- Email: [email protected]
Risk of major osteoporotic fracture after first, second and third fracture in Swedish women aged 50 years and older
Journal Pre-proof Risk of major osteoporotic fracture after first, second and third fracture in Swedish women aged 50 years and older
Emma Söreskog, Oskar Ström, Anna Spångéus, Kristina E. Åkesson, Fredrik Borgström, Jonas Banefelt, Emese Toth, Cesar Libanati, Mata Charokopou PII:
S8756-3282(20)30066-1
DOI:
https://doi.org/10.1016/j.bone.2020.115286
Reference:
BON 115286
To appear in:
Bone
Received date:
15 October 2019
Revised date:
16 January 2020
Accepted date:
14 February 2020
Please cite this article as: E. Söreskog, O. Ström, A. Spångéus, et al., Risk of major osteoporotic fracture after first, second and third fracture in Swedish women aged 50 years and older, Bone(2020), https://doi.org/10.1016/j.bone.2020.115286
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier.
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
Risk of Major Osteoporotic Fracture After First, Second and Third Fracture in Swedish Women Aged 50 Years and Older Emma Söreskog,a Oskar Ström,b Anna Spångéus,c Kristina E Åkesson,d Fredrik Borgström,e Jonas Banefelt,f Emese Toth,g Cesar Libanati,h Mata Charokopoui a
Quantify Research, Hantverkargatan 8, SE-112 21 Stockholm, Sweden,
[email protected]; bQuantify Research, Hantverkargatan 8, SE-
ro of
112 21 Stockholm, Sweden and Karolinska Institutet, Medical Management Centre, SE-171 77, Stockholm, Sweden, [email protected]; cLinköping University, Sandbäcksgatan 7, SE-581 83, Linköping, Sweden,
-p
[email protected]; dLund University, Skåne University Hospital, Inga Marie
re
Nilssons gata 22, SE-205 02 Malmö, Sweden, [email protected]; eQuantify
lP
Research, Hantverkargatan 8, SE-112 21 Stockholm, Sweden and Karolinska Institutet, Medical Management Centre, SE-171 77, Stockholm, Sweden,
na
[email protected]; fQuantify Research, Hantverkargatan 8, SE-112 21 Stockholm, Sweden, [email protected]; gUCB Pharma,
ur
Allée de la Recherche, 60 1070 Brussels, Belgium, [email protected]; hUCB
i
Jo
Pharma, Allée de la Recherche, 60 1070 Brussels, Belgium, [email protected]; UCB Pharma, Allée de la Recherche, 60 1070 Brussels, Belgium,
[email protected]
Correspondence to: Emma Söreskog; [email protected] Short title: Fracture after Fracture Funding: UCB Pharma and Amgen Inc. Abbreviations: Anatomical Therapeutic Chemical Classification System (ATC); bone mineral density (BMD); confidence interval (CI); hazard ratio (HR); International 1
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
Classification of Diseases, 10th Revision (ICD-10); major osteoporotic fracture
Jo
ur
na
lP
re
-p
ro of
(MOF); standard deviation (SD)
2
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
Abstract
Background Osteoporosis affects approximately one in five European women and leads to fragility fractures, which result in poor health, social and economic consequences. Fragility fractures are a strong risk factor for subsequent major osteoporotic fracture (MOF), with risk of MOF being elevated in the 1–2 years following an earlier fracture, a concept described as “imminent risk”. This study examines risk of subsequent MOF in patients with one, two or three prior fractures by age and type of fracture.
ro of
Methods
In this retrospective, observational cohort study, Swedish women aged ≥50 years with ≥1 any clinical fragility fracture between July 1, 2006–December 31, 2012 were
-p
identified from Sweden’s National Patient Register. Each patient was age- and sexmatched to three controls without history of fracture. Group 1 women included those
re
with one fragility fracture during the study period; Group 2 included those with two fragility fractures; and Group 3 included those with three fragility fractures. “Index
lP
fracture” was defined as the first fracture during the study period for Group 1; the second for Group 2; and the third for Group 3. Patients in each cohort and matched
na
controls were followed for up to 60 months or until subsequent MOF (hip, vertebra,
Results
ur
forearm, humerus), death or end of data availability.
Jo
231,769 women with at least one fracture were included in the study and therefore constituted Group 1; of these, 39,524 constituted Group 2 and of those, 7,656 constituted Group 3. At five years, cumulative incidence of subsequent MOF was higher in patients with a history of fracture as compared to controls (Group 1: 20.7% vs 12.3%; Group 2: 32.0% vs 15.3%). Three-year cumulative incidence for Group 3 was 12.1% (vs 10.7% for controls). After adjusting for baseline covariates, risk of subsequent MOF was highest within 0–24 months following an index fracture, then decreased but remained elevated as compared to controls. Having two prior fractures, vertebral fractures and younger age at time of index fracture were associated with greater relative risk.
3
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
Conclusions Women with a history of osteoporotic fracture are at increased risk of subsequent fracture, which is highest during the first 24 months following a fracture. Younger women and those with vertebral fractures are at greatest relative risk, suggesting that treatment should target these patients and be timely enough to impact the period of imminent risk.
