Osteoporosis International (2022) 33:123–137 https://doi.org/10.1007/s00198-021-06077-0 ORIGINAL ARTICLE One‑ and 2‑year incidence of osteoporotic fracture: a multi‑cohort observational study using routinely collected real‑world data Sara Khalid1 · Carlen Reyes2,3 · Martin Ernst4 · Antonella Delmestri1 · Emese Toth5 · Cesar Libanati5 · Bo Abrahamsen1,6,7 · Daniel Prieto‑Alhambra1,2,3,8 Received: 4 January 2021 / Accepted: 19 July 2021 / Published online: 15 August 2021 © The Author(s) 2021 Abstract Summary We estimated and characterized the imminent fracture risk (1–2 years) of high-risk fracture patients through a multinational (UK, Spain, Denmark) cohort study. Older individuals with newly diagnosed osteoporosis and individuals who had a fracture while on treatment with a bisphosphonate were at a high risk of imminent fracture. Purpose To characterize and estimate 1- to 2-year fracture risk in high-risk fracture patients. Methods Multi-cohort study in (database/study period) UK (CPRD/1995–2017), Spain (SIDIAP/2006–2016) and Denmark (DHR/1995–2016) including individuals ≥ 50 years old in NDO (newly diagnosed osteoporosis), OFx (incident osteoporotic fracture), BP (incident oral bisphosphonates use) or FWOT (fracture while on treatment with bisphosphonates). Outcomes (ICD-10/READ): hip, clinical spine, non-hip, non-spine and hip/humerus/distal forearm fracture. Follow-up: from cohort entry until death, migration/transfer or end of the study. Statistics: baseline characteristics and incidence rate (IR per 1000 persons). Results (1‑year IR) NDO included 69,899 (UK), 37,901 (Spain) and 158,191 (Denmark) individuals. Spanish-IR was low- est for hip (4.7), clinical spine (2.5) and major osteoporotic fracture (MOF) (17.3) and highest in Denmark (74.2, 26.0 and 120.1, respectively). OFx included 83,514 (UK), 51,044 (Spain) and 509,551 (Denmark) individuals. IR in Denmark was highest for hip (24.1) and MOF (47.2), in Spain was highest for the clinical spine (9.4) and lowest for hip (9.5) and in the UK was lowest for the clinical spine (2.8) and MOF (20.7). BP included 148,507 (UK), 52,037 (Spain) and 204,010 (Denmark) individuals. Spanish-IR was lowest for hip (5.0) and MOF (21.1) and highest in Denmark (20.3 and 48.6, respectively). FWOT included 28,930 (UK), 1,865 (Spain) and 31,882 (Denmark) individuals. Clinical spine-IR was highest for Spain (12.0). Hip-IR was lowest for Spain (7.6) and highest for Denmark (33.6). Comparing young subjects, those who have FWOT started with an increased fracture rate. Conclusion OFx and FWOT individuals experience higher re-fracture incidence rates than those with osteoporosis with or without treatment. Keywords Osteoporosis · General population studies · Anti-resorptive · Fracture prevention 5 * Daniel Prieto‑Alhambra UCB Biopharma Sprl, Brussels, Belgium Daniel.prietoalhambra@ndorms.ox.ac.uk 6 Open Patient Data Explorative Network, University 1 of Southern Denmark and Odense University Hospital, Nuffield Department of Orthopaedics, Rheumatology Odense, Denmark and Musculoskeletal Sciences, Centre for Statistics 7 in Medicine, Nuffield Orthopaedic Centre, Oxford, UK Department of Medicine, Holbæk Hospital, Holbæk, 2 Denmark Fundació Institut Universitari per a la recerca a ÍAtenció 8 Primària de Salut Jordi Gol i Gorina (IDIAPJ Gol), MRC Lifecourse Epidemiology Unit, Southampton General CIBERFES, Barcelona, Spain Hospital, University of Southampton, Southampton, UK 3 Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain 4 Department of Public Health, Clinical Pharmacology and Pharmacy, University of Southern Denmark, Odense, Denmark 13 Vol.:(0123456789) 124 Osteoporosis International (2022) 33:123–137 Introduction bisphosphonates, we selected patients with an incident fracture while using bisphosphonate treatment. Osteoporotic fractures entail a social and economic bur- den for populations and health care systems worldwide Materials and methods [1]. Costs reached €37 billion in Europe already by 2010 [2]. In addition to societal costs, there is also the question We conducted a retrospective cohort study using data from of the individual implications of these fractures, such as three European electronic health care databases from the an increased disability and mortality of the individuals United Kingdom (UK), Denmark and Spain. The databases affected [1]. Effective management centred on high frac- used were as follows: ture risk patients would improve outcomes and reduce the burden of fragility fractures. (1) The Clinical Practice Research Datalink (CPRD) from In this context, the identification of the population at the UK, which contains anonymized and computer- greater risk of fracture has been the focus of extensive ized primary care outpatient records for a representa- research in the last decades. In recent years, particular tive sample of the UK population (> 7 million patients, attention has been drawn to the analysis of the imminent approximately 8% of UK population) in addition to risk of fracture, defined as the markedly increased risk in demographic information, medication prescriptions, the 2-year period that follows a fragility fracture [3–9]. clinical events, referrals and hospital admissions. Pre- Many risk factors for fracture have been identified; vious reports have demonstrated the availability and however, while some of them appear to confer an increased quality of CPRD [14]. For completeness and improved risk at any time (age [3–5, 7], previous falls [4, 8, 9], low data quality, we linked CPRD to the Hospital Episode bone mineral density [10], previous fractures and most Statistics Admitted Patient Care, which contains clini- specifically vertebral fractures [3, 5, 8, 9]), others (comor- cal diagnoses during hospital admissions in England, bidities [3–5], use of benzodiazepines [4, 5] or nitrates [8]) to the Office for National Statistics mortality records seem to have a significant impact in the short term. and to the Index of Multiple Deprivation dataset. Fracture risk can be quantified with risk assessment (2) The Sistema d’informació per al desenvolupament de tools such as FRAX [11] and QFracture [12]. However, la investigació en Atenció primaria (SIDIAP) which these tools focus on the 5- or 10-year future risk and are comprises primary health care anonymised electronic untested for imminent risk of fracture. Scaling risk to medical records for a representative > 80% of the popu- a different time period is more challenging than it may lation of Catalonia, as well as demographic informa- seem at first—risk factors might not remain constant over tion, clinical events (ICD10 codes), prescriptions and a 10-year period, with a higher risk of fracture during the community pharmacy dispensations. Previous reports first year after the index fracture. A population-based study have demonstrated the availability and quality data of carried out in postmenopausal women found a higher rela- SIDIAP [15]. In order to avoid incomplete fracture tive risk of fracture in the first year with a further decrease records, we linked SIDIAP to regional hospital admis- in the following years [13]. To design effective preventive sions data available for those hospitals part of the same strategies and help the early post-fracture management of trust/health care provider covering around 30% of the affected individuals, it is therefore important to