design 8 Confidential:commonly used term in Forthe title or the Review Only 9 abstract 10 (b) Provide in the abstract an informative Abstract and balanced summary of what was done 11 and what was found 12 13 Introduction 14 Background/rationale 2 Explain the scientific background and Introduction 15 rationale for the investigation being reported 16 Objectives 3 State specific objectives, including any Abstract & Introduction prespecified hypotheses 17 18 Methods 19 Study design 4 Present key elements of study design early Methods, Design & Setting 20 in the paper 21 Setting 5 Describe the setting, locations, and relevant Methods, Design & Setting dates, including periods of recruitment, and Patient Population 22 exposure, followup, and data collection 23 Participants 6 (a) Cohort study —Give the eligibility criteria, Methods, Patient Population, 24 and the sources and methods of selection of eFigure 1 25 participants. Describe methods of followup 26 27 Variables 7 Clearly define all outcomes, exposures, Methods, Study outcomes, 28 predictors, potential confounders, and effect Table 2, eTable 4 29 modifiers. Give diagnostic criteria, if 30 applicable Data sources/ 8* For each variable of interest, give sources eTable 2ac 31 measurement of data and details of methods of 32 assessment (measurement). Describe 33 comparability of assessment methods if 34 there is more than one group 35 Bias 9 Describe any efforts to address potential Methods, Statistical analysis 36 sources of bias 37 Study size 10 Explain how the study size was arrived at Methods, eFigure 1, 38 population based study 39 40 Quantitative 11 Explain how quantitative variables were handled Methods variables in the analyses. If applicable, describe which 41 groupings were chosen and why eTable 5 42 Statistical 12 (a) Describe all statistical methods, including Methods, Statistical analysis 43 methods those used to control for confounding 44 (b) Describe any methods used to examine Methods, Statistical analysis 45 subgroups and interactions 46 (c) Explain how missing data were addressed Methods, Data sources 47 (d) Cohort study —If applicable, explain how loss Methods, Patient Population 48 to followup was addressed 49 (e) Describe any sensitivity analyses Methods, Statistical analysis 50 Participants 13* (a) Report numbers of individuals at each stage eFigure 1 of study—eg numbers potentially eligible, 51 examined for eligibility, confirmed eligible, 52 included in the study, completing followup, and 53 analysed 54 (b) Give reasons for nonparticipation at each NA 55 stage 56 (c) Consider use of a flow diagram eFigure 1 57 Descriptive data 14* (a) Give characteristics of study participants (eg eTable 5 58 demographic, clinical, social) and information on 59 exposures and potential confounders 60 2 https://mc.manuscriptcentral.com/bmj Page 27 of 44 BMJ

1 2 (b) Indicate number of participants with missing Methods, Data sources 3 data for each variable of interest 4 (c) Cohort study —Summarise followup time NA. 5 (eg, average and total amount) 6 Outcome data 15* Cohort study —Report numbers of outcome Table 1 events or summary measures over time 7 Main results 16 (a) Give unadjusted estimates and, if applicable, Table 1, Table 2 8 Confidential:confounderadjusted estimates For and their Review Only 9 precision (eg, 95% confidence interval). Make 10 clear which confounders were adjusted for and 11 why they were included 12 (b) Report category boundaries when Results 13 continuous variables were categorized Interquartile range 14 (c) If relevant, consider translating estimates of Results 15 relative risk into absolute risk for a meaningful Table 1, Table 2 time period 16

17 Other analyses 17 Report other analyses done—eg analyses of Table 2, eTable 7 18 subgroups and interactions, and sensitivity analyses 19 20 Discussion 21 Key results 18 Summarise key results with reference to study Discussion objectives 22 Limitations 19 Discuss limitations of the study, taking into account Discussion 23 sources of potential bias or imprecision. Discuss both 24 direction and magnitude of any potential bias 25 Interpretation 20 Give a cautious overall interpretation of results Discussion 26 considering objectives, limitations, multiplicity of 27 analyses, results from similar studies, and other 28 relevant evidence 29 Generalisability 21 Discuss the generalisability (external validity) of the Discussion 30 study results 31 Other information 32 Funding 22 Give the source of funding and the role of the Acknowledgements 33 funders for the present study and, if applicable, for 34 the original study on which the present article is 35 based 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 3 https://mc.manuscriptcentral.com/bmj BMJ Page 28 of 44

1 2 eTable 2a. Administrative data definitions used to define medical comorbidities. 3 4 5 Variable Database Codes 6 Acute kidney CIHIDAD ICD9 584 injury 7 ICD10 N17 Alcoholism CIHIDAD ICD9 303, 3050 8 Confidential: For Review Only ICD10 E244, E512, F10, G312, G621, G721, I426, K292, K70, K860, T510, X45, 9 X65, Y15, Y573, Z502, Z714, Z721 10 Atrial CIHIDAD ICD9 4273 11 fibrillation/flutter ICD10 I48 12 Cancer CIHIDAD ICD9 150, 154, 155, 157, 162, 174, 175, 185, 203, 204, 205, 206, 207, 208, 230 13 ICD10 971, 980, 982, 984, 985, 986, 987, 988, 989, 990, 991, 993, C15, C18, 14 C19, C20, C22, C25, C34, C50, C56, C61, C82, C83, C85, C91, C92, C93, C94, 15 C95, C00, D00, D01 D02, D05, D07 16 OHIP ICD9 203, 204, 205, 206, 207, 208, 150, 154, 155, 157, 162, 174, 175, 183, 185 17 Cataract CIHIDAD ICD9 366 18 ICD10 H25, H26, H27, H28 19 OHIP Fee code E214, E140, E141 20 Chronic kidney CIHIDAD ICD9 4030, 4031, 4039, 4040, 4041, 4049, 582, 583, 580, 581, 584, 585, 586, disease 21 587, 5880, 5888, 5889, 5937 ICD10 E102, E112, E132, E142, I12, I13, N08, N18, N19 22 OHIP ICD9 403, 585 23 Chronic liver CIHIDAD ICD9 4561, 4562, 070, 5722, 5723, 5724, 5728, 573, 7824, V026, 2750, 2751, 24 disease 7891, 7895, 571 25 ICD10 B16, B17, B18, B19, I85, R17, R18, R160, R162, B942, Z225, E831, E830, 26 K70, K713, K714, K715, K717, K721, K729, K73, K74, K753, K754, K758, K759, 27 K76, K77 28 OHIP ICD9 571, 573, 070, Fee code Z551, Z554 29 Chronic lung CIHIDAD ICD9 491, 492, 493, 494, 495, 496, 500, 501, 502, 503, 504, 505, 5064, 5069, 30 disease 5081, 515, 516, 517, 5185, 5188, 5198, 5199, 4168, 4169 31 ICD10 I272, I278, I279, J40, J41, J42, J43, J44, J45, J47, J60, J61, J62, J63, J64, J65, J66, J67, J68, J701, J703, J704, J708, J709, J82, J84, J92, J941, J949, J953, 32 J961, J969, J984, J988, J989, J99 33 OHIP ICD9 491, 492, 493, 494, 496, 501, 502, 515, 518, 519, Fee code J689, J889 34 Congestive CIHIDAD ICD9 425, 5184, 514, 428 35 heart failure ICD10 I500, I501, I509, I255, J81 36 CCP 4961, 4962, 4963, 4964 37 CCI 1HP53, 1HP55, 1HZ53GRFR, 1HZ53LAFR, 1HZ53SYFR 38 OHIP ICD9 428, Fee code R701, R702, Z429 39 Coronary artery CIHIDAD ICD9 410 412, 414, 4292, 4295, 4296, 4297, 413 40 disease or ICD10 I21, I22, I23, I24, I25, Z955, Z958, Z959, R931, T822, I20 41 angina CCP 4801, 4802, 4803, 4804, 4805, 481, 482, 483 42 CCI 1IJ26, 1IJ27, 1IJ54, 1IJ57, 1IJ50, 1IJ76 43 OHIP ICD9 410, 412, 413, Fee code R741, R742, R743, G298, E646, E651, E652, 44 E654, E655, G262, Z434, Z448 45 Dementia CIHIDAD ICD9 2900, 2901, 2903, 2904, 2908, 2909, 2948, 2949, 3310, 3311, 3312, 2941, 797 46 ICD10 F065, F066, F068, F069, F09, F00, F01, F02, F03, F051, G30, G31, R54 47 OHIP ICD9 290, 331, 797 48 Diabetes ODB Insulin (Isophane) Beef & Pork, Insulin (Isophane) Human Biosynthetic, Insulin 49 mellitus (Isophane) Human Semisynthetic, Insulin (Isophane) Pork, Insulin (Neutral) 50 Human Semisynthetic, Insulin (Neutral) Pork, Insulin, (Protamine Zinc) Beef & 51 Pork, Insulin (Sulfated) Beef, Insulin (Zinc) Beef, Insulin (Zinc) Beef & Pork, Insulin 52 (Zinc) Human Biosynthetic, Insulin (Zinc) Human Semisynthetic, Insulin (Zinc) 53 Pork, Insulin Aspart, Insulin Aspart & Insulin Aspart Protamine, Insulin Detemir, 54 Insulin Glargine, Insulin Glulisine, Insulin Human Biosynthetic, Insulin Human Semisynthetic, Insulin Injection Pork, Insulin Isophane Injection Pork, Insulin 55 Lispro, Acarbose, Acetohexamide, Chlorpropamide, Gliclazide, Glimepir, 56 Glyburide, Metformin HCL, Nateglinide, Pioglitazone HCL, Repaglinide, 57 Rosiglitazone Male, Sitagliptin Phosphate Monohydrate, Tolbutamide 58 Glaucoma CIHIDAD ICD9 365 59 ICD10 H25, H40, H42 60 4 https://mc.manuscriptcentral.com/bmj Page 29 of 44 BMJ