Jo
ur
na
lP
re
-p
ro of
Key words: fracture incidence; fragility fracture; imminent risk; osteoporosis
4
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
1. INTRODUCTION Osteoporosis, as defined by densitometric measurement, is a major cause of fragility fractures, and affects approximately 21% of women aged 50–84 in the EU and 15.4% of women aged 50 and older in the US [1, 2]. Fragility fractures not only lead to poor health and social outcomes such as pain and disability [3-5], but are also associated with a significant economic burden. For example, in the EU, fragility fractures accounted for €26 billion and 26,300 life-years lost in 2010 alone [1]. Fragility fractures occur as a consequence of insufficient bone strength due to
ro of
deficits in bone mass and microstructure [6]. Risk factors for fracture include increased age, low bone mineral density (BMD), female sex and glucocorticoid use, among others [7-10]. Additionally, a history of fragility fracture has been identified as a strong driver of subsequent fracture risk [7, 9-14]. The risk of any subsequent
-p
fracture was reported to be twice that of a first fracture using pooled data from a large meta-analysis of studies published between 1966 and 1999 [15], and was even
re
higher (approximately four-fold) for subsequent vertebral fractures in women with a
lP
prior vertebral fracture. Similar findings were observed by others; Ahmed et al. (2013) found that an initial fracture was associated with 1.3–2.0 times the risk (for women and men, respectively) as compared to patients without an earlier fracture,
na
with that risk concentrated within the first five years post-fracture [14]. Additionally, van Geel et al. (2009) estimated that risk of subsequent fracture within 15 years of a
ur
first fracture is approximately double that of patients without a first fracture [16].
Jo
It is now clear that the increase in subsequent fracture risk following an initial fracture is not linear over time. While van Geel et al. demonstrated a doubling of risk resulting from a prior fracture, the same study found a five-fold risk of subsequent major osteoporotic fracture (MOF) in the year following the first fracture [16]. Other studies have reported that up to one-fourth of patients will sustain a second fracture during the 1–2 years following an earlier fracture, depending on the type of first fracture [17, 18]. This pattern of elevated subsequent fracture risk in the period immediately following a fragility fracture has been described as “imminent fracture risk” [12, 19-22]. What is less clear is how the risk of a subsequent fracture changes following a second or third fracture, and how a patient’s age and prior fracture type influence the level of imminent risk.
5
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
This study examines the risk of subsequent MOF, defined as those at the hip, vertebra, forearm or humerus, in patients with a history of one, two or three prior fractures of any type, compared with age- and sex-matched controls without history of fracture. 2. METHODS 2.1. Patients and study design In this retrospective, observational cohort study, Swedish females aged ≥50 years with at least one fragility fracture occurring during the study period of July 1, 2006–
ro of
December 31, 2012 were identified from the Swedish National Patient Register (Figure 1). Women with clinical fragility fractures at the hip, vertebra, shoulder, upper arm, pelvis, femur, lower leg, wrist, hand, rib and/or sternum were included.
-p
Women identified to have a fracture during the period of January 1, 2001–June 30, 2006 were excluded. Women identified to have Paget’s disease or any malignancy
re
(other than basal cell carcinoma) at any time during the study were excluded from
lP
the analysis.
Three separate analysis cohorts were identified. Group 1 included all women with at least one fragility fracture during the study period; Group 2 was constituted of the
na
women in Group 1 who sustained a second fragility fracture during the study period; and Group 3 was constituted of the women in Group 2 who sustained a third fragility
ur
fracture during the study period. “Index fracture” was defined as the fragility fracture
Jo
that classified a patient in a given Group. Therefore, the index fracture for Group 1 was their first fragility fracture to occur during the study period. For Group 2, the index fracture was their second to occur during the study period, and for Group 3, the index fracture was their third fracture during the study period. The date of the index fracture was defined as the index date. The baseline period during which patient characteristics data were retrieved was defined as the 12-month period before the index fracture date. For each group, each woman was age- and sex-matched with replacement to three controls without a history of fracture on, or before (dating back to January 1, 2001), the index fracture date. Controls were randomly selected from a population of 260,401 women ≥50 years old included in the Swedish Total Population Register and were assigned the index date of their matched counterpart. Patients in each cohort and their matched 6
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
controls were followed for up to 60 months or until subsequent MOF, death or end of data availability. Index fracture was defined as any fragility fracture because all types are associated with increased fracture risk [15]. For the outcome fracture, we selected MOFs as they represent the group of fractures with the greatest burden. 2.2. Data sources The National Patient Register was used to identify patients for inclusion; extracted data included sex, patient identification numbers, diagnosis codes, dates of diagnosis, duration of hospitalization and specialist visits. Diagnosis codes for each index and
ro of
subsequent fracture type are listed in Supplementary 1. Patient dispensed drug data were obtained from Sweden’s Prescribed Drug Register; extracted data included patient identification numbers, Anatomical Therapeutic Chemical Classification System (ATC) codes, prescription dates, dispensing dates, defined daily doses per
-p
prescribed package and number of pills/injections, etc. Sweden’s Cause of Death Register was used to establish patients’ dates of death using the national person
re
identification numbers based on birthdate. Clinical and demographic characteristics
lP
were collected at the date of each index fracture or in the 12-month period prior to the index fracture date. Matched data were retrieved for controls from the Swedish
na
Total Population Register.