accurately Catalan population (CMBD-AH ICS for its acronym in identify the population with an imminent risk of fracture. the Catalan language). In this context, we aimed to characterize patients at (3) The Danish Health Registry (Lægemiddel-registret) high imminent risk of fracture and to estimate their 1- [16], which contains all filled prescriptions in the and 2-year risk of fracture. We considered a number of country, is linked to the National Hospital Discharge scenarios with an increased risk of fracture and selected Register (Landspatientsregistret—all diagnosis codes appropriate cohorts for assessment. In the cohort of newly and treatments for inpatients and outpatients) and to diagnosed osteoporosis, we sought to identify rates of the National Cause of Death Register (Dødsårsagsreg- imminent fracture in patients following their first assign- istret). It represents the total Danish population, and ment of an osteoporosis diagnosis. In the cohort of inci- previous reports have demonstrated the validity of the dent osteoporosis fracture, we focused on patients experi- DHR database for epidemiological studies [17]. encing their first incident osteoporotic fracture and again focused on subsequent event rates for new fractures. In Participants the cohort of incident oral bisphosphonates, we included incident users of oral bisphosphonate therapy, and in The source population for this study included individuals the cohort of incident fracture while on treatment with aged at least 50 years old at cohort entry, with at least 1 year of patient records available prior to inclusion. 13 Osteoporosis International (2022) 33:123–137 125 The study period for CPRD was from 1 January 1995 forearm during the first and second years of follow-up. Any to 31 January 2017; for SIDIAP from 1 January 2006 to two fractures of the same bone/site coded within 6 months 31 December 2016 and for DHR from 1 January 1995 to of each other were considered duplicate records, and the one 31 December 2016. The study population included cohorts recorded the latest was dismissed in regard to the primary of individuals at potentially high fracture risk, as identified outcome. from the literature, divided into 4 cohorts (Online Resource The aim of the aforementioned groups of outcomes is to 1). capture all types of fractures; however, for clinical purposes, these groups are not practical. For this reason, we grouped Cohorts (non‑exclusive) the outcomes considered as ‘major osteoporotic fractures’ (MOF), which included fractures at the hip/humerus/distal (1) NDO (newly diagnosed osteoporosis) patients with forearm and clinical spine. an incident recorded diagnosis of osteoporosis which For each study outcome, patients were followed from the might include those with or without a fracture (diag- index date until the earliest of the study outcome, death, nosed through read codes or ICD-10 codes). migration/transfer out and end of the study period (based (2) OFx (osteoporotic fracture) patients with a first incident on data availability at extraction date), whichever came first. fracture (all except face, skull and digits), diagnosed Covariates of interest included risk factors of fracture either through read codes or ICD-10 codes and without studied in the literature and those included in the QFracture fracture coded at the same anatomical site in the prior [12] and FRAX [11] risk score calculators: age, sex, ethnic- 6 months. ity, smoking status, alcohol drinking, diabetes mellitus, fam- (3) BP (oral bisphosphonates use) incident users of oral ily history of osteoporosis/hip fracture, residence in a nurs- bisphosphonates (ATC 05B*) without bisphosphonate ing/care home, previous history of distal forearm/spine/hip/ use in the prior year. shoulder fracture, history of falls, dementia, cancer, asthma (4) FWOT (fracture while on treatment) patients with an or COPD, ischaemic heart disease, chronic liver disease, incident fracture (as in the osteoporotic fracture cohort) chronic kidney disease (stage 4+), Parkinson’s disease, while on treatment with oral bisphosphonates. This rheumatoid arthritis or SLE, malabsorption and endocrine cohort was defined by a fracture diagnosed after bis- problems including thyrotoxicosis, hyperparathyroidism, phosphonate initiation and during continued therapy Cushing’s, epilepsy or use of anticonvulsants, use of anti- (with a gap of 90 days between dispensations allowed) depressants, steroid tablets use, use of oestrogen or hormone and without fracture at the same site in the prior replacement therapy body mass index, weight, height, any 6 months. previous fracture, secondary osteoporosis, recent fracture (in the previous 6 months and 1 year), use of proton pump For a given individual, the date of entry into a cohort inhibitors, vitamin D deficiency, previously described pre- was defined as the index date (Table S2). Participants could dictors of imminent fracture and comorbidities: ankylosing potentially be present in more than one of the cohorts earlier, spondylitis, anxiety disorders, Charlson co-morbidity index with different index dates. and medication: previous anti-osteoporosis drug use, hepa- rin, beta-blockers, nitrates and calcium channel blockers. Exclusion criteria In terms of the look-back period, we considered all med- ical history, e.g. if diabetes mellitus was ever diagnosed, Individuals with Paget disease of bone and those with a his- it was coded ‘yes’. For smoking, drinking, BMI, nursing tory of breast or prostate cancer (at any time before the index home, falls and medication, a look-back period of 1 year date) were excluded. Furthermore, those with the use of any was used, given that the most recent recorded status would anti-osteoporotic drug (except Calcium and Vitamin D sup- be the most correct at baseline. For fractures, a look-back plements) in the previous year were also excluded from the period of 6 months, 1 year and ever before the index date BP cohort. (baseline) were all included. Exposures, outcomes and covariates Statistical analysis All study variables (exposures, outcomes and covariates) Baseline demographics were reported. These included socio- were identified using pre-specified lists of diagnosis codes economics, medical history and fracture risk factors strati- (ICD10, READ and ICD9) and medication codes (ATC and fied by cohort and country. BNF/Prodcodes) (Online Resource 2). For each outcome, the 1-year and 2-year incidence The main study outcome was the first fracture of the hip, rates were calculated. The incidence rate was defined as clinical spine, non-hip, non-spine and hip/humerus/distal the number of events of the outcome of interest divided 13 126 Osteoporosis International (2022) 33:123–137 by the summation of patient years contributed by all Results patients in the cohort. Plots of the 1- and 2-year cumula- tive incidence (95% CI) function were also produced for NDO: newly diagnosed osteoporosis patients (with each outcome. or without a fracture) The NDO cohort consisted of a total of 69,899 from the Sub‑group analysis CPRD database, 37,901 from the SIDIAP database and 158,191 individuals from the DHR database (Fig. 1). One-year incidence rates were also calculated for each Baseline characteristics of the population included are of the following sub-groups: age groups, type of index reported in Table 1. The proportion of men among these fracture (hip and clinical spine fracture); and for the OBP newly diagnosed patients was low, especially in Spain cohort only, previous fracture history. (8.6%). There were also pronounced differences in comor- A common data dictionary was used to extract vari- bidity between the countries with Denmark having the low- ables from all three databases, and a single R script was est proportion of individuals with > 3 points in the Charlson used to run the analyses in all three databases; all analy- Comorbidity Index. ses were performed using RStudio version 1.1.380. Overall, 1-year incidence rates of all individual fracture categories (Table 2) were highest in Denmark and lowest in Fig. 1  NDO, OFx, BP and FWOT study cohorts 13 Table 1  Baseline characteristics of the individuals included in the study NDOa Ofxb BPc FWOTd CPRD SIDIAP DHR CPRD SIDIAP DHR CPRD SIDIAP DHR CPRD SIDIAP DHR N 69,899 37,901 158,191 83,514 51,044 509,551 148,507 52,037 204,010 28,930 1865 31,882 Age, mean 72.7 (10.6) 69.4 (10.6) 72.6 (11.0) 78.9 (11.2) 72.2 (12.3) 70.8 (12.9) 74.2 (10.8) 69.6 (10.3) 71.0 (10.0) 79.6 (9.7) 74.1 (9.8) 75.9 (10.1) (SD) Men (%) 8870 (12.7) 3242 (8.6) 28,062 18,547 14,841 157,307 29,034 9820 (18.9) 37,122 3020 (10.4) 164 (8.8) 3543 (11.1) (18.7) (22.2) (29.1) (30.9) (19.6) (18.2) Comorbid conditions Diabetes 4534 (6.5) 4652 (12.3) 9856 (6.2) 7698 (9.2) 10,683 30,789 (6.0) 11,136 (7.5) 7723 (14.8) 11,144 (5.5) 2747 (9.5) 336 (18.0) 2185 (6.9) (20.9) Osteoporosis International (2022) 33:123–137 Osteopo- – – – 8,017 (9.6) 8,013 (15.7) 22,929 (4.5) 24,206 13,217 69,363 13,828 986 (52.9) 15,941 (50) rosis (16.3) (25.4) (34.0) (47.8) Dementia 1325 (1.9) 906 (2.4) 3621 (2.3) 6398 (7.7) 3551 (7.0) 15,004 (2.9) 3433 (2.3) 1242 (2.4) 3375 (1.7) 2142 (7.4) 102 (5.5) 1539 (4.8) Cancer 5271 (7.5) 1785 (4.7) 13,972 (8.8) 7391 (8.9) 3216 (6.3) 35,686 (7.0) 11,613 (7.8) 2719 (5.2) 16,930 (8.3) 3560 (12.3) 113 (6.1) 3166 (9.9) Asthma 7717 (11.0) 1996 (5.3) 7562 (4.8) 6314 (7.6) 2461 (4.8) 11,356 (2.2) 15,648 2687 (5.2) 9973 (4.9) 3844 (13.3) 155 (8.3) 1632 (5.1) (10.5) ­COPDe 5002 (7.2) 1691 (4.5) 20,256 4805 (5.8) 3230 (6.3) 26,965 (5.3) 11,360 (7.6) 2650 (5.1) 25,357 2815 (9.7) 120 (6.4) 4300 (13.5) (12.8) (12.4) Chronic 623 (0.9) 1743 (4.6) 3418 (2.2) 566 (0.7) 2946 (5.8) 7934 (1.6) 1135 (0.8) 2276 (4.4) 4164 (2.0) 307 (1.1) 104 (5.6) 925 (2.9) liver disease Chronic 5467 (7.8) 1418 (3.7) 2581 (1.6) 9899 (11.9) 4016 (7.9) 6254 (1.2) 12,613 (8.5) 2054 (3.9) 2145 (1.1) 5134 (17.7) 124 (6.6) 451 (1.4) kidney disease Parkinson’s 666 (1.0) 805 (2.1) 2371 (1.5) 1472 (1.8) 1866 (3.7) 6558 (1.3) 1330 (0.9) 1142 (2.2) 2547 (1.2) 544 (1.9) 84 (4.5) 770 (2.4) disease Rheu- 5466 (7.8) 818 (2.2) 6532 (4.1) 3949 (4.7) 682 (1.3) 8853 (1.7) 17,947 1188 (2.3) 10,854 (5.3) 4579 (15.8) 58 (3.1) 2147 (6.7) matoid (12.1) arthritis Lupus 154 (0.2) 86 (0.2) 420 (0.3) 74 (0.1) 81 (0.2) 478 (0.1) 333 (0.2) 113 (0.2) 696 (0.3) 82 (0.3) 7 (0.4) 126 (0.4) Epilepsy 1277 (1.8) 319 (0.8) 3090 (2.0) 1834 (2.2) 760 (1.5) 9139 (1.8) 2308 (1.6) 476 (0.9) 3588 (1.8) 816 (2.8) 28 (1.5) 965 (3.0) Prior frac- 10,591 1982 (5.2) 24,351 0 (0.0)* 1 (0.0) 0 (0.0) 22,843 2902 (5.6) 27,871 2149 (7.4) 0 (0.0%) 3390 (10.6) tures, N (15.2) (15.4) (15.4) (13.7) (%)f Type of index fracture, N (%) Hip – – – 56,500 8577 (16.8) 106,997 – – – 8790 (30.4) 256 (13.6) 7735 (24.3) (67.6) (21.0) Clinical – – – 3502 (4.2) 2813 (5.1) 15,480 (3.0) – – – 2335 (8.1) 229 (12.2) 1941 (6.1) Spine Non-hip, – – – 28,106 24,863 176,134 – – – 18,945 672 (36.1) 9313 (29.2) non-spine (33.6) (45.0) (34.6) (65.5) 13 127 Table 1  (continued) 128 NDOa Ofxb BPc FWOTd 13 CPRD SIDIAP DHR CPRD SIDIAP DHR CPRD SIDIAP DHR CPRD SIDIAP DHR Distal – – – 16,459 12,440 136,715 – – – 5022 (17.3) 527 (28.0) 7936 (24.9) forearm (19.7) (22.5) (26.8) Shoulder – – – 7521 (9.0) 4943 (8.9) 54,741(10.7) – – – 1157 (4.0) 161 (8.6) 3310 (10.4) Ankylosing 410 (0.6) 98 (0.3) 302 (0.2) 302 (0.4) 130 (0.3) 533 (0.1) 857 (0.6) 129 (0.2) 465 (0.2) 210 (0.7) 5 (0.3) 72 (0.2) spondy- litis Anxiety 7205 (10.3) 6701 (17.7) 929 (0.6) 7697 (9.2) 9405 (18.4) 1976 (0.4) 12,829 (8.6) 8589 (16.5) 997 (0.5) 3944 (13.6) 390 (20.9) 227 (0.7) disorder CCI, N (%) 0 42,875 22,894 122,357 49,455 23,386 440,183 87,330 29,428 155,729 12,412 878 (47.1) 24,566 (77.1) (61.3) (60.4) (77.3) (59.2) (45.8) (86.4) (58.8) (56.6) (76.3) (42.9) 1 14,653 8456 (22.3) 22,859 15,107 13,177 40,291 (7.9) 32,030 12,605 32,522 6977 (24.1) 462 (24.8) 4657 (14.6) (21.0) (14.5) (18.1) (25.8) (21.6) (24.2) (15.9) 2 6889 (9.9) 3871 (10.2) 8324 (5.3) 9005 (10.8) 7205 (14.1) 18,193 (3.6) 15,598 5744 (11.0) 10,279 (5.0) 4461 (15.4) 275 (14.7) 1730 (5.4) (10.5) 3+ 5482 (7.8) 2680 (7.0) 4651 (2.9) 9947 (11.9) 7276(14.2) 10,884 (2.2) 13,549 (9.1) 4260 (8.3) 5480 (2.7) 5080 (17.6) 250 (13.5) 929 (2.9) Medications Steroid/s, N 10,379 4151 (10.7) 32,247 6500 (7.8) 4429(8.7) 42,535 (8.3) 38,837 6692 (12.9) 54,036 6160 (21.3) 310(16.6) 6729 (21.1) (%)f (14.8) (20.4) (26.2) (26.5) Anti-OP 11,996 13,506 35,541 7717 (9.2) 5635 (11.03) 24,609 (4.8) 0 (0.0) 0 (0.0) 0 (0.0) 28,930 (100) 1865 (100) 31,567 (99.0) medica- (17.2) (34.8) (22.5) tion, N (%)f *In CPRD, there were 6895 (8.3%) fractures which were found to be related to historical coding and were dismissed a Newly diagnosed osteoporosis b Incident osteoporotic fracture c Incident oral bisphosphonates d Fractures while on treatment with bisphosphonates e Chronic obstructive pulmonary disease f In the previous year g CCI, Charlson Comorbidity Index Osteoporosis International (2022) 33:123–137 Osteoporosis International (2022) 33:123–137 129 Spain. The difference between the countries was observed The incidence rate for hip fracture was highest in Den- for all fracture types but most pronounced for hip fractures, mark (24.1 per 1000 persons-years) followed by the UK ranging from 4.7 to 74.2 per 1000 persons-years in Spain (18.1) and Spain (9.5); however, the opposite was found for and Denmark, respectively. clinical spine fractures which rates were highest in Spain The major osteoporotic fracture was created by adding (9.4 per 1000 persons-years) followed by Denmark (5.1) the incidence rate of hip/humerus/distal forearm and clini- and the UK (2.8). The incidence rate of major osteoporotic cal spine. The 1-year incidence rate of major osteoporotic fracture was highest for Denmark, followed by Spain and the fracture was highest for Denmark, followed by the UK and UK (42.7, 38.3 and 20.7, respectively). The 1-year incidence Spain (127.8, 33.4 and 17.3 per 1000 persons-year, respec- rate of major osteoporosis fracture per sex is depicted in tively). One-year incidence rate per sex is depicted in the the supplemental data (Online Resource 6); incidence rate supplemental data (Online Resource 6); incidence rate of (per 1000 persons-year) was greater in women in SIDIAP major osteoporosis fracture was greater in women in CPRD (41.4 vs 30.8) and DHR (44.35 vs 36.8 in women and men, (16.3 vs 9.1 per 1000 persons-year in women and men, respectively) with no significant differences found in CPRD. respectively) and in men in DHR (123.3 vs 150.1 per 1000 Age-specific 1-year incidence rates of major osteoporotic persons-year in women and men, respectively) with no sig- fracture (Fig. 3) increased similarly in Spain and Denmark nificant differences found in SIDIAP. and were both higher than the ones found for the UK, whose The