1 2 OHIP Fee code E132, E133 3 Hypertension ODB Acebutolol HCL, Aliskiren Fumarate, Amlodipine Besylate, Amlodipine Besylate & 4 Atrovastatin, Atenolol, Atenolol & Chlorthalidone, Benazepril HCL, Bisoprolol 5 Fumurate, Brimonidine Tartrate & Timolol Maleate, Brizolamide Tartrate & Timolol Maleate, Bumetanide, Candesartan Cilexetil, Candesartan Cilexetil & 6 Hydrochlorothizide, Captopril, Carvedilol, Chlorthalidone, Cilazapril, Cilazapril & 7 Hydrochlorothizide, Diltiazem HCL, Enalapril Sodium, Enalapril Maleate, 8 Confidential:Eprosartan Mesylate, For Eprosartan MesylateReview & Hydrochlorothizide, Only Erythrityl 9 Tetranitrate, Felodipine, Flunarizine HCL, Fosinopril Sodium, Hydralazine HCL, 10 Hydrochlorothizide, Indapamide, Irbesartan, Irbesartan & Hydrochlorothiazide, 11 Hydrochlorothiazide , Hydrochlorothiazide, Labetalol HCL, Lisinopril, Lisinopril & 12 Hydrochlorothiazide, Losartan Potassium, Losartan Potassium & 13 Hydrochlorothiazide, Methyldopa & Hydrochlorothiazide, Metoprolol Tartrate, Minoxidil, Nadolol, Nicardipine HCL, Nifedipine, Nimodipine, Olmesartan 14 Medoxomil, Olmesartan Medoxomil & Hydrochlorothiazide, Oxprenolol HCL, 15 Perindopril Erbamine, Perindopril Erbamine & Indapamine, Phenoxybenzamine, 16 Pindolol, Pindolol & Hydrochlorothiazide, Propanolol HCL, Quinapril HCL & 17 Hydrochlorothiazide, Ramipril, Ramipril & Hydrochlorothiazide, Spironolactone, 18 Spironolactone & Hydrochlorothiazide, Telmisartan, Telmisartan & 19 Hydrochlorothiazide, Timolol Maleate, Timolol Maleate & Hydrochlorothiazide, 20 Timolol Maleat Hydrochlorothiazide e & Travoprost, Trandolapril, Triamterene, 21 Triamterene & Hydrochlorothiazide, Valsartan, Valsartan & Hydrochlorothiazide, Verapamil HCL. 22 Hypotension CIHIDAD ICD9 458 23 ICD10 I95 24 Macular CIHIDAD ICD9 362 25 degeneration ICD10 H35 26 OHIP Fee code E154, E125, E126, E147 27 Obesity CIHIDAD ICD9 278 28 ICD10 E66 29 OHIP ICD9 278 30 Osteoporosis CIHIDAD ICD9 733 31 ICD10 M80, M81 32 OHIP ICD9 733 33 Parkinson’s CIHIDAD ICD9 332 34 disease ICD10 G20, F023 35 Peripheral CIHIDAD ICD9 4402, 4408, 4409, 5571, 4439, 444 vascular 36 ICD10 I700, I702, I708, I709, I731, I738, I739, K551 disease 37 CCP 5125, 5126, 5129, 5159, 5014, 5016, 5018, 5028, 5038 CCI 1KA76, 1KA50, 1KE76, 1KG50, 1KG57, 1KG76MI, 1KG87, 1IA87, 1IB87, 38 1IC87, 1ID87, 1KA87, 1KE57 39 OHIP Fee code R787, R780, R797, R804, R809, R875, R815, R936, R783, R784, R785, 40 E626, R814, R786, R937, R860, R861, R855, R856, R933, R934, R791, E672, 41 R794, R813, R867, E649 42 Prostate Cancer OHIP Feecode A355, A356, A353, A935, A354, C355, C356, C353, C935, C354 A345, 43 A745, A346, A343, A340, A341, A348 C345, C745, C346, C343, C340, C341, 44 C348 A445, A845, A446, A443, A444, A441, A448, C445, C845, C446, C443, 45 C444, C441, C448 with ICD9 185 46 Rheumatoid CIHIDAD ICD9 714 arthritis 47 ICD10 M05, M06 OHIP ICD9 714 48 Seizure CIHIDAD ICD9 345, 7803 49 ICD10 G40, G41 50 Stroke or TIA CIHIDAD ICD9 430, 431, 436, 432, 4340, 4341, 4349, 3623, 435 51 ICD10 I600, I601, I602, I603, I604, I605, I606, I607, I609, I61, I630, I631, I632, 52 I633, I634, I635, I638, I639, I64, I92, H341, G450, G451, G452, G453, G458, 53 G459, H340 54 Data sources included CIHIDAD (inpatient discharge abstracts), OHIP (physician billings) and ODB (universal medication 55 coverage). CIHIDAD used ICD10 (for diagnostic coding) and CCI coding (for interventions) after April 1, 2002, and ICD9/CCP 56 coding prior to this. Specific ICD 9/10 and CCP/CCI codes, and medication classes used to define hypertension and diabetes are 57 defined above. The presence of any data element was considered evidence of the comorbidity. A lookback window of 5 years prior 58 to index date was used for all conditions, aside from prior falls and fractures (1 year lookback), and hypertension and diabetes (6 59 month lookback). 60 5 https://mc.manuscriptcentral.com/bmj BMJ Page 30 of 44