2.3. Identifying fractures
ur
To avoid counting the same index fracture more than once, fractures at the same
Jo
site in the body and identified during an inpatient hospitalization were counted only if they occurred at least six months after the index fracture. Fractures at the same site of an earlier fracture and identified during an outpatient visit were counted only if they occurred at least 12 months after the index fracture. Hip fractures required an inpatient hospital admission. 2.4. Data analysis MOFs included hip, vertebral, forearm and humerus fractures. Incidence of subsequent MOF was calculated for Groups 1, 2 and 3 and compared with the respective control group. The relationship between having an index fracture and risk of subsequent fracture was assessed via the Royston-Parmar parametric spline model with MOF as a single failure event [23]. Data were right-censored at time of death or end of follow-up. Hazard ratios were assessed in separate time windows: 0– 7
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
6 months, 7–12 months, 13–18 months, 19–24 months, 25–36 months, 37–48 months and 49–60 months after index fracture. The time windows were chosen after iteratively testing the proportional hazard assumption by different windows. These analyses were performed on each group overall and by age (50–59, 60–69, 70–79 and ≥80 years). The risk of subsequent MOF was adjusted for several baseline covariates known to be associated with increased risk of fragility fracture: attained age; use of osteoporosis medication in the past 12 months; use of drugs known to increase the risk of falls in the past 12 months; glucocorticoid use in the past 12 months (defined
ro of
according to the FRAX® algorithm as past or present exposure of ≥5 mg daily dosage of prednisolone or equivalent for ≥3 months [24]); number of different medications prescribed in the past 12 months; secondary osteoporosis; rheumatoid arthritis;
-p
Charlson comorbidity index measured two years prior to index date [25], and prepackaged drug dispensing (defined as use of the ApoDos system within the past 12
re
months), which was used as a proxy for being “dependent”.
lP
3. RESULTS
3.1. Baseline characteristics
na
A total of 231,769 women with at least one index fracture during the study period were included in the study and constituted Group 1. Of these, 39,524 women
ur
experienced at least two fractures during the study period and constituted Group 2;
Jo
of those, 7,656 had at least three fractures during the study period and constituted Group 3 (Figure 1). The mean ages for women in Groups 1, 2 and 3 were 74.1, 79.3 and 81.8 years, respectively (Table 1). In Group 1, 22.8% of index fractures were at the hip and 14.7% were vertebral. In Group 2, 28.0% of index (second) fractures were at the hip and 24.5% were vertebral (Table 1). Group 3 index (third) fractures were not analyzed by fracture type due to small sample size. At least threequarters of women were not receiving any osteoporosis treatment in the 12 months prior to index fracture, even when they had suffered three prior fractures (Table 1). The proportions of patients with fracture risk factors (e.g. comorbidities, glucocorticoid use, recent hospitalization, outpatient specialist visits) were larger in Groups 1, 2 and 3 than in controls. These proportions increased with each additional prior fracture. For example, the proportions of patients with glucocorticoid use within 8
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
the last 12 months was 3.6%, 5.1% and 6.2% in Groups 1, 2 and 3, respectively (Table 1). Generally, the proportion of patients with fracture risk factors increased with age across all Groups and their respective controls (Supplementary 2).
a
Jo
ur
na
lP
re
-p
ro of
Figure 1. Patient population
Group 3 index (third) fractures were not analyzed by fracture type due to small sample size.
9
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
Journal Pre-proof
15 February 2020
Table 1. Baseline patient characteristics Group 1
Group 2
1 fracture during study period (n=231,769)
Controlsb (n=231,769)
2 fractures during study period (n=39,524)
74.1 (74.1–74.2)
74.1 (74.1–74.2)
79.3 (79.2–79.4)
Secondary osteoporosis, % (95% CI)
9.4 (9.3–9.5)
6.3 (6.2–6.3)
11.9 (11.6–12.2)
Rheumatoid arthritis, % (95% CI)
2.4 (2.4–2.5)
1.6 (1.6–1.7)
Charlson comorbidity index, mean (95% CI)
0.4 (0.4–0.4)
0.3 (0.3–0.3)
Group 3
Controlsb (n=39,524)
3 fractures during study period (n=7,656)
Controlsb (n=7,656)
79.3 (79.2–79.4)
81.8 (81.5–82.0)
81.8 (81.5–82.0)
6.3 (6.1–6.4)
13.8 (13.0–14.6)
6.1 (5.8–6.4)
2.9 (2.7–3.0)
1.6 (1.6–1.7)
3.2 (2.8–3.6)
1.6 (1.4–1.7)
0.6 (0.6–0.6)
0.4 (0.4–0.4)
0.7 (0.6–0.7)
0.4 (0.4–0.4)
n r u
1.4 (1.4–1.4)
7.2 (7.0–7.3)
1.7 (1.6–1.7)
10.8 (10.4–11.3)
1.8 (1.7–1.9)
1.5 (1.5–1.6)
1.1 (1.1–1.1)
2.5 (2.4–2.5)
1.2 (1.2–1.2)
3.1 (3.0–3.3)
1.2 (1.2–1.3)
No osteoporosis medications during the 12 months before index fracture date, % (95% CI)
81.8 (81.6–81.9)
82.4 (82.4–82.5)
77.8 (77.4–78.3)
83.4 (83.2–83.6)
74.9 (73.9–75.9)
83.3 (82.8–83.8)