age-specific 1-year incidence of major osteoporo- individuals had lower incidence rates in all age bands. tic fractures is reported in Fig. 2. While all three countries Regarding 2-year incidence rates of fracture (Table 3), showed an increase in rates with increasing age, there were overall, it was highest in Denmark, followed by the UK and nation-specific differences, with the highest rates for Den- Spain. Major osteoporotic fracture followed the same trend mark compared to the UK and Spain. These differences were with incidence rates of 41, 30 and 24.9 in Denmark, Spain more pronounced between the youngest (subjects between and the UK, respectively. Of note is the slight increase in 50 and 60 years old) and the oldest (subjects over 90 years the UK from the 1st year to the 2nd year (incidence rate of old). Young Danish subjects had a double risk of suffering 20.3 to 24.9). a fracture compared to subjects the same age in the UK or Spain, and the oldest Danish individuals had higher inci- dence rates of fracture (over 400 per 1000 persons-year) BP: new users of oral bisphosphonates compared to the UK (75) and Spain (57). The 2-year incidence rate of fracture is reported in The BP cohort was formed by a total of 148,507 individu- Table 3. Incidence rates of all fracture categories remained als from the CPRD database, 52,037 individuals from the highest for Denmark followed by the UK and Spain but SIDIAP database and 204,010 individuals from the DHR were still lower than those found for the 1st year; the great- database (Fig. 1). est reduction was found for Denmark, with approximately Danish and Spanish individuals in this cohort were 3 and a 40% decrease in the incidence rates of hip and clinical 4 years younger than the UK individuals, and Spanish indi- spine fracture. The same was observed when grouping by viduals had a lower proportion of steroid users compared major osteoporotic fracture (IR of 84.5, 28.8 and 15.4 for to the other countries (Table 1); 16.3%, 25.4% and 34.0% Denmark, the UK and Spain, respectively). patients in the UK, Spain and Denmark had an osteoporosis diagnosis. The mean time to fracture from index date was OFx: patients with a first incident fracture 181 days, 144 days and 141 days for the UK, Spain and Dan- at osteoporotic sites ish individuals, respectively. The 1-year incidence rates of individual fracture catego- The OFx cohort consisted of a total of 83,514 from the ries were highest in Denmark followed by the UK and lowest CPRD database, 51,044 from the SIDIAP database and in Spain (Table 2). When considering hip fractures, both the 509,551 individuals from the DHR database (Fig. 1). UK and Denmark had higher rates than the ones experienced UK individuals were on average 8 and 6 years older than by Spanish individuals. The incidence rate of clinical spine the Danish and Spanish individuals, and there were fewer fractures in Denmark (8.3 per 1000 persons-year) doubled men (around 1 in 5 participants) compared to Denmark or the ones found in the UK or Spain (4.3 in both countries). Spain (about 1/3 in both); 9.6%, 15.7% and 4.5% patients in Overall, the 1-year rates of major osteoporotic fractures the UK, Spain and Denmark had an osteoporosis diagnosis. were highest for Denmark, followed by the UK and Spain Compared to the newly diagnosed osteoporosis cohort, (48.6, 31.3 and 21.1, respectively). The 1-year incidence individuals with an incident osteoporosis fracture had lower rate of major osteoporosis fracture per sex is depicted in use of steroids and bisphosphonates and had an overall the supplemental data (Online Resource 6); incidence rate worse health status (Table 1). (per 1000 persons-year) was greater in women in SIDIAP 13 130 13 Table 2  One-year incident rates (IR per 1000 persons-year, 95% confidence interval (CI)) of primary outcomes CPRD (UK) SIDIAP (Spain) DHR (Denmark) N Per year at risk IR 95% CI N Per year at risk IR 95% CI N Per year at risk IR 95% CI NDOa Hip fracture 1025 67,493 15.2 14.2–16.1 178 37,503 4.7 4.0–5.4 10,436 140,560 74.2 72.8–75.7 Spine (clinical) 445 67,739 6.6 5.9–7.2 93 37,540 2.5 1.9–2.9 3776 145,210 26.0 25.2–26.8 Non-hip, non-spine 2425 66,697 36.4 34.9–37.8 721 37,204 19.3 17.9–20.7 8489 142,850 59.4 58.2–60.7 Hip/humerus/distal forearm 1797 67,074 26.8 25.5–28.0 554 37,302 14.8 13.6–16.0 14,083 138,298 101.8 100.1–103.5 OFxb Hip fracture 1281 70,874 18.1 17.1–19.1 466 48,754 9.5 8.6–10.5 11,195 463,758 24.1 23.7–24.6 Spine (clinical) 197 71,249 2.8 2.4–3.1 459 48,630 9.4 8.6–10.3 2414 471,643 5.1 4.9–5.3 Non-hip, non-spine 5433 67,691 80.3 78.1–82.2 3216 46,544 69.0 66.7–71.4 24,852 455,208 54.6 53.9–55.3 Hip/humerus/distal forearm 1271 71,003 17.9 16.9–18.9 1392 48,096 28.9 27.4–30.4 17,491 465,546 37.6 37.0–38.1 BPc Hip fracture 2305 140,595 16.4 15.7–17.1 259 51,399 5.0 4.4–5.7 3969 195,372 20.3 19.7–20.9 Spine (clinical) 614 141,357 4.3 4.0–4.7 220 51,414 4.3 3.7–4.8 1646 196,442 8.3 8.0–8.9 Non-hip, non-spine 4694 139,175 33.7 32.8–34.7 1121 50,865 22.0 20.8–23.3 7123 193,773 36.8 35.9–37.6 Hip/humerus/distal forearm 3781 139,797 27.0 26.2–27.9 857 51,019 16.8 15.7–17.9 7787 193,340 40.3 39.4–41.2 FWOTd Hip fracture 728 25,209 28.9 26.8–31.0 14 1824 7.6 4.6–12.1 992 29,481 33.6 31.5–35.7 Spine (clinical) 204 25,454 8.0 6.9–9.1 22 1816 12.0 7.0–17.2 304 29,783 10.2 9.1–11.3 Non-hip, non-spine 1370 24,776 55.3 52.4–58.2 135 1731 77.9 64.3–91.1 1980 28,847 68.6 65.6–71.6 Hip/humerus/distal forearm 1095 24,948 43. 9 41.3–46.5 52 1799 28.8 21.0–36.7 1158 29,453 39.3 37.1–41.6 a Newly diagnosed osteoporosis b Incident osteoporotic fracture c Incident oral bisphosphonates d Fractures while on treatment with bisphosphonates Osteoporosis International (2022) 33:123–137 Osteoporosis International (2022) 33:123–137 131 clinical spine, where Spain had the highest incidence rates in this population. One-year IR of major osteoporotic frac- ture was highest in the UK followed by Denmark and Spain (51.9, 49.5 and 40.8, respectively). The 1-year incidence rate of major osteoporosis fracture per sex is depicted in the supplemental data (Online Resource 6); incidence rate (per 1000 persons-year) did not show significant differences between men and women. The age-specific major osteoporotic fracture rates were very similar for all three countries up to the age of 90 years old, over which the Spanish individuals suffer a drop in their incidence rates reaching the lowest levels in this cohort prob- ably influenced by the low number of individuals included in this age band (Fig. 3). Interestingly, the risk appeared much higher than in previous cohorts among the younger groups Fig. 2  Age-specific 1-year incidence rate per 1000 persons-year of of 50 to 60 years of age: from 37 (Spain) to 44 per 1000 major osteoporotic fractures in newly diagnosed osteoporosis cohort persons-years (UK and Denmark) for major osteoporotic fracture (Fig. 3). Two-year incidence rates (Table 3) of hip fracture were (22.7 vs 14.1) and DHR (49.8 vs 43.4 in women and men, highest in Denmark followed by the UK and Spain. How- respectively) with no significant differences found in CPRD. ever, similar rates of clinical spine fracture were demon- The age-specific 1-year risk of major osteoporotic frac- strated in all three countries. ture showed increasing IR with age and higher rates for Den- mark followed by the UK and Spain (Fig. 3). Site of index fracture (OFx and FWOT) When stratifying this cohort depending on their previ- ous fracture history (Online Resource 3), those who had a OFx: incident osteoporotic fracture cohort previous fracture had high incidence rates of major osteo- porotic fracture in all countries and for subjects in the UK When analyzing data by the site of index fracture, in all and Denmark higher incidence rates of the hip compared to countries, those who had an index hip fracture (Online clinical spine fracture. Resource 4) or clinical spine fracture (Online Resource 5) Two-year IR (Table 3) of major osteoporotic fracture had higher incidence rates of major osteoporotic fracture group was highest in Denmark and the UK than in Spain than of hip or clinical spine fractures. (40.1, 26.9 and 16.6 in Denmark, the UK and Spain, Incidence rates of hip fracture were higher among indi- respectively). viduals with an index clinical spine fracture compared with those with an index hip fracture. FWOT: patients with incident fracture while on treatment with oral bisphosphonates FWOT: fracture while on treatment with bisphosphonates cohort The FWOT study cohort composed a total of 28,930 from the CPRD database, 1,865 from the SIDIAP database and As in the incident osteoporotic fracture cohort, those who 31,882 individuals from the DHR database (Fig. 1). had index hip fracture had more frequently a subsequent As with the other three cohorts, Spanish patients were major osteoporotic fracture and among these a second hip the youngest and had a lower proportion of men. Compared fracture (Online Resource 4). Those who had an index clini- to Denmark, Spain and UK individuals had poorer health cal spine fracture had more frequently a hip fracture in the status, with 17% and 13% of individuals with a Charlson UK and Denmark but had higher rates of clinical spine frac- index 3+ respectively (Table 1); 47.8%, 52.9% and 50% of ture in Spain (Online Resource 5). patients in the UK, Spain and Denmark had an osteoporosis diagnosis. The mean duration of bisphosphonate use from index date to incident fracture was 100.8, 43.7 and 61 days Discussion for the UK, Spain and Denmark, respectively. As in subjects with an incident osteoporotic fracture, In this large multi-database and multinational cohort study, the 1-year incidence rate of hip fracture was higher in Den- we report 1-year and 2-year fracture risk in four different mark and lowest in Spain; however, this was not true for the cohorts: (1) newly diagnosed osteoporosis, (2) incident 13 132 13 Table 3  Two-year incidence rate (IR per 1000 persons-year, 95% confidence interval (CI)) of primary outcomes CPRD (UK) SIDIAP (Spain) DHR (Denmark) N Per year at risk IR 95% CI N Per year at risk IR 95% CI N Per year at risk IR 95% CI NDOa Hip 1740 131,120 13.3 12.6–13.9 340 73,986 4.6 4.1–5.0 12,510 268,633 46.6 45.7–47.4 Clinical spine 681 131,963 5.2 4.8–5.5 161 74,126 2.1 1.8–2.5 4328 278,622 15.5 15.0–16.0 Non-hip, non-spine 4049 128,463 31.5 30.5–32.5 1295 72,939 17.8 16.8–18.7 12,607 270,481 46.6 45.8–47.4 Hip/humerus/distal forearm 3057 129,717 23.6 22.7–24.4 977 73,291 13.3 12.4–14.2 18,095 262,188 69.0 68.0–70.0 OFxb Hip 2499 131,868 18.9 18.2–19.7 886 94,423 9.3 8.7–10.1 18,919 887,510 21.3 21.0–21.6 Clinical spine 384 133,523 2.9 2.6–3.2 571 94,413 6.0 5.6–6.5 3757 908,720 4.1 4.0–4.3 Non-hip, non-spine 7269 124,946 58.2 56.8–59.5 4158 89,244 46.5 45.1–48.0 39,973 861,638 46.4 45.9–46.8 Hip/humerus/distal forearm 2899 131,720 22.0 21.2–22.8 2220 92,649 24.0 23.0–25.0 32,264 875,089 36.9 36.5–37.3 BPc Hip 3776 269,447 14.0 13.6–14.5 443 101,391 4.3 3.9–4.7 6251 378,368 16.5 16.1–16.9 Clinical spine 997 271,788 3.7 3.4–3.9 295 101,469 2.9 2.6–3.2 2205 381,941 5.7 5.5–6.0 Non-hip, non-spine 7650 264,866 28.9 28.2–29.5 1843 99,718 18.5 17.7–19.3 12,229 372,027 32.9 32.3–33.4 Hip/humerus/distal forearm 6198 266,862 23.2 22.6 -23.8 1380 100,261 13.7 13.0–14.5 12,777 371,440 34.4 33.8–35.0 FWOTd Hip 1107 46,846 23.6 22.2–25.0 34 3558 9.5 6.3–12.7 1602 54,383 29.4 28.0–30.9 Clinical spine 346 47,551 7.3 6.5–8.0 26 3551 7.3 4.5–10.1 490 56,708 8.6 7.9–9.4 Non-hip, non-spine 2169 45,603 47.6 45.6–49.6 185 3331 55.5 47.5–63.5 3142 52,562 59.8 57.7–61.9 Hip/humerus/distal forearm 1675 46,142 36.3 34.6–38.0 90 3485 25.8 20.4–31.1 2109 53,961 39.1 37.4–40.7 a Newly diagnosed osteoporosis b Incident osteoporotic fracture c Incident oral bisphosphonates d Fractures while on treatment with bisphosphonates Osteoporosis International (2022) 33:123–137 Osteoporosis International (2022) 33:123–137 133 Fig. 3  Age-specific one-year incidence rate per 1000 persons- year of major osteoporotic fractures in the osteoporotic fracture, bisphosphonate treat- ment and fracture while on treatment cohorts osteoporotic fracture, (3) incident use of oral bisphospho- (13.4% and 13.3% in the fracture while on treatment and nates and (4) fractures while on treatment with oral bispho- incident osteoporotic fracture cohort, respectively). The sphonates. The study highlights that at the end of the 1st and majority of the index fractures in these cohorts were non- 2nd years, patients who fracture while on treatment are par- hip, non-spine. ticularly prone to further major osteoporotic fractures; this In Spain, individuals who experienced the highest 1-year is true for two of the three databases (SIDIAP and CPRD). incidence rate of fracture and therefore at greater risk of For DHR, individuals at greater risk of major osteoporotic major osteoporotic fractures were those who fractured while fractures are those with newly diagnosed osteoporosis. In on treatment (40.8 per 1000 persons-year) followed by those accordance with the global epidemiology of osteoporosis, who had had an incident osteoporotic fracture (38.3 per 1000 we consistently found an increased risk of fracture with persons-year). The same is seen at the end of the 2nd year. increasing age, with relatively little difference in the age- risk gradient between the different cohorts. Of particular Fractures in the UK note is the highly elevated risk of fracture found especially among the younger individuals (50–60 years old) included in The CPRD database is a primary health care population the fracture while on treatment cohort, which was twice that database. Individuals with an incident fracture and those found in the same age category in the incident osteoporotic who fractured while on treatment were older than the rest fracture cohort. of the cohorts (mean age of 78–79 years old) and had poorer health status, with a higher proportion of individuals with Fractures in Spain > 3 points in the Charlson Comorbidity Index (17.6% and 11.9% for those who fractured while on treatment and the The SIDIAP database is a primary health care population- incident osteoporotic fracture cohort, respectively). based database. Overall, Spanish individuals were diagnosed Regarding the previous fractures, 15.2% of the individu- and treated for osteoporosis at younger ages compared to the als with an incident diagnose of osteoporosis already had a other countries. fracture, and 8.3% of the individuals in the incident osteo- Individuals who fractured while on treatment were porotic fracture cohort had already a previous fracture which older than the rest of cohorts in this country (mean age of is due to difficulties in accurately dating the history of previ- 73 years) and had, as for