1 2 eTable 2b. Specific drugs included within each medication class. Approximately 7,400 individual 3 Drug Identification Numbers (DINs) were used to identify relevant brands and doses. 4 5 6 Medication class Medications 7 5 alpha reductase inhibitor Finasteride, Dutasteride 8 ACE inhibitorsConfidential: or ARBs Ramipril, Ramipril For & Hydrochlorothiazide, Review Trandolapril, Candesartan Only Cilexetil, 9 Candesartan Cilexetil & Hydrochlorothiazide, 10 Eprosartan Mesylate, Eprosartan Mesylate & Hydrochlorothiazide, Irbesartan, Irbesartan & Hydrochlorothiazide, Losartan potassium, Losartan 11 potassium & Hydrochlorothiazide, Olmesartan Medoxomil, Olmesartan 12 Medoxomil & Hydrochlorothiazide, Telmisartan , Telmisartan & 13 Hydrochlorothiazide, Valsartan, Valsartan & Hydrochlorothiazide 14 Androgen deprivation therapy Buserelin Acetate, Goserelin Acetate, Leuprolide, Leuprolide Acetate 15 Anti -inflammatory Cannabidiol & Dronabinol, Celecoxib, Diclofencac sodium, Diclofencac sodium & 16 Misoprostol, Diflunisal, Etodolac, Fenoprofen calcium, Floctafenine, Flurbiprofen, 17 Glucosamine & Chondroitin, Ibuprofen, Indomethacin, Ketoprofen, Ketoralac 18 Tromethamine, Mefenamic Acid, Meloxicam, Nabumetone, Naproxen, Naproxen sodium, Oxaprpozin, Phenylbutazone, Piroxicam, Rofecoxib, Sulindac, 19 Tenoxicam, Tiaprofenic acid, Tolmetin sodium, Valdecoxib 20 Antiandroge ns Bicalutamide, Cyproterone, Cyproterone Acetate, Flutamide, Nilutamide 21 Antibiotic Amikacin Sulphate, Amoxicillin, Amoxicillin & Clavulanic Acid Potassium, 22 Amoxicillin Trihydrate, Amoxicillin Trihydrate & Clavulanic Acid Potassium, 23 Ampicillin, Ampicillin Sodium, Ampicillin Trihydrate, Azithromycin, Azithromycin 24 Dihydrate, Bacampicillin HCL, Bacitracin, Bacitracin Zinc & Cysteine & Glycine & 25 Neomycin Sulphate & Threonine, Bacitracin Zinc & Neomycin Sulphate & 26 Polymyxin B Sulphate, Bacitracin Zinc & Polymyxin B Sulphate, Carbenicillin, 27 Carbenicillin Disodium, Cefaclor, Cefadroxil, Cefadroxil Monohydrate, Cefazolin Sodium, Cefepime HCL, Cefixime, Cefoperazone Sodium, Cefotaxime Sodium, 28 Cefoxitin Sodium, Cefprozil, Ceftazidime Sodium, Ceftazidme Hydrate, 29 Ceftriaxone Sodium, Cefuroxime Axetil, Cephalexin, Cephalexin Monohydrate, 30 Cephradine, Ciprofloxacin, Ciprofloxacin HCL, Ciprofloxacin HCL & 31 Dexamethasone, Clindamycin, Clindamycin Phosphate, Clindamycin Phosphate 32 & Glycolic Acid, Cloxacillin, Cloxacillin Sodium, Colistin Sodium 33 Methanesulfonate, Daptomycin, Dicloxacillin Sodium, Erythromycin, 34 Erythromycin Estolate, Erythromycin Ethyl Succinate, Erythromycin Ethyl Succinate & Sulfixoxazole, Erythromycin Gluceptate, Erythromycin Lactobionate, 35 Erythromycin Stearate, Flucloxacillin Sodium, Fluocinolone Acetonide & 36 Neomycin Sulphate & Polymyxin B Sulphate, Framycetin Sulphate, Fusidic Acid, 37 Fusidic Acid Sodium, Gatifloxacin, Gentamicin, Gentamicin & Colistin, 38 Gentamicin Sulphate, Garamicidin & Neomycin & Polymyxin B Sulphate, 39 Garamicidin & Polymyxin B Sulphate, Grepafloxacin HCL, Levofloxacin, 40 Linezolid, Moxifloxicin HCL, Mupirocin, Neomycin Sulphate, Neomycin Sulphate 41 & Polymyxin B Sulphate, Netilmicin Sulphate, Norfloxacin, Ofloxacin, 42 Paromomycin, Penicillin G Benzathine, Penicillin G Potassium, Penicillin G Procain Salt, Penicillin G Sodium, Penicillin V, Penicillin V Benzathine, Penicillin 43 V Potassium, Piperacillin, Piperacillin Sodium & Tazobactam Sodium, 44 Pivampicillin, Pivmecillinam, Polymyxin B Sulphate & Trimethoprim, 45 Spectinomycin HCL, Spiramycin, Streptomycin, Streptomycin Sulphate, 46 Sulfabenzamide & Sulfacetamide & Sulfathiazole, Sulfacetamide Sodium, 47 Sulfadiazine, Sulfadiazine & Trimethoprim, Sulfamethoxazole, Sulfamethoxazole 48 & Trimethoprim, Sulfapyridine, Sulfisoxazole, Telithromycin, Tobramycin, 49 Tobramycin Sulphate, Trimethoprim, Azithromycin, Azithromycin Dihydrate, 50 Clarithromycin Anticoagulant Enoxaparin, Dalteparin Sodium, Tinzaparin Sodium, Nadroparin Calc, 51 Enoxaparin Sodium, Dabigatran Etexilate, Rivaroxaban, Fondaparinux Sodium, 52 Lepirudin, Heparin Sodium, Coumadin, Heparin, Heparin Calcium, Danaparoid 53 Sodium 54 Anticonvulsants Carbamazepine, Clobazam, Clonazepam, Divalproex sodium, Ethosuximide, 55 Fosphenytoin, Gabapentin, Lacosamide, Lamotrigine, Levetiracetam, 56 Magnesium pyrogluconate, Magnesium sulphate, Mephenytoin, Mephobarbital, 57 Methsuximide, Oxcarbazepine, Phenobarbital, Phensuximide, Phenytoin, 58 Phenytoin sodium, Pregabalin, Primidone, Secobarbital sodium, Topiramate, Valproate sodium, Valproic acid, Vigabatrin 59 60 6 https://mc.manuscriptcentral.com/bmj Page 31 of 44 BMJ

1 2 Antidepressants Amitriptyline, Amitriptyline HCL, Amitriptyline HCL & Baclofen, Amitriptyline HCL 3 & Perphenazine, Amoxapine, Bupropion HCL, Clomipramine HCL, Desipramine 4 HCL, Doxepine HCL, Duloxetine, Imipramine, Imipramine HCL, Isocarboxazid, 5 Maprotiline HCL, Mirtazapine, Moclobemide, Nortriptyline, Nortiptyline HCL, 6 Phenelzine sulphate, Protriptyline HCL, Selegiline HCL, Tranylcypromine 7 sulphate, Trazadone HCL, Trimipramine, Trimipramine maleate Antineoplastic Abatacept, Aldesleukin, Alemtuzumab, Altretamine, Aminoglutethimide, 8 Confidential:Anastrozole, Aspariginase, For Azathioprine, Review Bicalutamide, Bleomycin Only Sulphate, 9 Buserelin Acetate, Busulfan, Capecitabine, Carmustine, Chlorambucil, 10 Cladribine, Cyclophosphamide, Cyproterone Acetate, Cytarabine, Decarbazine, 11 Dasatinib Monohydrate, Daunorubicin, Degarelix Acetate, Diethylstilbestrol, 12 Diethylstilbestrol Diphosphate Sodium, Doxorubicin HCL, Efalizumab, Epirubicin 13 HCL, Erlotinib HCL, Estramustine Phosphate Disodium, Etoposide, Everolimus, 14 Exemestane, Fludarabine Phosphate Sodium, Fluorouracil, Flutamide, 15 Formestane, Goserelin Acetate, Hydroxyurea, Imatinib Mesylate, Interferon, Interferon Alfa2B, Irinotecan HCL, Lanreotide, Lenalidomide, Letrozole, 16 Leuprolide Acetate, Levamisole HCL, Lomustine, Mechlorethamine HCL, 17 Megestrol Acetate, Melphalan, Mercaptopurine, Methotrexate Sodium, 18 Mitomycin, Mitotane, Nilotinib HCL Monohydrate, Nilutamide, Profimer Sodium, 19 Procarbazine HCL, Rituximab, Sirolimus, Sorafenib, Sunitinib Malate, 20 Tacrolimus, Tamoxifen Citrate, Temozolomide, Thioguanine, Thiotepa, Tretinoin, 21 Triptorelin, Triptorelin Pamoate, Vaccine B.C.G, Vinblastine Sulphate, 22 Vincristine Sulphate Antiparkinson drug s Benserazide HCL & Levodopa, Benztropine Mesylate, Biperiden HCL, 23 Bromocriptine Mesylate, Carbidopa & Entacapone & Levodopa, Carbidopa & 24 Levodopa, Entacapone, Pergolide Mesylate, Pramipexole HCL, Procyclidine 25 HCL, Profenamine HCL, Rasagiline Mesylate, Ropinirole HCL, Selegiline HCL, 26 Trihexyphenidyl HCL 27 Antiplatelet Ticlopidine HCl, Ticlopidine, Clopidogrel, Clopidogrel Bisulphate, Prasugrel HCL, 28 Dipyridamole, Acetylsalicylic Acid & Dipyridamole 29 Antipsychotic (atypical) Risperidone, Quetiapine fumarate, Olanzapine 30 Antipsychotic (other) Chlorpromazine HCL, Clozapine, Flupenthixol decanoate, Flupenthixol HCL, 31 Fluphenazine decanoate, Fluphenazine HCL, Haloperidol, Haloperidol decanoate, Loxapine HCL, Loxapine succinate, Mesoridazine besylate, 32 Methotrimeprazine, Olanzapine tartrate, Paliperidone, Pericyazine, 33 Perphenazine, Pimozide, Pipotiazine palmitate, Prochlorperazine maleate, 34 Prochlorperazine mesylate, Risperidone, Thioridazine HCL, Thiothixene, 35 Trifluoperazine HCL, Ziprasidone HCL, Zuclopenthixol acetate, Zuclopenthixol 36 decanoate, Zuclopenthixol dihyrochloride 37 Benzodiazepine Alprazolam, Bromazepam, Chlordiazepoxide HCL, Chlordiazepoxide HCL & 38 Clidinium Bromide, Clobazam, Clonazepam, Clorazepate Dipotassium, 39 Diazepam, Diazepam & Methycellulose, Estazolam, Flumazenil, Flurazepam HCL, Ketazolam, Lorazepam, Midazolam HCL, Nitrazepam, Oxazepam, 40 Temazepam, Triazolam, Zaleplon, Zopiclone 41 Beta blockers Acebutolol HCL, Atenolol, Atenolol & Chlorthalidone, Betaxolol HCL, 42 Bisoprolol fumurate, Brimonidine tartrate & Timolol maleate, 43 Carvedilol, Brinzolamide & Timolol maleate, Labetalol HCL, Metoprolol 44 tartrate, Nadolol, Oxprenolol HCL, Pindolol, Pindolol & 45 Hydrochlorothiazide, Propanolol HCL, Propanolol HCL & 46 Hydrochlorothiazide, Sotalol HCL, Timolol maleate, Timolol maleate & Travoprost 47 Bisphosphonate Alendronate, Alendronate Sodium, Alendronate Sodium & 48 Cholecalciferol, Calcium Carbonate & Etidronic Acid, Calcium 49 Carbonate & Etidronic Acid Sodium, Calcium Carbonate & 50 Risedronate Sodium, Clodronate Disodium, Clodronate Acid Disodium, 51 Etidronate & Calcium Carbonate, Etidronate Disodium, Etidronic Acid 52 Disodium, Ibandronate, Pamidronate Disodium, Pamidronic Acid 53 Disodium, Risedronate Sodium & Calcium, Risedronate Sodium & 54 Calcium & Vitamin D3, Zoledronic Acid Calcium channel blockers Amlodipine besylate, Amlodipine besylate & Atorvastatin, Diltiazem HCL, 55 Erythrityl Tetranitrate, Felodipine, Flunarizine HCL, Nifedipine, Nicardipine HCL, 56 Nimodipine, Verapamil HCL 57 Cholinesterase inhibitors Aricept, MylanGalantamine ER, PatGalantamine ER, SandozRivastigmine, 58 RatioRivastigmine, PMSRivastigmine, TevaRivastigmine, APORivastigmine, 59 MylanRivastigmine, Exelon, Reminyl ER 60 7 https://mc.manuscriptcentral.com/bmj BMJ Page 32 of 44