Glucocorticoid use during the 12 months before index fracture date, % (95% CI)c
3.6 (3.6–3.7)
2.5 (2.5–2.6)
5.1 (4.8–5.3)
2.9 (2.8-3.0)
6.2 (5.7–6.8)
3.3 (3.1–3.5)
73.7 (73.6–73.9)
68.2 (68.1–68.3)
85.7 (85.4–86.1)
75.1 (74.9–75.4)
90.8 (90.1–91.4)
78.6 (78.1–79.1)
Patient characteristicsa
Mean age at index fracture date, years (95% CI)
Days of all-cause hospitalization during the 12 months before index fracture date, mean (95% CI)
o J
Number of outpatient physician specialist visits during the 12 months before index fracture date (all-cause), mean (95% CI)
Exposure to drugs increasing the risk of falls during the 12 months before index fracture date, % (95% CI)
2.7 (2.7–2.7)
l a
o r p
e
r P
f o
10
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
15.3 (15.2–15.5)
9.7 (9.6–9.8)
30.0 (29.5–30.4)
14.2 (14.0–14.4)
42.6 (41.4–43.7)
16.6 (16.1–17.1)
6.2 (6.2–6.2)
5.3 (5.3–5.3)
8.2 (8.2–8.3)
5.9 (5.9–5.9)
9.6 (9.5–9.7)
6.3 (6.2–6.3)
Hip, n (%)
52,381 (22.6)
NA
11,054 (28.0)
NA
NRd
NA
Vertebral, n (%)
33,965 (14.7)
NA
9,675 (24.5)
NA
NRd
NA
Pre-packaged drug dispensing, % (95% CI) Number of different medications prescribed within the 12 months before index fracture date, mean (95% CI) Index fracture type
Journal Pre-proof
a
o r p
f o
Index fracture refers to the first fragility fracture during the study period for Group 1; the second for Group 2; and the third for Group 3. Index fracture date was defined as
the date of index fracture. bSampled from a pool of 260,401 women from the general population without a history of fracture at the index date. cGlucocorticoid use defined by
e
the FRAX® algorithm. dNot shown due to insufficient sample size. CI: confidence interval; NA: not applicable; NR: not reported.
l a
r P
n r u
o J
11
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) 3.2. Cumulative fracture incidence These descriptive data demonstrate that at all time points, cumulative incidence of subsequent MOF was higher in patients with a history of fracture as compared to controls with no history of fracture (Figure 2). One-year cumulative incidence of subsequent MOF was 5.1% for Group 1 as compared to 2.5% in matched controls. For Group 2 and matched controls, cumulative incidence was 8.6% and 3.3%, respectively. For Group 3 and matched controls, cumulative incidence was 5.4% and 3.5%, respectively.
ro of
Two-year cumulative incidence of subsequent MOF was 9.7% for Group 1 as compared to 5.0% in matched controls. For Group 2 and matched controls, cumulative incidence was 15.5% and 6.4%, respectively. For Group 3 and matched
-p
controls, cumulative incidence was 9.0% and 7.0%, respectively.
The five-year cumulative incidence of any subsequent MOF in Group 1 was 20.7% as
re
compared to 12.3% in matched controls. In Group 2, five-year cumulative incidence was twice as high compared to controls (32.0% vs 15.3%). Cumulative incidence of
lP
subsequent MOF at five years for Group 3 is not reported due to limited follow-up for
na
this group.
Cumulative incidence of subsequent MOF differed based on index fracture type. In
ur
Group 1, five-year cumulative incidence of subsequent MOF in women with an index vertebral fracture was 44.2% versus 27.3% in women with an index hip fracture. In
Jo
Group 2, five-year cumulative incidence of subsequent MOF was 51.8% in women with an index vertebral fracture versus 29.1% in women with an index hip fracture (not shown).
12
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Figure 2. Cumulative incidence of subsequent MOF after one, two and
Jo
ur
na
lP
re
-p
ro of
three fractures during study period compared to no-fracture controlsa
a
Shaded areas show 95% confidence intervals. MOF: major osteoporotic fracture.
13
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) 3.3. Risk of subsequent MOF Risk of subsequent MOF in patients with an index fracture, after adjusting for baseline covariates and as compared to controls, are stratified by age and index fracture type in Table 2. Excess relative risk was highest in the year following the index fracture and decreased thereafter in all Groups. In Groups 1 and 2, risk of subsequent MOF for most age groups and time periods was elevated as compared to controls. In patients younger than 80 years in Group 3, risk was generally elevated for 0–24 months as compared to controls, while for patients ≥80 years it was lower
ro of
than in controls at 7–12 months and beyond. When comparing index fracture types, vertebral fractures were associated with the greatest amount of excess risk relative to controls, particularly during Months 0–24 following the index fracture. The excess risk then decreased over time, but risk
-p
remained elevated throughout the five-year follow-up period (Table 2). The excess risk of MOF following an index vertebral fracture was higher in the unadjusted model
re
throughout the five-year follow-up period than in the adjusted model which
lP
accounted for covariates, though adjusted risk remained elevated as compared to any other index fracture (Figure 3). In patients with an index hip fracture, excess
Jo
ur
fracture (Table 2).