the incident osteoporotic fracture ous fractures with the read codes used in CPRD. cohort, poorer health status with a higher proportion of indi- When analysing the distribution of the index fractures viduals with > 3 points in the Charlson Comorbidity Index in the incident osteoporotic fracture and fracture while on 13 134 Osteoporosis International (2022) 33:123–137 treatment cohort, 67.6% were hip in the incident osteoporo- comorbidity which could have contributed to the increased tic fracture, while the majority of the index fractures in the risk of fracture. fracture while on treatment cohort were non-hip, non-spine The difference in the rates of major osteoporotic frac- (65.5%). tures—highest in Denmark and lowest in Spain—likely In the UK, the highest 1 year incidence rates of major reflect the already known geographical worldwide variation osteoporotic fractures are found among those who fractured of fracture incidence with greater incidence rates of fracture while on treatment followed by those who had a new diag- in Northern Europe [23]. The reasons for this remain elu- nose of osteoporosis (51.9 and 33.4 per 1000 py, respec- sive and are beyond the scope of this paper but may include tively). The same is seen at the end of the 2nd year. genetic differences including differences in body height and obesity, socio-economic differences, lifestyle habits, nutri- tion and even intrinsic differences in bone quality and bone Fractures in Denmark geometry [24–26]. Age is an already known risk factor for fracture, and it As in Spain and the UK, individuals who fractured while on is unsurprising that this also affects the immediate risk of treatment were older than the rest of the cohorts (mean age major osteoporotic fractures [3, 4, 19, 27]. We found age- of 75 years old) and had poorer health status with a higher specific major osteoporotic fractures rates very similar proportion of individuals with > 3 points in the Charlson between all the countries; however, we found that our young- Comorbidity Index, which was true for those with incident est population, those aged 50–60 years old, started with an osteoporotic fracture and for those who fractured while on overall increased fracture rate (5 and 10 per 1000 per year). treatment but was especially relevant in this last group (2.9% Nevertheless, these incident rates are still lower than what and 2.2% for the fractures while on treatment and incident was found in a study carried out in Norway on this same age osteoporotic fracture cohort, respectively). The majority of band (35 to 50 per 1000 persons-year) [28]. the index fractures in the incident osteoporotic fracture and In the last years, there is a growing awareness of the in the fracture while on treatment cohort were non-hip, non- importance of fracture prevention in the imminent fracture spine. Incidence rates of fracture in Denmark were higher risk period. Our results show that the imminent fracture risk than in the two other countries, and as will be discussed increased with age in all patients included (with or without in the succeeding text, the highest incidence rate of major a fracture) but was especially pronounced in those who had osteoporotic fracture was found for those individuals with already fractured while on treatment with bisphosphonates. a newly diagnosed osteoporosis (127.8 per 1000 persons- These subjects even those at younger ages were twice as year); this was true for the 1st- and 2nd-year IR. likely to fracture compared to the rest of the cohorts; there- fore, focusing the preventive strategies on these patients Interpretation and comparison with other studies could help optimize the health care resources in each coun- try. Future studies are needed to stratify the results per sex, The increased rate of fracture found, in all the countries, or the Charlson Comorbidity Index and multivariate models’ among individuals with a previous fracture (incident osteo- analysis to assess the association between the imminent frac- porotic fracture and fracture while on treatment cohort) has ture risk and baseline characteristics in each of the cohorts been already described, regardless of the bisphosphonate would contribute to more accurately define the population adherence [18], and even among highly adherent individu- at risk. als included in some randomized controlled trials of bis- phosphonates [19, 20]. Fewer are the studies using real- Strengths and limitations world data; in 2014, a study based on the SIDIAP database reported an increased risk of fracture among bisphosphonate To our knowledge, this is the first multinational and multi- users (3.4% of these individuals fractured every year) with database study that analyses the imminent risk of fracture older age and previous fractures identified as risk factors in four different groups of individuals at risk using real- [21], and another carried out in Denmark in 2016 found a world data. The large number of individuals included and higher incidence of subtrochanteric and femoral shaft frac- given the data gathered (routinely collected in primary tures among highly adherent alendronate users after 13 years care with hospital linkage) enables us to easily extrapolate of treatment; however, further analysis justified this increase our results. However, this study must also be interpreted by the worse profile of risk factors among the older pop- in light of some limitations. First, there are differences in ulation rather than by a detrimental effect of alendronate the number of individuals provided by each database for [22]. As in these previous studies, in our study, individuals each cohort; in the incident osteoporotic fracture cohort, who fractured while on treatment in all the countries were the DHR database contributed with 509,551 individuals, older than the rest with poorer health status and higher which is over sixfold higher compared to CPRD (83,514) 13 Osteoporosis International (2022) 33:123–137 135 and over ninefold higher compared to SIDIAP (55,304). Sec- across all three databases. At last, the stratified results could ond, the different nature of the databases (Online resource not be driven by the strata definition itself but underlying 6), with CPRD and SIDIAP from primary care and DHR differences; hence, multivariate models are required to truly from hospital bases, renders comparison between them dif- identify the patients with high risk. ficult. The DHR is a hospital-based database coupled with a nationwide prescription register that includes primary prac- tice, compared to the primary care database of CPRD and SIDIAP. In some respects, it affects cohort definitions only Conclusions as regards the newly diagnosed osteoporosis cohort where a small number of patients will be given an osteoporosis code Individuals who already have sustained an initial fracture in hospitals compared that with primary practice, selecting and especially those who suffer a fracture while on treatment for a smaller and more severely affected population. This can with bisphosphonates experience higher re-fracture inci- also be reflected in the proportion of patients with a higher dence rates than those with a newly diagnosed osteoporosis Charlson Comorbidity Index (> 3) which is similar in the with or without treatment. Spanish individuals demonstrated newly diagnosed osteoporosis, in the fracture while on treat- an