1 2 Denosumab Denosumab, Denosumab Recombinant, Pamidronic Acid Disodium, 3 Tiludronate Disodium 4 Digoxin Digoxin 5 Glucocorticoids Budenoside, Methylprednisone, Hydrocortisone, Dexamethasone, 6 Betamethasone, Triamcin, Cortisone, Fludrocortisone, 7 Glucose test strip Diagnostic Agent Diabetes, Glucose, Blood Tests, Glucose, Urine Tests, Non Pharmaceutical Ingredient 8 Confidential: For Review Only Inhaled acetylcholine Ipratropium Bromide, Tiotropium Bromide 9 Inhaled beta -agonist Methaproterenol Sulfate, Albuterol, Albuterol & Albuterol Sulfate, Albuterol 10 Sulfate, Fenoterol HBR, Zenhale, Foradil, Oxeze, Pirbuterol Acetate, Procaterol 11 HCL, Salmeterol Xinafoate, Formoterol & Mometasone, Formoterol Fumarate, 12 Terbutaline Sulfate, Fluticasone Propionate & Salmeterol Xinafoate, Budesonide 13 & Formoterol Fumarate, Albuterol Sulfate & Ipratropium Bromide 14 Inhaled corticosteroid Beclomethasone Dipropionate, Budesonide, Ciclesonide, Flunisolide, 15 Fluticasone Propionate, Triamcinolone Acetonide Narcotics 16 Acetaminophen & Caffeine & Codeine, Acetaminophen & Caffeine & Codeine Phosphate, Acetaminophen & Caffeine Citrate & Codeine Phosphate, 17 Acetaminophen & Chlorzoxazone & Codeine, Acetaminophen & Chlorzoxazone 18 & Codeine Phosphate, Acetaminophen & Codeine & Doxylamine, 19 Acetaminophen & Codeine Phosphate, Acetaminophen & Dextromethorphan Hbr 20 & Doxylamine & Pseudoephedrine Hcl, Acetaminophen & Dextromethorphan Hbr 21 & Doxylamine Succinate & Pseudoephedrine Hcl, Acetaminophen & Oxycodone 22 Hcl, Acetaminophen & Pseudoephedrine Hcl, Acetylsalicylic Acid & Caffeine & 23 Codeine Phosphate, Acetylsalicylic Acid & Caffeine & Dextropropoxyphene Hcl, Acetylsalicylic Acid & Caffeine & Propoxyphene Hcl, Acetylsalicylic Acid & 24 Codeine Phosphate, Acetylsalicylic Acid & Oxycodone Hcl, Alfentanil Hcl, 25 Ammonium Chloride & Codeine Phosphate & Guaifenesin, Analgesics, 26 Anileridine Hcl, Atropine Sulfate & Attapulgite & Hyoscyamine Sulfate & Opium 27 Powder & Pectin & Scopolamine Hbr, Belladona & Opium, Belladona Extract For 28 Oral Use & Opium Powder, Codeine, Codeine Phosphate, Dextromethorphan 29 Hbr & Guaifenesin & Menthol, Dextropropoxyphene Hcl, Dextropropoxyphene 30 Napsylate, Diamorphine Hcl, Fentanyl, Fentanyl & Baclofen, Fentanyl & 31 Bupivacaine, Fentanyl & Bupivacaine & Baclofen, Fentanyl & Bupivacaine & Clonidine, Fentanyl & Bupivicaine & Clonidin & Baclofen, Fentanyl & Clonidine 32 Hcl, Fentanyl Citrate, Hydrocodone Bitartrate & Ibuprofen, Hydromorphone, 33 Hydromorphone & Baclofen, Hydromorphone & Bupivacaine & Baclofen, 34 Hydromorphone & Bupivicaine, Hydromorphone & Bupivicaine & Baclofen, 35 Hydromorphone & Bupivicaine & Baclofen & Clonidine, Hydromorphone & 36 Bupivicaine & Clonidine, Hydromorphone Hbr, Hydromorphone Hcl, Ibuprofen & 37 Diphenhydramine Hcl, Levorphanol Tartrate, Meperidine Hcl, Meperidine Hcl & 38 Promethazine Hcl, Morphine, Morphine & Baclofen, Morphine & Bupivicaine, 39 Morphine & Bupivicaine & Baclofen, Morphine & Lioresal, Morphine Hcl, Morphine Succinate, Morphine Sulfate, Naloxone Hcl & Oxycodone Hcl, Opium, 40 Narcotic Compound, Opium & Belladonna, Opium & Camphor, Opium Tincture, 41 Oxycodone Hcl, Oxymorphone Hcl, Propoxyphene Hcl, Remifentanil Hcl, 42 Sufentanil Citrate, Tapentadol Hcl, Tramadol, Tramadol Hcl, Bupivacaine & 43 Fentanyl, Fentanyl & Baclofen & Clonidine, Fentanyl & Bupivacaine, Fentanyl & 44 Bupivicaine & Clonidine, Fentanyl & Bupivicaine & Clonidine & Baclofen, 45 Hydromorphine & Bupivicaine & Clonidine, Morphine & Bupivicaine, 46 Acetylsalicylic Acid & Caffeine & Pentazocine Hcl, Butorphanol Tartrate, 47 Nalbuphine Hcl, Pentazocine Lactate, Pentazocine Hcl Non -potassium sparing diuretics Bumetanide, Ethacrynic Acid, Furosemide, Chlorthalidone, Hydrochlorothiazide, 48 Indapamide, Methyldopa & Hydrochlorothiazide, Metolazone, Timolol Maleate & 49 Hydrochlorothiazide 50 Overactive bladder medication Darifenacin, Flavoxate HCL, Oxybutynin, Oxybutynin Chloride, Solifenacin 51 Succinate, Tolterodine Tartrate, Trospium Chloride, Trospium Hydroxide 52 Chloride 53 Potassium sparing diuretics Amiloride HCL, Amiloride HCL & Hydrochlorothiazide, Epleronone, 54 Spironolactone, Spironolactone & Hydrochlorothiazide, Triamterene, Triamterene 55 & Hydrochlorothiazide, Proton Pump inhibitors Amoxicillin Trihydrate & Clarithromycin & Lansoprazole, Dexlansoprazole, 56 Esomeprazole Magnesium, Lansoprazole, Lansoprazole Sodium, Omeprazole, 57 Omeprazole Magnesium, Pantoprazole, Pantoprazole Magnesium, Pantoprazole 58 Sodium, Rabeprazole Sodium 59 60 8 https://mc.manuscriptcentral.com/bmj Page 33 of 44 BMJ

1 2 Selective Serotonin Reuptake Fluoxetine HCL, Fluvoxamine Maleate, Paroxetine HCL, Sertraline HCL, 3 Inhibitors Fluoxetine, Sertraline, Citalopram HBR, Escitalopram Oxalate, Citalopram 4 Hydrobromide, Nefazodone HCL, 5 Smoking cessation aid Varenicline Tartrate, Bupropion HCL, Nicotine 6 Statins Atorvastatin, Cerivastatin Sodium, Fluvastatin Sodium, Lovastatin, Provastatin 7 Sodium, Rosuvastatin Calcium, Simvastatin Testosterone replacement Androderm, Testosterone Cipionate, Androgel, Testim, Testosterone Enanthate, 8 Confidential: For Review Only Testosterone Propionate, Testosterone Cypionate, Testosterone, 9 Methyltestosterone, Testosterone Powder, Testosterone Decanoate 10 Data source was the ODB, which provides universal medication coverage to all people over 65 years of age. Each medication is 11 identified with a unique and permanent DIN. The presence of any medication within the medication class in the 6 months prior to the 12 index date was considered evidence of recent or current medication usage. 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 9 https://mc.manuscriptcentral.com/bmj BMJ Page 34 of 44