na
risk was also elevated but was more stable than in patients with an index vertebral
14
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
Journal Pre-proof
15 February 2020
Table 2. Adjusted risk of subsequent MOF after one, two and three fractures compared with no fractures, by index fracture site and time since index fracturea Index fracture site
Months after index fracture Age (years) 0–6
7–12
13–18
f o
19–24
37–48
4960
3.0 (2.3–3.7)
2.9 (2.2–3.6)
2.8 (2.0–3.6)
2.5 (2.2–2.8)
2.4 (2.1–2.8)
2.3 (1.9–2.7)
2.3 (1.8–2.8)
1.8 (1.6–1.9)
1.7 (1.5–1.9)
1.7 (1.4–1.9)
1.6 (1.3–2.0)
1.2 (1.2–1.3)
1.2 (1.1–1.3)
1.2 (1.0–1.3)
1.2 (0.9–1.4)
ro
25–36
Group 1: Risk of MOF after one fracture during study period compared to no fractures, HR (95% CI)
Hip
Vertebral
Any
50
3.4 (2.7–4.2)
3.9 (3.1–4.8)
3.7 (2.9–4.5)
60
2.8 (2.5–3.1)
3.2 (2.8–3.6)
3.0 (2.6–3.4)
70
2.0 (1.8–2.1)
2.3 (2.1–2.5)
2.1 (1.9–2.4)
80
1.4 (1.3–1.5)
1.6 (1.5–1.7)
1.5 (1.4–1.6)
All ages (n=52,831)
1.6 (1.4–1.9)
1.8 (1.6–2.1)
1.7 (1.5–2.0)
1.4 (1.2–1.7)
1.4 (1.2–1.6)
1.3 (1.1–1.6)
1.3 (1.0–1.7)
50
8.2 (6.5–9.8)
5.9 (4.7–7.2)
5.5 (4.3–6.7)
4.8 (3.8–5.9)
4.3 (3.3–5.3)
3.6 (2.6–4.6)
2.4 (1.5–3.3)
60
6.5 (5.6–7.4)
4.8 (4.0–5.5)
4.4 (3.7–5.1)
3.9 (3.3–4.5)
3.5 (2.8–4.1)
2.9 (2.2–3.5)
1.9 (1.3–2.6)
70
5.1 (4.6–5.6)
3.7 (3.2–4.2)
3.4 (3.0–3.9)
3.0 (2.6–3.4)
2.7 (2.3–3.1)
2.2 (1.8–2.7)
1.5 (1.0–2.0)
80
3.9 (3.6–4.1)
2.8 (2.5–3.1)
2.6 (2.3–2.9)
2.3 (2.0–2.5)
2.0 (1.7–2.3)
1.7 (1.4–2.0)
1.1 (0.8–1.5)
All ages (n=9,675)
4.1 (3.4–5.0)
3.0 (2.5–3.7)
2.8 (2.3–3.4)
2.5 (2.0–3.0)
2.2 (1.8–2.7)
1.8 (1.4–2.4)
1.2 (0.8–1.7)
50
2.8 (2.7–3.0)
2.8 (2.6–3.0)
2.6 (2.4–2.7)
2.3 (2.2–2.4)
2.2 (2.0–2.3)
2.0 (1.9–2.2)
2.0 (1.8–2.2)
60
2.4 (2.3–2.5)
2.3 (2.2–2.4)
2.1 (2.0–2.2)
1.9 (1.8–2.0)
1.8 (1.7–1.9)
1.7 (1.6–1.8)
1.7 (1.5–1.8)
70
2.2 (2.1–2.2)
2.1 (2.0–2.2)
2.0 (1.9–2.1)
1.8 (1.7–1.8)
1.6 (1.6–1.7)
1.5 (1.4–1.6)
1.5 (1.4–1.7)
80
1.9 (1.8–1.9)
1.9 (1.8–1.9)
1.7 (1.6–1.8)
1.5 (1.5–1.6)
1.4 (1.4–1.5)
1.3 (1.2–1.4)
1.3 (1.2–1.4)
All ages (n=231,769)
2.1 (1.9–2.2)
2.0 (1.9–2.2)
1.9 (1.7–2.0)
1.7 (1.6–1.8)
1.6 (1.5–1.7)
1.5 (1.3–1.6)
1.5 (1.3–1.6)
l a
o J
n r u
3.1 (2.4–3.8)
e r P
-p
15
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
Index fracture site
Journal Pre-proof
15 February 2020
Months after index fracture Age (years)
0–6
7–12
13–18
19–24
25–36
37–48
4960
3.2 (1.1–5.4)
2.3 (0.5–4.1)
2.2 (-0.4–4.8)
3.0 (1.9–4.1)
2.1 (0.9–3.3)
2.0 (-0.1–4.2)
2.1 (1.4–2.7)
1.5 (0.7–2.2)
1.4 (-0.1–2.8)
1.3 (0.9–1.6)
0.9 (0.4–1.3)
0.8 (0.0–1.7)
Group 2: Risk of MOF after two fractures during study period compared to no fractures, HR (95% CI)
Hip
Vertebral
Any
50
2.6 (0.7–4.5)
3.9 (1.4–6.4)
3.0 (1.1–5.0)
3.1 (1.1–5.1)
60
3.9 (2.8–5.1)
3.6 (2.5–4.8)
2.8 (1.8–3.7)
2.9 (1.9–3.8)
70
2.7 (2.2–3.2)
2.5 (1.9–3.1)
1.9 (1.4–2.4)
2.0 (1.5–2.5)
80
1.7 (1.5–1.8)
1.5 (1.3–1.8)
1.2 (0.9–1.4)
1.2 (1.0–1.4)