overall lower risk of major osteoporotic fractures except ment with bisphosphonates and in the incident osteoporosis for those individuals who re-fractured, where the incident fracture cohort in Denmark suggesting a possible selection rate was found even higher than in the UK. Similar age- bias of a more severely affected population at baseline com- specific major osteoporotic fractures were seen for all three pared to what is found in the other countries, which would countries; however, the risk was much higher among the attenuate or confound the differences between the cohorts younger patients (50–60 years old). found with regard to fracture risk. As a minimum, the data- base differences will also affect the threshold for contacts Supplementary Information The online version contains supplemen- tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 00198-0​ 21-0​ 6077-0. with the health service, given that patients, in general, will have more contacts with primary practice than with hospi- Acknowledgements The authors would like to acknowledge Paloma tal clinics, so the same patient is likely to achieve a higher Doherty for project management; CPRD, SIDIAP and DHR for their Charlson score in a general practice database than in a hos- support; and UCB Pharma for funding this study. pital system database since some conditions do not require referral to specialists. By contrast and critical to the epide- Author contributions Study design: DPA, BA and ET. Data analysis: SK, DPA and ME. Data interpretation: SK, DPA, BA, ET, CL, CC and miological value of the study, all three databases capture CR. Drafting manuscript: SK, DPA and CR. Revising manuscript con- hospital-treated fractures, and all three databases are able tent: all co-authors. Approving final version of manuscript: all authors. to accurately capture initiation of osteoporosis treatment. DPA takes responsibility for the integrity of the data analysis. Third, among the incident fracture cohort, incident frac- tures were, as defined, the first fracture sustained in all Funding This study received funding from UCB Pharma. Danish patients and in almost all patients in Spain. How- Data availability The lead author (SK) affirms that the manuscript is an ever, approximately 8% of the UK patients experienced a honest, accurate and transparent account of the study being reported. fracture previous to the inclusion in the cohort (Table 1); No important aspects of the study have been omitted, and any discrep- these patients’ fractures were found to be related to histori- ancies from the study as planned (and, if relevant, registered) have been cal coding and were dismissed from the analysis. Fourth, explained. Regarding data sharing, there is no additional data available. the duration of the treatment of bisphosphonates in the inci- Code availability No additional code is available. dent bisphosphonate treatment cohort was not assessed, and therefore, we cannot exclude that the fracture rates reported could be partly due to a previous baseline risk of the patient Declarations (before the maximum effect of the bisphosphonate treat- Conflict of interest BA reports grants from UCB, during the conduct ment). However, evidence suggests that the anti-fracture of the study, personal fees from UCB, grants from Novartis, personal efficacy starts at the beginning of the treatment and reaches fees from Amgen, personal fees from Eli-Lilly and personal fees from its maximum of 3 months after the initiation of the bisphos- Pharmacosmos, outside the submitted work. ME reports institutional grants from UCB, during the conduct of the study. CL reports being phonates [29]. In the worst-case, fracture rates in this cohort an employee and stockholder of UCB Pharma. DPA reports grants and would be slightly lower, unlikely to affect our conclusions. others from AMGEN; grants, non-financial support and others from Moreover, potential unknown discrepancies in coding could UCB Biopharma; and grants from Les Laboratoires Servier, outside have taken place, leading to an under or overestimation of the submitted work; Janssen, on behalf of IMI-funded EHDEN and EMIF consortiums, and Synapse Management Partners have support- the fracture rates. There are also potential analytical dis- ed training programmes organized by DPA’s department and open for crepancies, which were minimized by the use of a single external participants. ET is an employee of UCB Pharma, who spon- standardized and double-checked (by two analysts) script sored the study. AD, SK and CR have no conflicts to declare. 13 136 Osteoporosis International (2022) 33:123–137 Consent to participate Consent to participate was not required as only 6. Kanis JA, Johansson H, Odén A, Harvey NC, Gudnason V, Sand- anonymised retrospective data was used for this study, and no patient ers KM, Sigurdsson G, Siggeirsdottir K, Fitzpatrick LA, Borg- or GP contact was required. ström F, McCloskey EV (2018) Characteristics of recurrent frac- tures. Osteoporos Int 29(8):1747–1757. https://​doi.​org/​10.​1007/​ Consent for publication All authors have given their consent for the s00198-​018-​4502-0 publication. 7. Johansson H, Siggeirsdóttir K, Harvey NC, Odén A, Gudnason V, McCloskey E, Sigurdsson G, Kanis JA (2017) Imminent risk Ethical approval All procedures performed in this study were in of fracture after fracture. Osteoporos Int 28(3):775–780. https://​ accordance with the ethical standards of the institutional and/or doi.​org/​10.​1007/​s00198-​016-​3868-0 national research committees of each country (Spain, UK and Den- 8. Hannan MT, Weycker D, McLean RR, Sahni S, Bornheimer mark) and with the 1964 Helsinki Declaration and its later amendments R, Barron R, Travison TG, Kiel DP (2019) Predictors of immi- or comparable ethical standards. More specifically, for the UK, the nent risk of nonvertebral fracture in older, high-risk women: the study was approved by the Independent Scientific Advisory Committee Framingham Osteoporosis Study. JBMR Plus 3(6):e10129. https://​ (ISAC 17_128R); for Spain, the study was approved by the IDIAP J doi.​org/​10.​1002/​jbm4.​10129 Gol Clinical Research Ethics Committee (P17_152); and for Denmark, 9. Bliuc D, Alarkawi D, Nguyen TV, Eisman JA, Center JR (2015) the study was approved by Statistics Denmark (Reference 706638). Risk of subsequent fractures and mortality in elderly women and men with fragility fractures with and without osteoporotic Informed consent Informed consent was not required as only bone density: the Dubbo Osteoporosis Epidemiology Study. J anonymised retrospective data was used for this study, and no patient Bone Miner Res 30(4):637–646. https://​doi.​org/​10.​1002/​jbmr.​ or GP contact was required. 2393 10. Kanis JA, Johnell O, De Laet C, Johansson H, Oden A, Delmas P, Eisman J, Fujiwara S, Garnero P, Kroger H, McCloskey EV, Open Access This article is licensed under a Creative Commons Attri- Mellstrom D, Melton LJ, Pols H, Reeve J, Silman A, Tenenhouse bution-NonCommercial 4.0 International License, which permits any A (2004) A meta-analysis of previous fracture and subsequent non-commercial use, sharing, adaptation, distribution and reproduction fracture risk. Bone 35(2):375–382. https://d​ oi.o​ rg/1​ 0.1​ 016/j.b​ one.​ in any medium or format, as long as you give appropriate credit to the 2004.​03.