1 2 eTable 2c. Administrative data definitions used to define medical investigations and procedures. 3 4 5 Variable Database Codes 6 Bone mineral density test OHIP Fee code J654, J688, J854, J888, X149, X152, X153, X155, Y654, 7 Y688, Y854, Y888 Bone scan OHIP Fee code J850, J650, J852, J862 8 Confidential: For Review Only Cardiac catheterization CIHIDAD/SDS CCP 4995, 4996, 4997 9 CCI 3IJ30GP, 3HZ30GP, 2HZ24GPKJ, 2HZ24GPKL, 2HZ24GPKM, 10 2HZ24GPXJ, 2HZ28GPPL, 2HZ71GP 11 OHIP Fee code G296, G297, G299, G300, G301, G304, G305, G306 12 Cardiac stress test CIHIDAD/SDS CCP 0341, 0342, 0343, 0344, 0605 13 CCI 2HZ08, 3IP70 14 OHIP Fee code G315, G174, G111, G112, G319, G582, G583, G584, 15 J604, J606, J607, J608, J611, J612, J613, J667, J807, J808, J809, 16 J804, J811, J812, J813, J866, J867, J609, J666, J814, J614, J810, 17 J610 18 Carotid endarterectomy OHIP Fee code N220, R792 19 Carotid ultrasound CIHIDAD/SDS CCP 0281 CCI 3JE30, 3JG30 20 OHIP Fee code J201, J501, J190, J191, J490, J491, J492 21 Chest x -ray OHIP Fee code X090, X091, X092, X195 22 CT head OHIP Fee code X188, X400, X401, X402, X405, X408 23 CT spine OHIP Fee code X128, X415, X416 24 Echocardiography CIHIDAD/SDS CCP 0282 25 CCI 3IP30 26 OHIP Fee code G560, G561, G562, G566, G567, G568, G570, G571, 27 G572, G574, G575, G576, G577, G578, G581 28 EEG OHIP Fee code G414, G415, G416, G417, G418, G540, G542, G544, 29 G545, G546, G554, G555 30 Heart valve replacement OHIP Fee code R728, R735, R738, R772 31 Holter monitoring CIHIDAD/SDS CCP 0354 32 CCI 2HZ24JAKH 33 OHIP Fee code G311, G320, G647, G648, G649, G650, G651, G652, G653, G654, G655, G656, G657, G658, G659, G660, G661, G682, 34 G683, G684, G685, G686, G687, G688, G689, G690, G692, G693 35 Prostate -specific antigen OHIP Fee code L354, L358 36 (PSA) test 37 Pulmonary function test OHIP Fee code J301, J303, J304, J305, J306, J307, J308, J309, J310, 38 J311, J313, J315, J316, J317, J318, J319, J320, J322, J323, J324, 39 J327, J328, J330, J331, J332, J333, J334, J335, J340, J341, E450, 40 E451 41 Postvoid urine measurement OHIP Fee code G900 42 TRUS biopsy OHIP Fee code Z712 43 Urodynamic study OHIP Fee code G192, G193, G447 44 Urine culture OHIP Fee code L633, L634, L641 Noninvasive uroflowmetry OHIP Fee code G475 45 46 Data sources included CIHIDAD (inpatient discharge abstracts) and OHIP (physician billings). CIHIDAD used ICD10 (for 47 diagnostic coding) and CCI coding (for interventions) after April 1, 2002, and ICD9/CCP coding prior to this. Specific ICD 9/10 and CCP/CCI codes are defined above. The presence of any data element within the year prior to the index date was considered 48 evidence of the medical test or intervention. 49 50 51 52 53 54 55 56 57 58 59 60 10 https://mc.manuscriptcentral.com/bmj Page 35 of 44 BMJ

1 2 eTable 3. Variables included in the propensity score. 3 4 5 Demographics 6 Year of cohort entry (based on 2 year increments) Region of residence (Using 1 of 14 Local Health Integration Networks (LHINs)) 7 Rural location 8 ComorbidConfidential: medical conditions For Review Only 9 Cancer 10 Chronic lung disease 11 Dementia 12 Diabetes 13 Hypertension 14 Osteoporosis Prior fall 15 Prostate Cancer 16 Comorbidity Index 17 John Hopkins ACG count 18 Medication use 19 ACE inhibitor or ARB 20 Androgen deprivation therapy 21 Antiandrogen therapy 22 Antibiotic Benzodiazepine 23 Bisphosphonates 24 Denosumab 25 Glucocorticoids 26 Narcotics 27 Overactive bladder medication 28 Proton Pump Inhibitors 29 Statin 30 Total medication count Hospital and physician utilization 31 Number of hospital admissions 32 Number of emergency room visits 33 Number of family physician visits 34 Number of geriatric medicine visits 35 Number of neurology visits 36 Number of cardiology visits 37 Number of urologist visits 38 Medical investigations and procedures 39 Echocardiography Bone mineral density test 40 Bone scan 41 Cardiac stress test 42 Chest Xray 43 CT head 44 Prostate specific antigen test 45 Transrectal ultrasound guided prostate biopsy 46 Urine culture 47 Urodynamic test or post void residual measurement 48 Selected based on review of standardized differences between exposed and unexposed men prior to matching, and known association with falls or fractures. 49 50 51 52 53 54 55 56 57 58 59 60 11 https://mc.manuscriptcentral.com/bmj BMJ Page 36 of 44

1 2 eTable 4. Coding definition for primary and secondary outcomes 3 4 5 Data sources Codes 6 Fall* NACRS or CIHIDAD ICD 10 W00W19 Any Fracture** NACRS or CIHIDAD or Hip (ICD 10 S720, S721; CCI 1VA73, 1VC73, 1VA74, 1VA53, 1VC74, 7 OHIP 1VA80) 8 Confidential:Forearm For (ICD 10 S52; Review CCI 1TV73, 1TV74, 1TV03; Only OHIP F014, F022, 9 F023, F025, F026, F028, F030, F032, F033, F046, F024, F027, 10 F031, Z203) 11 Humerus (ICD 10 S422, S423, S424) 12 Vertebral (S220, S221, S320, T080, T081) 13 Femur (ICD 10 S723; CCI 1VC73, 1VC74, 1VC03, 1VC80; OHIP 14 F095, F096, F097, Z211) Pelvis (ICD 10 S321, S322, S324, S323, S325, S327, S328) 15 Ankle (ICD 10 S825, S826, S827, S828, S829) 16 Patella (ICD 10 S820) 17 Tibia/Fibula (ICD 10 S821, S822, S823, S824) 18 Ribs/sternum (ICD 10 S222, S223, S224) 19 Trunk (ICD 10 S228, S29) 20 Scapula (ICD 10 S421, S29) 21 Clavicle (ICD 10 S420) 22 Other (ICD 10 M844, M843) Major Osteoporotic NACRS or CIHIDAD or Hip (ICD 10 S720, S721; CCI 1VA73, 1VC73, 1VA74, 1VA53, 1VC74, 23 fracture** OHIP 1VA80) 24 Forearm (ICD 10 S52; CCI 1TV73, 1TV74, 1TV03; OHIP F014, F022, 25 F023, F025, F026, F028, F030, F032, F033, F046, F024, F027, 26 F031, Z203) 27 Humerus (ICD 10 S422, S423, S424) 28 Pelvis (ICD 10 S321, S322, S324, S323, S325, S327, S328) 29 Hip fracture** NACRS or CIHIDAD ICD 10 S720, S721; CCI 1VA73, 1VC73, 1VA74, 1VA53, 1VC74, 30 1VA80 Hypotension*** CIHIDAD or NACRS ICD 10 I95 31 Head injury*** CIHIDAD or NACRS ICD 10 S00S09 32 CIHI DAD (hospital discharge abstracts) and NACRS (all emergency room visits) data is extracted from medical records by trained 33 abstractors, and data quality is maintained though auditfeedback systems. Data on diagnoses are coded using ICD 10, and data for 34 interventions are coded using Canadian Classification of Health Intervention (CCI). Up to 25 diagnoses and 25 interventions can be 35 coded per patient. Data elements have shown >80% agreement with medical records during reabstraction studies 1. 36 *Using medical chart abstraction as the reference standard, PPV for ICD 10 codes for falls in NACRS is 91.4% (email 37 communication, Dr LisaAnn Fraser, June 2015). **PPV for the coding combinations for fractures ranges between 0.810.96 depending on fracture site. Hip/femur fracture has an 38 2 39 estimated PPV of 0.83 . ***Using reabstracted information written in a patient’s chart as the reference standard, the codes for hypotension (I95) and head 40 injury (S06) have a sensitivity of 72% and 77% respectively, and a positive predictive value of 39% and 79% respectively 3. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 12 https://mc.manuscriptcentral.com/bmj Page 37 of 44 BMJ