All ages (n=11,054)
1.6 (1.1–2.5)
1.5 (1.0–2.3)
1.2 (0.7–1.8)
1.2 (0.8–1.8)
1.2 (0.8–2.0)
0.9 (0.5–2.6)
0.9 (0.3–2.5)
50
7.4 (4.2–10.7)
6.3 (3.4–9.2)
4.3 (2.3–6.3)
4.9 (2.5–7.4)
5.0 (2.1–7.8)
2.8 (0.1–5.5)
60
6.2 (4.7–7.8)
5.3 (3.8–6.8)
3.6 (2.5–4.7)
3.0 (2.1–3.9)
4.2 (2.7–5.6)
4.2 (2.2–6.1)
2.3 (0.2–4.5)
70
4.7 (3.8–5.5)
4.0 (3.1–4.9)
2.7 (2.0–3.4)
2.3 (1.7–2.8)
3.1 (2.1–4.1)
3.1 (1.8–4.5)
1.8 (0.2–3.3)
80
3.5 (3.1–3.9)
3.0 (2.4–3.5)
r P
3.6 (1.9–5.3)
2.0 (1.6–2.5)
1.7 (1.3–2.1)
2.3 (1.7–3.0)
2.3 (1.4–3.3)
1.3 (0.2–2.5)
All ages (n=9,675)
4.2 (3.0–5.9)
3.6 (2.5–5.2)
2.4 (1.7–3.6)
2.0 (1.4–3.0)
2.9 (1.8–4.3)
2.8 (1.7–4.7)
1.6 (0.6–4.0)
50
4.6 (3.9–5.3)
4.1 (3.4–4.8)
3.5 (2.9–4.1)
3.3 (2.7–3.8)
3.2 (2.6–3.8)
2.9 (2.2–3.7)
3.0 (1.7–4.2)
60
3.8 (3.4–4.2)
3.3 (2.9–3.8)
2.9 (2.5–3.3)
2.7 (2.3–3.0)
2.6 (2.2–3.0)
2.4 (1.9–2.9)
2.4 (1.4–3.4)
70
3.2 (2.9–3.4)
2.8 (2.5–3.1)
2.4 (2.2–2.7)
2.3 (2.0–2.5)
2.2 (1.9–2.5)
2.0 (1.6–2.5)
2.0 (1.2–2.9)
80
2.2 (2.1–2.4)
2.0 (1.8–2.2)
1.7 (1.6–1.9)
1.6 (1.4–1.7)
1.6 (1.3–1.8)
1.4 (1.1–1.7)
1.4 (0.9–2.0)
All ages (n=39,524)
2.4 (2.0–2.8)
2.1 (1.8–2.6)
1.8 (1.5–2.2)
1.7 (1.4–2.0)
1.7 (1.3–2.0)
1.5 (1.2–2.0)
1.5 (1.0–2.4)
l a
o J
n r u
o r p
f o
e
16
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
Index fracture site
Journal Pre-proof
15 February 2020
Months after index fracture Age (years)
0–6
7–12
13–18
19–24
25–36
37–48
4960
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
Group 3: Risk of MOF after three fractures during study period compared to no fractures, HR (95% CI)
Any
50
4.7 (2.4–7.0)
3.3 (1.5–5.1)
2.8 (1.2–4.4)
2.3 (1.0–3.6)
60
4.0 (2.9–5.2)
2.8 (1.8–3.8)
2.5 (1.5–3.4)
2.0 (1.2–2.8)
70
2.6 (2.0–3.3)
1.8 (1.3–2.4)
1.6 (1.0–2.2)
1.3 (0.8–1.7)
80
1.1 (0.9–1.3)
0.8 (0.6–1.0)
0.7 (0.5–0.9)
0.5 (0.4–0.7)
All ages (n=7,656)
1.8 (1.0–3.1)
1.2 (0.7–2.2)
1.1 (0.6–2.0)
0.9 (0.5–1.6)
a
p e
ro
f o
Table shows the predicted HRs (95% CI) for a woman aged 50, 60, 70 or 80 years old based on parametric survival models adjusted for baseline covariates. All women in the
r P
study were included in the models. Index fracture refers to the first fragility fracture during the study period for Group 1; the second for Group 2; and the third for Group 3. Index fracture date was defined as the date of index fracture. bNot shown due to insufficient sample size. CI: confidence interval; HR: hazard ratio; MOF: major osteoporotic
l a
fracture; NR: not reported.
n r u
o J
17
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Figure 3. Risk of subsequent MOF after index vertebral fracture by time
lP
re
-p
ro of
since index fracturea
a
Models adjusted and unadjusted for potentially confounding demographic and clinical covariates are
shown. Hazard ratios are calculated versus non-fracture controls. Shaded areas show 95% confidence
ur
4. DISCUSSION
na
intervals. MOF: major osteoporotic fracture.