​024 original author(s) and the source, provide a link to the Creative Com- 11. Kanis JA, Johnell O, Oden A, Johansson H, McCloskey E (2008) mons licence, and indicate if changes were made. The images or other FRAX and the assessment of fracture probability in men and third party material in this article are included in the article's Creative women from the UK. Osteoporos Int 19(4):385–397. https://​doi.​ Commons licence, unless indicated otherwise in a credit line to the org/​10.​1007/​s00198-​007-​0543-5 material. If material is not included in the article's Creative Commons 12. Hippisley-Cox J, Coupland C (2012) Derivation and validation licence and your intended use is not permitted by statutory regula- of updated QFracture algorithm to predict risk of osteoporotic tion or exceeds the permitted use, you will need to obtain permission fracture in primary care in the United Kingdom: prospective open directly from the copyright holder. To view a copy of this licence, visit cohort study. BMJ 344:e3427. https://​doi.​org/​10.​1136/​bmj.​e3427 http://​creat​iveco​mmons.​org/​licen​ses/​by-​nc/4.​0/. 13. van Geel TA, van Helden S, Geusens PP, Winkens B, Dinant GJ (2009) Clinical subsequent fractures cluster in time after first frac- tures. Ann Rheum Dis 68(1):99–102. https://​doi.​org/​10.​1136/​ard.​ 2008.​092775 14. Hippisley-Cox J, Coupland C, Brindle P (2014) The perfor- References mance of seven QPrediction risk scores in an independent external sample of patients from general practice: a validation 1. Dyer SM, Crotty M, Fairhall N, Magaziner JS, Beaupre LA, Cam- study. BMJ Open 4(8):e005809. https://​doi.​org/​10.​1136/​bmjop​ eron I, Sherrington C (2017) A critical review of the long-term en-​2014-​005809 disability outcomes following hip fracture. Innov Aging 1(1):736. 15. García-Gil Mdel M, Hermosilla E, Prieto-Alhambra D, Fina F, https://​doi.​org/​10.​1093/​geroni/​igx004.​2656 Rosell M, Ramos R, Rodriguez J, Williams T, Van Staa T, Bolíbar 2. Hernlund E, Svedbom A, Ivergård M, Compston J, Cooper C, B (2011) Construction and validation of a scoring system for the Stenmark J, McCloskey EV, Jönsson B, Kanis JA (2013) Osteo- selection of high-quality data in a Spanish population primary porosis in the European Union: medical management, epidemi- care database (SIDIAP). Inform Prim Care 19(3):135–145. https://​ ology and economic burden. A report prepared in collaboration doi.​org/​10.​14236/​jhi.​v19i3.​806 with the International Osteoporosis Foundation (IOF) and the 16. Pottegård A, Schmidt SAJ, Wallach-Kildemoes H, Sørensen HT, European Federation of Pharmaceutical Industry Associations Hallas J, Schmidt M (2017) Data resource profile: the Danish (EFPIA). Arch Osteoporos 8(1–2):136. https://​doi.​org/​10.​1007/​ National Prescription Registry. Int J Epidemiol 46(3):798–798f. s11657-​013-​0136-1 https://​doi.​org/​10.​1093/​ije/​dyw213 3. Banefelt J, Åkesson KE, Spångéus A, Ljunggren O, Karlsson L, 17. Schmidt M, Schmidt SA, Sandegaard JL, Ehrenstein V, Pedersen Ström O, Ortsäter G, Libanati C, Toth E (2019) Risk of immi- L, Sørensen HT (2015) The Danish National Patient Registry: a nent fracture following a previous fracture in a Swedish database review of content, data quality, and research potential. Clin Epi- study. Osteoporos Int 30(3):601–609. https://​doi.​org/​10.​1007/​ demiol 7:449–490. https://​doi.​org/​10.​2147/​CLEP.​S91125 s00198-​019-​04852-8 18. Imel EA, Eckert G, Modi A, Li Z, Martin J, de Papp A, Allen K, 4. Bonafede M, Shi N, Barron R, Li X, Crittenden DB, Chandler D Johnston CC, Hui SL, Liu Z (2016) Proportion of osteoporotic (2016) Predicting imminent risk for fracture in patients aged 50 women remaining at risk for fracture despite adherence to oral or older with osteoporosis using US claims data. Arch Osteoporos bisphosphonates. Bone 83:267–275. https://​doi.​org/​10.​1016/j.​ 11(1):26. https://​doi.​org/​10.​1007/​s11657-​016-​0280-5 bone.​2015.​11.​021 5. Balasubramanian A, Zhang J, Chen L, Wenkert D, Daigle SG, 19. Black DM, Thompson DE, Bauer DC, Ensrud K, Musliner T, Grauer A, Curtis JR (2019) Risk of subsequent fracture after prior Hochberg MC, Nevitt MC, Suryawanshi S, Cummings SR, Frac- fracture among older women. Osteoporos Int 30(1):79–92. https://​ ture Intervention Trial (2000) Fracture risk reduction with alen- doi.​org/​10.​1007/​s00198-​018-​4732-1 dronate in women with osteoporosis: the Fracture Intervention 13 Osteoporosis International (2022) 33:123–137 137 Trial: FIT Research Group. J Clin Endocrinol Metab 85(11):4118– among Norwegian women explained by impaired bone material 4124. https://​doi.​org/​10.​1210/​jcem.​85.​11.​6953 properties? J Bone Miner Res 30(10):1784–1789. https://​doi.​org/​ 20. Harris ST, Watts NB, Genant HK, McKeever CD, Hangartner T, 10.​1002/​jbmr.​2537 Keller M, Chesnut CH 3rd, Brown J, Eriksen EF, Hoseyni MS, 26. Fajar JK, Taufan T, Syarif M, Azharuddin A (2018) Hip geometry Axelrod DW, Miller PD (1999) Effects of risedronate treatment and femoral neck fractures: a meta-analysis. J Orthop Translat on vertebral and nonvertebral fractures in women with postmen- 13:1–6. https://​doi.​org/​10.​1016/j.​jot.​2017.​12.​002 opausal osteoporosis: a randomized controlled trial: Vertebral 27. Brozek W, Reichardt B, Zwerina J, Dimai HP, Klaushofer K, Efficacy With Risedronate Therapy (VERT) Study Group. JAMA Zwettler E (2016) Antiresorptive therapy and risk of mortality 282(14):1344–1352. https://​doi.​org/​10.​1001/​jama.​282.​14.​1344 and refracture in osteoporosis-related hip fracture: a nationwide 21. Prieto-Alhambra D, Pagès-Castellà A, Wallace G, Javaid MK, study. Osteoporos Int 27(1):387–396. https://​doi.​org/​10.​1007/​ Judge A, Nogués X, Arden NK, Cooper C, Diez-Perez A (2014) s00198-​015-​3415-4 Predictors of fracture while on treatment with oral bisphos- 28. Ahmed LA, Center JR, Bjørnerem A, Bluic D, Joakimsen RM, phonates: a population-based cohort study. J Bone Miner Res Jørgensen L, Meyer HE, Nguyen ND, Nguyen TV, Omsland TK, 29(1):268–274. https://​doi.​org/​10.​1002/​jbmr.​2011 Størmer J, Tell GS, van Geel TA, Eisman JA, Emaus N (2013) 22. Abrahamsen B, Eiken P, Prieto-Alhambra D, Eastell R (2016) Progressively increasing fracture risk with advancing age after Risk of hip, subtrochanteric, and femoral shaft fractures among initial incident fragility fracture: the Tromsø study. J Bone Miner mid and long term users of alendronate: nationwide cohort and Res 28(10):2214–2221. https://​doi.​org/​10.​1002/​jbmr.​1952 nested case-control study. BMJ 353:i3365. https://​doi.​org/​10.​ 29. Drake MT, Clarke BL, Khosla S (2008) Bisphosphonates: mech- 1136/​bmj.​i3365 anism of action and role in clinical practice. Mayo Clin Proc 23. Dhanwal DK, Dennison EM, Harvey NC, Cooper C (2011) Epide- 83(9):1032–1045. https://​doi.​org/​10.​4065/​83.9.​1032 miology of hip fracture: worldwide geographic variation. Indian J Orthop 45(1):15–22. https://​doi.​org/​10.​4103/​0019-​5413.​73656 Publisher's Note Springer Nature remains neutral with regard to 24. Curtis EM, Moon RJ, Harvey NC, Cooper C (2017) Reprint of jurisdictional claims in published maps and institutional affiliations. the impact of fragility fracture and approaches to osteoporosis risk assessment worldwide. Int J Orthop Trauma Nurs 26:7–17. https://​ doi.​org/​10.​1016/j.​ijotn.​2017.​04.​004 25. Duarte SD, Vilaplana L, Güerri R, Nogués X, Wang-Fagerland M, Diez-Perez A, Eriksen FE (2015) Are the high hip fracture rates 13