1 2 eTable 5. Complete baselines for unmatched and matched cohorts. 3 4 Unmatched Cohort Matched Cohort 5 * 6 Unexposed Exposed SD Unexposed Exposed SD n=426,106 n=186,004 n=147,084 n=147,084 7 Demographics 8 Confidential: For Review Only Age Median (IQR) 73 (6979) 74 (6980) 75 (7080) 75 (7080) 9 6669.9 118718 27.9% 47798 25.7% 5% 35520 24.1% 35800 24.3% 0% 10 7074.9 116482 27.3% 46751 25.1% 5% 36549 24.8% 36306 24.7% 0% 11 7579.9 89097 20.9% 41518 22.3% 3% 33075 22.5% 33130 22.5% 0% 12 8084.9 60538 14.2% 29281 15.7% 4% 24412 16.6% 24329 16.5% 0% 13 8589.9 29644 7.0% 15171 8.2% 5% 12843 8.7% 12824 8.7% 0% 14 ≥90 11627 2.7% 5485 2.9% 1% 4685 3.2% 4695 3.2% 0% 15 Fiscal year of 20032004 35310 8.3% 18969 10.2% 7% 14002 9.5% 14598 9.9% 1% 16 cohort entry 20052006 59789 14.0% 30036 16.1% 6% 22316 15.2% 23616 16.1% 2% 17 (index date) 20072008 72225 17.0% 34202 18.4% 4% 25289 17.2% 27490 18.7% 4% 18 20092010 72156 16.9% 31478 16.9% 0% 24377 16.6% 24927 16.9% 1% 19 20112012 125153 29.4% 49057 26.4% 7% 41603 28.3% 38526 26.2% 5% 20 2013 61473 14.4% 22262 12.0% 7% 19497 13.3% 17927 12.2% 3% 21 Rural Residence 66897 15.7% 23937 12.9% 8% 20064 13.6% 19968 13.6% 0% 22 Socioeconom 1Lowest 76410 17.9% 32566 17.5% 1% 26256 17.9% 25871 17.6% 1% ic Status 23 2 86180 20.2% 37196 20.0% 0% 29685 20.2% 29543 20.1% 0% 3 84010 19.7% 36322 19.5% 1% 28675 19.5% 28804 19.6% 0% 24 4 87012 20.4% 38332 20.6% 0% 30106 20.5% 30205 20.5% 0% 25 5Highest 91099 21.4% 41056 22.1% 2% 31924 21.7% 32244 21.9% 0% 26 Missing 1395 0.3% 532 0.3% 0% 438 0.3% 417 0.3% 0% 27 Resides in long term care 13565 3.2% 4799 2.6% 4% 4068 2.8% 4068 2.8% 0% 28 Local Health 1 23315 5.5% 11464 6.2% 3% 8940 6.1% 8941 6.1% 0% 29 Integration 2 36753 8.6% 12419 6.7% 7% 10854 7.4% 10585 7.2% 1% 30 Network 3 22425 5.3% 7668 4.1% 6% 6474 4.4% 6390 4.3% 0% (LHIN) 31 4 54938 12.9% 21993 11.8% 3% 18107 12.3% 18226 12.4% 0% 32 5 19085 4.5% 9652 5.2% 3% 6992 4.8% 7282 5.0% 1% 33 6 27924 6.6% 14313 7.7% 4% 11082 7.5% 10999 7.5% 0% 34 7 30823 7.2% 16495 8.9% 6% 12191 8.3% 12286 8.4% 0% 35 8 47803 11.2% 24870 13.4% 7% 18494 12.6% 18417 12.5% 0% 36 9 52636 12.4% 20688 11.1% 4% 16832 11.4% 16596 11.3% 0% 37 10 21035 4.9% 8542 4.6% 1% 6964 4.7% 6982 4.7% 0% 38 11 39183 9.2% 17725 9.5% 1% 13964 9.5% 14016 9.5% 0% 39 12 17218 4.0% 7696 4.1% 1% 6053 4.1% 6179 4.2% 1% 40 13 23854 5.6% 9878 5.3% 1% 8000 5.4% 8035 5.5% 0% 14 9114 2.1% 2601 1.4% 5% 2137 1.5% 2150 1.5% 0% 41 Comorbid medical conditions 42 Acute Kidney injury 8913 2.1% 5611 3.0% 6% 4351 3.0% 4393 3.0% 0% 43 Alcoholism 4661 1.1% 2022 1.1% 0% 1702 1.2% 1616 1.1% 1% 44 Atrial fibrillation/flutter 27665 6.5% 13945 7.5% 4% 12034 8.2% 11391 7.7% 2% 45 Cancer 69442 16.3% 38230 20.6% 11% 32379 22.0% 30194 20.5% 4% 46 Cataract 74269 17.4% 34179 18.4% 3% 28455 19.3% 27981 19.0% 1% 47 Chronic kidney disease 32710 7.7% 16670 9.0% 5% 14429 9.8% 13425 9.1% 2% 48 Chronic liver disease 14329 3.4% 7092 3.8% 2% 5853 4.0% 5516 3.8% 1% 49 Chronic lung disease 103516 24.3% 51517 27.7% 8% 41298 28.1% 41344 28.1% 0% 50 Congestive heart failure 49675 11.7% 23908 12.9% 4% 20714 14.1% 19577 13.3% 2% 51 Coronary artery disease or 166116 39.0% 76510 41.1% 4% 63040 42.9% 62404 42.4% 1% 52 angina Dementia 38377 9.0% 18690 10.0% 3% 14590 9.9% 15461 10.5% 2% 53 Diabetes mellitus 92023 21.6% 36461 19.6% 5% 30307 20.6% 30784 20.9% 1% 54 Number of 0 422786 99.2% 183224 98.5% 7% 145058 98.6% 145036 98.6% 0% 55 previous falls 1 2986 0.7% 2460 1.3% 6% 1794 1.2% 1817 1.2% 0% 56 2 274 0.1% 268 0.1% 0% 187 0.1% 188 0.1% 0% 57 ≥3 60 0.0% 52 0.0% 0% 45 0.0% 43 0.0% 0% 58 Prior Fracture 2256 0.5% 1838 1.0% 6% 1294 0.9% 1294 0.9% 0% 59 Glaucoma 27244 6.4% 11507 6.2% 1% 9956 6.8% 9556 6.5% 1% 60 13 https://mc.manuscriptcentral.com/bmj BMJ Page 38 of 44