Jo
The descriptive data reported in this study suggest that amongst women with one fragility fracture, one in five will sustain a subsequent MOF within five years; amongst women with two fragility fractures, this increases to one in three. Cumulative incidence of subsequent MOF was higher following an index vertebral fracture than following an index hip fracture or any other index fracture. In women with three fragility fractures, 12.1% sustained subsequent MOF within three years. After adjusting for covariates, relative risk of subsequent fracture following a vertebral fracture for a woman aged 70 years or older with two previous fractures is approximately 50% greater as compared to a woman of the same age but with one previous fracture. This suggests that prevention of the first and/or second fracture could alleviate the recurrence of fractures that are indicated by these data and emphasizes the importance of early treatment after a fracture. 18
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) 4.1. Fracture history and age impact subsequent fracture risk Women <80 years of age with a history of one or more fractures demonstrated increased risk of subsequent MOF as compared to controls at all time points. However, in women aged ≥80 years with a history of three fractures, relative risk of subsequent MOF between 7 and 24 months was lower than in women with no prior fractures. It was not possible in this study to determine the reason for the reduced risk of subsequent MOF in women aged 80 years and older and three fractures, but a possible reason is limited mobility (and thereby reduced exposure to falls); 42% of women in this group utilized assisted drug dispensing, indicating some level of
ro of
dependence. Additionally, older patients are inherently at greater risk of death. Therefore, competing mortality risk may have reduced the time at risk of subsequent MOF for this group, leading to fewer subsequent MOFs experienced by this age
-p
group and contributing to the lower risk seen for women ≥80 years, whereas younger women did not exhibit a reduction in relative risk.
re
This study accounts for several factors that increase fracture risk, yet still the excess
lP
risk in women with a history of fracture remains, indicating that prior fracture represents an independent risk factor for future fracture for at least five years. However, while our results echo earlier findings that show increased subsequent
na
fracture risk in women with a history of fracture [15, 26], and specifically in the two years immediately following a fracture, the magnitude of this increase was not as
ur
dramatic as that reported elsewhere [12, 16]. This may be a result of other studies only adjusting for age and not for other potential risk factors, which may have led to
Jo
higher estimated risk in those analyses. 4.2. Strengths and limitations This study compared the risk of subsequent fractures in a large sample of patients with a history of fracture versus non-fractured controls of the same age while adjusting for key clinical and demographic characteristics. Gehlbach et al. also compared groups of patients with one or more previous fractures, controlling for age and type of previous fracture; however, unlike our study, the Gehlbach study was questionnaire-based and excluded patients who declined, or were unable, to complete the survey [27].
19
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Additional strengths of our study derive from the large datasets, the robustness and completeness of the registries, and the fact that all inpatient and specialized outpatient visits associated with fractures occurring in Sweden between 2006 and 2012 were included, improving the reliability of the fracture counts. This study also compared data with a large random sample from the total population. Furthermore, unlike previous studies [12, 16], adjustments were made in this study for several potentially confounding clinical and demographic variables, though not all potential risk variables were universally available (e.g. BMD, body mass index). In this study, the statistical adjustments had a clear impact on the results, suggesting that
ro of
adjusting for covariates is important for these types of analyses and should be included in future research.
This study also has limitations. Despite a large initial study sample, fracture data for
-p
women with three prior fractures were only reportable up to three years due to limited follow-up. In addition, fracture data were based on diagnostic coding and
re
were not adjudicated. To avoid counting the same incident fracture more than once, fractures at the same site in the body were counted only if they occurred at least six
lP
months (inpatient) or 12 months (outpatient) after the index fracture. This may have resulted in either underestimation or overestimation of fracture incidence but is
na
difficult to determine with the available data. In addition, asymptomatic vertebral fractures are often undiagnosed and therefore may have been undercounted. This
ur
may have led to either an overestimate or underestimate of the hazard ratios, depending on the population (fractured patients versus controls) that sustained an
Jo
undiagnosed fracture. Lastly, high-trauma and pathological fractures were not specifically excluded, though all included fractures were associated with osteoporosis, so this is unlikely to have had a large impact on the results. 4.3. Conclusions The results from this study indicate that women with a recent fracture are at increased risk of MOF as compared to age- and sex-matched women without history of fracture. The relative risk increased with each subsequent fracture and was highest in the two years following the fracture; for younger women, it remained high within 2–5 years. These results support treatment initiation early after a fracture and beyond, suggesting that treatment timing should be accounted for in health economic assessments of fracture prevention. 20
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) ACKNOWLEDGEMENTS The authors thank the patients, the investigators and their teams who took part in this study. The authors also acknowledge Helen Chambers, PhD, Costello Medical, UK for publication coordination and Kristian Clausen, MPH, and Simon Foulcer, PhD from Costello Medical, UK, for medical writing and editorial assistance based on the authors’ input and direction. This study was funded by UCB Pharma and Amgen Inc. FUNDING This study was sponsored by UCB Pharma and Amgen Inc. Support for third-party
ro of
writing assistance for this article, provided by Kristian Clausen, MPH, Costello Medical, UK was funded by UCB Pharma and Amgen Inc. in accordance with Good Publication
AUTHORS’ CONTRIBUTIONS
-p
Practice (GPP3) guidelines (http://www.ismpp.org/gpp3).