1 2 Hypertension 304887 71.6% 120457 64.8% 15% 102623 69.8% 102036 69.4% 1% 3 Hypotension 6068 1.4% 3310 1.8% 3% 2796 1.9% 2615 1.8% 1% 4 Macular degeneration 12028 2.8% 5097 2.7% 1% 4428 3.0% 4144 2.8% 1% 5 Obesity (BMI >40) 18567 4.4% 8740 4.7% 1% 6854 4.7% 6835 4.6% 0% 6 Osteoporosis 21179 5.0% 11257 6.1% 5% 9010 6.1% 8795 6.0% 0% 7 Parkinson's disease 2251 0.5% 1477 0.8% 4% 1113 0.8% 1169 0.8% 0% Peripheral vascular disease 9107 2.1% 4017 2.2% 1% 3691 2.5% 3331 2.3% 1% 8 Confidential: For Review Only Prostate cancer 39758 9.3% 21332 11.5% 7% 19112 13.0% 17173 11.7% 4% 9 Rheumatoid arthritis 15407 3.6% 7734 4.2% 3% 6248 4.2% 6216 4.2% 0% 10 Seizure 2599 0.6% 1306 0.7% 1% 1090 0.7% 1040 0.7% 0% 11 Stroke or TIA 12592 3.0% 6687 3.6% 3% 5195 3.5% 5368 3.6% 1% 12 Comorbidity Index (based on 3 year lookback window) 13 Charlson Median (IQR) 0 (02) 0 (02) 0 (02) 0 (02) 14 comorbidity 0 325765 76.5% 134274 72.2% 10% 102838 69.9% 105567 71.8% 4% 15 index 1 38987 9.1% 18379 9.9% 3% 15051 10.2% 14841 10.1% 0% 16 2 31279 7.3% 16564 8.9% 6% 14598 9.9% 13138 8.9% 3% 17 ≥3 30075 7.1% 16787 9.0% 7% 14597 9.9% 13538 9.2% 2% 18 Number of Median (IQR) 7 (510) 9 (611) 8 (611) 19 Johns 04 103199 24.2% 25020 13.5% 28% 19630 13.3% 19152 13.0% 1% Hopkins ACG 59 97005 22.8% 61420 33.0% 7% 68262 46.4% 70116.00 47.7% 2% 20 Aggregated 21 Diagnosis 1014 14866 3.5% 13165.00 7.1% 23% 48762 33.2% 47961.00 32.6% 1% 22 Groups 1519 424 0.1% 565 0.3% 16% 10085 6.9% 9482 6.4% 2% 23 ≥20 210612 49.4% 85834 46.1% 4% 345 0.2% 373 0.3% 2% Medication use (in the 180 days prior to the index date) 24 5 alpha reductase inhibitor 10838 2.5% 20474 11.0% 34% 10826 7.4% 10826 7.4% 0% 25 ACE inhibitors or ARBs 218787 51.3% 85697 46.1% 10% 72764 49.5% 72529 49.3% 0% 26 Androgen deprivation therapy 6728 1.6% 5120 2.8% 8% 4051 2.8% 3974 2.7% 1% 27 Anti -inflammatory 61927 14.5% 29786 16.0% 4% 23196 15.8% 24681 16.8% 3% 28 Antiandrogens 4479 1.1% 3676 2.0% 7% 2736 1.9% 2743 1.9% 0% 29 Antibiotic 118029 27.7% 71863 38.6% 23% 54917 37.3% 56281 38.3% 2% 30 Anticoagulant 5411 1.3% 2526 1.4% 1% 2372 1.6% 2047 1.4% 2% 31 Anticonvulsants 18254 4.3% 9170 4.9% 3% 7411 5.0% 7429 5.1% 0% 32 Antidepressants 25650 6.0% 13242 7.1% 4% 10568 7.2% 10846 7.4% 1% 33 Antineoplastic 15852 3.7% 9293 5.0% 6% 8139 5.5% 7419 5.0% 2% 34 Antiparkinson drug 8326 2.0% 4813 2.6% 4% 3552 2.4% 3970 2.7% 2% 35 Antiplatelet 27649 6.5% 12724 6.8% 1% 10700 7.3% 10561 7.2% 0% 36 Antipsychotic (atypical) 12243 2.9% 4917 2.6% 2% 4572 3.1% 4153 2.8% 2% 37 Antipsychotic (other) 3644 0.9% 1854 1.0% 1% 1663 1.1% 1497 1.0% 1% 38 Benzodiazepine 49030 11.5% 25979 14.0% 8% 20868 14.2% 21220 14.4% 1% 39 Beta blockers 135582 31.8% 53959 29.0% 6% 47435 32.3% 45448 30.9% 3% 40 Bisphosphonate 22573 5.3% 12004 6.5% 5% 10024 6.8% 9590 6.5% 1% Calcium channel blockers 111886 26.3% 47549 25.6% 2% 39507 26.9% 40141 27.3% 1% 41 Cholinesterase inhibitors 13344 3.1% 5727 3.1% 0% 5169 3.5% 4893 3.3% 1% 42 Denosumab 93 0.0% 41 0.0% 5% 41 0.0% 34 0.0% 1% 43 Digoxin 14499 3.4% 6405 3.4% 0% 6062 4.1% 5378 3.7% 2% 44 Glucocorticoids 32037 7.5% 17262 9.3% 6% 14149 9.6% 14000 9.5% 0% 45 Glucose test strip 4736 1.1% 1939 1.0% 0% 1626 1.1% 1604 1.1% 0% 46 Inhaled acetylcholine 26936 6.3% 13319 7.2% 4% 10994 7.5% 11138 7.6% 0% 47 Inhaled beta -agonist 47862 11.2% 22750 12.2% 3% 19544 13.3% 18805 12.8% 1% 48 Inhaled corticosteroid 21522 5.1% 10007 5.4% 1% 8764 6.0% 8145 5.5% 2% 49 Narcotics 63767 15.0% 35285 19.0% 11% 28254 19.2% 28523 19.4% 1% 50 Non -potassium sparing 107470 25.2% 41600 22.4% 7% 38851 26.4% 35062 23.8% 6% 51 diuretics Overactive bladder medication 5342 1.3% 6439 3.5% 14% 4057 2.8% 4411 3.0% 1% 52 Potassium sparing diuretics 16418 3.9% 6306 3.4% 3% 6224 4.2% 5321 3.6% 3% 53 Proton Pump inhibitors 91823 21.5% 45558 24.5% 7% 38025 25.9% 37415 25.4% 1% 54 Selective Serotonin Reuptake 25783 6.1% 12471 6.7% 2% 10368 7.0% 10315 7.0% 0% 55 Inhibitors 56 Smoking cessation aid 501 0.1% 200 0.1% 0% 151 0.1% 158 0.1% 0% 57 Statins 211307 49.6% 85185 45.8% 8% 71769 48.8% 71449 48.6% 0% 58 Testosterone replacement 2401 0.6% 1326 0.7% 1% 996 0.7% 1064 0.7% 0% 59 Number of Median (IQR) 6 (39) 6 (310) 7 (410) 7 (410) 60 14 https://mc.manuscriptcentral.com/bmj Page 39 of 44 BMJ

1 2 unique 0 0 0.0% 10793 35% 35% 0 0.0% 0 0.0% 0% 3 medications 14 163603 38.4% 53418 21% 21% 45795 31.1% 47086 32.0% 2% 4 58 147738 34.7% 58162.00 7% 7% 50084 34.1% 49144.00 33.4% 1% 5 912 71489 16.8% 35343.00 6% 6% 29317 19.9% 28799.00 19.6% 1% 6 1316 27772 6.5% 16619 9% 9% 13350 9.1% 13150 8.9% 1% 7 ≥17 15504 3.6% 11669 12% 12% 8538 5.8% 8905 6.1% 1% Hospital and Physician Utilisation 8 Confidential: For Review Only Number of Median (IQR) 0 (01) 0 (01) 0 (01) 9 hospital 0 304363 71.4% 109456 58.8% 27% 89192 60.6% 88795 60.4% 0% 10 admissions 1 74362 17.5% 43584 23.4% 15% 32978 22.4% 33218 22.6% 0% 11 2 28981 6.8% 18345 9.9% 11% 14010 9.5% 14120 9.6% 0% 12 ≥3 18400 4.3% 14619 7.9% 15% 10904 7.4% 10951 7.4% 0% 13 Number of Median (IQR) 0 (01) 0 (01) 0 (01) 14 Emergency 0 306292 71.9% 115537 62.1% 21% 92414 62.8% 91941 62.5% 1% 15 Room visits 1 70553 16.6% 36929 19.9% 9% 29150 19.8% 29305 19.9% 0% 16 2 25618 6.0% 16289 8.8% 11% 12612 8.6% 12679 8.6% 0% 17 ≥3 23643 5.5% 17249 9.3% 15% 12908 8.8% 13159 8.9% 0% 18 Number of Median (IQR) 6 (311) 7 (413) 7 (413) 19 family 0 17514 4.1% 5791 3.1% 5% 3918 2.7% 3967 2.7% 0% physician 12 54261 12.7% 17635 9.5% 10% 12899 8.8% 12995 8.8% 0% 20 visits 21 34 79480 18.7% 27662 14.9% 10% 22140 15.1% 21980 14.9% 1% 22 56 75196 17.6% 30102 16.2% 4% 24651 16.8% 24344 16.6% 1% 23 78 54448 12.8% 24611 13.2% 1% 20198 13.7% 19881 13.5% 1% 24 910 36712 8.6% 18061 9.7% 4% 14772 10.0% 14506 9.9% 0% ≥11 108495 25.5% 62142 33.4% 17% 48506 33.0% 49411 33.6% 1% 25 Number of Median (IQR) 0 (00) 0 (00) 0 (00) 26 Geriatrician 0 415069 97.4% 177844 95.6% 10% 141198 96.0% 140922 95.8% 1% 27 visits 1 5097 1.2% 3352 1.8% 5% 2503 1.7% 2594 1.8% 1% 28 2 2418 0.6% 1671 0.9% 3% 1234 0.8% 1257 0.9% 1% 29 ≥3 3522 0.8% 3137 1.7% 8% 2149 1.5% 2311 1.6% 1% 30 Number of Median (IQR) 0 (00) 0 (00) 0 (00) 31 Neurologist 0 401060 94.1% 169928 91.4% 10% 135241 91.9% 134846 91.7% 1% 32 visits 1 13397 3.1% 7847 4.2% 6% 5974 4.1% 6104 4.2% 1% 33 2 6448 1.5% 4112 2.2% 5% 3053 2.1% 3127 2.1% 0% 34 ≥3 5201 1.2% 4117 2.2% 8% 2816 1.9% 3007 2.0% 1% 35 Number of Median (IQR) 0 (01) 0 (01) 0 (02) 36 Cardiologist 0 265403 62.3% 103906 55.9% 13% 81840 55.6% 82042 55.8% 0% visits 37 1 74142 17.4% 35826 19.3% 5% 28238 19.2% 28237 19.2% 0% 38 2 33092 7.8% 16785 9.0% 4% 13555 9.2% 13379 9.1% 0% 39 ≥3 53469 12.5% 29487 15.9% 10% 23451 15.9% 23426 15.9% 0% 40 Number of Median (IQR) 0 (00) 0 (02) 0 (01) Urologist 0 357268 83.8% 100518 54.0% 68% 89489 60.8% 90134 61.3% 1% 41 visits 1 27955 6.6% 32619 17.5% 34% 23241 15.8% 22430 15.2% 2% 42 2 17502 4.1% 21662 11.6% 28% 14517 9.9% 14257 9.7% 1% 43 ≥3 23381 5.5% 31205 16.8% 36% 19837 13.5% 20263 13.8% 1% 44 Number of Median (IQR) 2 (14) 2 (14) 2 (14) 45 Ophthalmolo- 0 306825 72.0% 129425 69.6% 5% 100668 68.4% 101731 69.2% 2% 46 gist visits 1 47864 11.2% 23761 12.8% 5% 18585 12.6% 18854 12.8% 1% 47 2 25373 6.0% 12011 6.5% 2% 9928 6.7% 9573 6.5% 1% 48 ≥3 46044 10.8% 20807 11.2% 1% 17903 12.2% 16926 11.5% 2% 49 Medical investigations and procedures 50 Bone mineral density test 15490 3.6% 8704 4.7% 6% 6842 4.7% 6636 4.5% 1% 51 Bone scan 9798 2.3% 9465 5.1% 15% 6446 4.4% 6668 4.5% 0% 52 Cardiac catheterization 8512 2.0% 4377 2.4% 3% 3373 2.3% 3565 2.4% 1% 53 Cardiac stress test 51037 12.0% 26955 14.5% 7% 21126 14.4% 21066 14.3% 0% 54 Carotid endarterectomy 402 0.1% 241 0.1% 0% 176 0.1% 191 0.1% 0% 55 Carotid ultrasound 20098 4.7% 10647 5.7% 5% 8495 5.8% 8466 5.8% 0% 56 Chest x -ray 128826 30.2% 71018 38.2% 17% 55893 38.0% 55992 38.1% 0% CT head 57 29622 7.0% 18993 10.2% 11% 14361 9.8% 14752 10.0% 1% CT spine 5304 1.2% 3833 2.1% 7% 2638 1.8% 2930 2.0% 1% 58 Echocardiography 73390 17.2% 37689 20.3% 8% 30330 20.6% 29944 20.4% 0% 59 60 15 https://mc.manuscriptcentral.com/bmj BMJ Page 40 of 44

1 2 EEG 2075 0.5% 1322 0.7% 3% 1042 0.7% 980 0.7% 0% 3 Heart valve replacement 622 0.1% 408 0.2% 3% 253 0.2% 329 0.2% 0% 4 Holter monitoring 24508 5.8% 13320 7.2% 6% 10305 7.0% 10578 7.2% 1% 5 Prostate -specific antigen test 39022 9.2% 28331 15.2% 18% 22420 15.2% 20168 13.7% 4% 6 Pulmonary function test 35519 8.3% 18704 10.1% 6% 15006 10.2% 14784 10.1% 0% 7 Postvoid urine measurement 4606 1.1% 8964 4.8% 22% 4490 3.1% 4689 3.2% 1% TRUS biopsy 4116 1.0% 8570 4.6% 22% 4026 2.7% 4710 3.2% 3% 8 Confidential: For Review Only Urodynamic study 462 0.1% 1087 0.6% 8% 451 0.3% 569 0.4% 2% 9 Urine culture 56320 13.2% 58941 31.7% 45% 39070 26.6% 40731 27.7% 2% 10 Noninvasive uroflowmetry 9192 2.2% 19372 10.4% 34% 8960 6.1% 9968 6.8% 3% 11 12 *Standardized differences (SD) describe differences in between group means relative to a pooled standard deviation, and better demonstrate significant differences in large samples. A SD > 10% is considered a meaningfully difference between groups. 13 14

15 16

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 16 https://mc.manuscriptcentral.com/bmj Page 41 of 44 BMJ

1 2 eTable 6. Study alpha antagonists and their doses and prescription details for the 147,084 3 exposed men. 4 5 6 7 Tamsulosin Silodosin Alfuzosin 8 Number (%)Confidential: 123537 (84.0%) For 2720 (1.8%) Review 20827 (14.2%)Only 9 Median Dose (IQR) 0.4mg (0.40.4) 8mg (88) 10mg (1010) 10 11 0.4mg: 114509 (92.7%) 4mg: 354 (13.0%) 10mg: 20418 (98.0%) 12 0.8mg: 6802 (5.5%) 8mg: 2324 (85.4%) 20mg: 135 (0.6%) 13 Other: 2226 (1.8%) Other: 42 (1.5%) Other: 274 (1.3%) 14 Labelled dose 0.4mg 4mg or 8mg 10mg 15 16 Included formulations Tamsulosin (generic), Rapaflo™ Alfuzosin (generic), and 17 Flomax™, and Flomax CR™ Xatral™ 18 Median Prescription 30 (3060) 30 (3030) 30 (3030) length, days (IQR) 19 Number (%) of men that 59558 (48.2%) 1465 (53.9%) 10174 (48.9%) 20 refill a prescription within 21 7 days of predicted end 22 date of initial prescription 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 17 https://mc.manuscriptcentral.com/bmj BMJ Page 42 of 44

1 2 eTable 7. Primary outcomes among exposed and unexposed patients in the 180 days prior to 3 alpha antagonist initiation or index date. 4 5 6 Odds Ratio Change in Percent Events n (%) 7 (95% CI) Absolute Risk (95% CI) 8 Confidential: For Review Only Future alpha Future nonalpha 9 antagonist users antagonist users 10 (Exposed cohort) (Unexposed cohort) 11 n=119,831 n=119,831 12 13 Primary outcome 14 Fall 1469 (1.23%) 1595 (1.33%) 0.92 (0.860.99) 0.11 (0.20 to 0.02) 15 Fracture 495 (0.41%) 577 (0.48%) 0.86 (0.760.97) 0.07 (0.12 to 0.01) 16 Exclusion criteria were reapplied to the study cohort based on a new index date 180 days prior to primary study index date (initiation 17 of alpha antagonist). All eligible matched pairs who were not excluded were retained. The primary outcomes of 90day risk of fall 18 and fracture were then determined to assess for a potential baseline increased risk of fall and fracture due to lower urinary tract 19 symptoms among men who go on to initiate alpha antagonist therapy in the future. 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 18 https://mc.manuscriptcentral.com/bmj Page 43 of 44 BMJ

1 2 eFigure 1. Flow diagram of cohort selection. 3 4 5 6 7 Exposed Cohort Unexposed Cohort 8 Confidential: For Review Only 9 n=249,689 n=695,223 10 11 12 13 14 15 16 Excluded: 17 Excluded: Death prior to index date n=15,365 18 Death prior to index date n=18 19 Prescription for >1 second generation No Ontario drug benefits prescription in 20 alpha blocker on index date n=401 90 days prior to index date n=205,560 21 Prescription for ≥ 1 second generation Prescription for ≥ 1 second generation 22 alpha blocker in the 180 days prior to alpha blocker on or 180 days prior to 23 index date n=18,329 index date n=11,644 24 Prescription for a first generation alpha Prescription for a first generation alpha 25 blocker in the 180 days prior to index blocker in the 180 days prior to index 26 date n=25,974 date n=31,554 27 Hospital discharge in the 2 days prior to Hospital discharge in the 2 days prior to 28 index date n=11,471 index date n=3,367 29 Emergency room visit in the 2 days Emergency room visit in the 2 days 30 prior to index date n=7,492 prior to index date n=1,627 31 32 33 34 35 36 37 38 39 40 Remaining Remaining 41 n=186,004 n=426,106 42 43 44 45 46 Matching 1:1 (without replacement) 47 1) Age at index date +/ 2 years 48 2) Residential status (community versus long term care) 49 3) Prior fracture in the last 365 days 50 4) 5 alpha reductase inhibitor usage 51 5) Logit of the propensity score +/ 0.2 times the standard deviation of the logit 52 53 54 55 56 Final exposed cohort Final unexposed cohort 57 n=147,084 n=147,084 58 59 60 19 https://mc.manuscriptcentral.com/bmj BMJ Page 44 of 44

1 2 References 3 4 5 6 7 1. Roos LL, Gupta S, Soodeen R-A, Jebamani L. Data quality in an information-rich environment: Canada 8 as anConfidential: example. Can J Aging . 2005;24 Suppl For 1:153–170. Review Only 9 10 2. Jean S, Candas B, Belzile É, et al. Algorithms can be used to identify fragility fracture cases in 11 physician-claims databases. Osteoporos Int . 2012;23(2):483–501. 12 13 3. Juurlink D, Preyra C, Croxford R, Chong A, Austin P, Tu J, et al. Canadian Institute for Health 14 15 Information Discharge Abstract Database: a validation study. Toronto, Ontario, Canada: Institute for 16 Clinical Evaluative Sciences; 2006. Accessed at www.ices.on.ca/flip-publication/canadian-istitute-for- 17 health -information-discharge/index.html#41/z on April 20, 2015.) 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 20 https://mc.manuscriptcentral.com/bmj