re
Substantial contributions to study conception and design: ES, OS, AS, KEA, FB, JB, ET, MC, CL; substantial contributions to analysis and interpretation of the data: ES,
lP
OS, AS, KEA, FB, JB, ET, MC, CL; drafting the article or revising it critically for important intellectual content: ES, OS, AS, KEA, FB, JB, ET, MC, CL; final approval of
ur
DISCLOSURES
na
the version of the article to be published: ES, OS, AS, KEA, FB, JB, ET, MC, CL.
ES, OS, FB, JB: employed by Quantify Research and funded by UCB Pharma and
Jo
Amgen Inc. to conduct this study; AS: received lecture fees from Amgen Inc., Eli Lilly and Mylan; KA: received lecture fees from Amgen Inc., Eli Lilly, Merck and UCB Pharma; ET, CL, MT: employed by and stockholders of UCB Pharma.
21
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) REFERENCES [1] E. Hernlund, A. Svedbom, M. Ivergard, J. Compston, C. Cooper, J. Stenmark, E.V. McCloskey, B. Jonsson, J.A. Kanis, Osteoporosis in the European Union: medical management, epidemiology and economic burden. A report prepared in collaboration with the International Osteoporosis Foundation (IOF) and the European Federation of Pharmaceutical Industry Associations (EFPIA), Archives of osteoporosis 8 (2013) 136. [2] N.C. Wright, A.C. Looker, K.G. Saag, J.R. Curtis, E.S. Delzell, S. Randall, B. Dawson-Hughes, The recent prevalence of osteoporosis and low bone mass in the United States based on bone mineral density at the femoral neck or lumbar spine, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 29(11) (2014) 2520-2526.
ro of
[3] A. Svedbom, E. Hernlund, M. Ivergard, J. Compston, C. Cooper, J. Stenmark, E.V. McCloskey, B. Jonsson, J.A. Kanis, Osteoporosis in the European Union: a compendium of country-specific reports, Archives of osteoporosis 8 (2013) 137.
re
-p
[4] C. Kerr, C. Bottomley, S. Shingler, L. Giangregorio, H.M. de Freitas, C. Patel, S. Randall, D.T. Gold, The importance of physical function to people with osteoporosis, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 28(5) (2017) 1597-1607.
na
lP
[5] S.L. Silverman, M.E. Minshall, W. Shen, K.D. Harper, S. Xie, The relationship of health-related quality of life to prevalent and incident vertebral fractures in postmenopausal women with osteoporosis: results from the Multiple Outcomes of Raloxifene Evaluation Study, Arthritis and rheumatism 44(11) (2001) 2611-9.
ur
[6] E. Seeman, P.D. Delmas, Bone quality--the material and structural basis of bone strength and fragility, The New England journal of medicine 354(21) (2006) 2250-61.
Jo
[7] J.A. Kanis, F. Borgstrom, C. De Laet, H. Johansson, O. Johnell, B. Jonsson, A. Oden, N. Zethraeus, B. Pfleger, N. Khaltaev, Assessment of fracture risk, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 16(6) (2005) 581-9. [8] J.R. Center, D. Bliuc, T.V. Nguyen, J.A. Eisman, Risk of subsequent fracture after low-trauma fracture in men and women, JAMA 297(4) (2007) 387-94. [9] F. Cosman, S.J. de Beur, M.S. LeBoff, E.M. Lewiecki, B. Tanner, S. Randall, R. Lindsay, Clinician's guide to prevention and treatment of osteoporosis, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 25(10) (2014) 2359-81. [10] J.A. Kanis, Diagnosis of osteoporosis and assessment of fracture risk, Lancet (London, England) 359(9321) (2002) 1929-36. [11] J.A. Eisman, E.R. Bogoch, R. Dell, J.T. Harrington, R.E. McKinney, Jr., A. McLellan, P.J. Mitchell, S. Silverman, R. Singleton, E. Siris, Making the first fracture 22
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) the last fracture: ASBMR task force report on secondary fracture prevention, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 27(10) (2012) 2039-46. [12] H. Johansson, K. Siggeirsdottir, N.C. Harvey, A. Oden, V. Gudnason, E. McCloskey, G. Sigurdsson, J.A. Kanis, Imminent risk of fracture after fracture, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 28(3) (2017) 775-780. [13] J.A. Kanis, O. Johnell, C. De Laet, H. Johansson, A. Oden, P. Delmas, J. Eisman, S. Fujiwara, P. Garnero, H. Kroger, E.V. McCloskey, D. Mellstrom, L.J. Melton, H. Pols, J. Reeve, A. Silman, A. Tenenhouse, A meta-analysis of previous fracture and subsequent fracture risk, Bone 35(2) (2004) 375-82.
-p
ro of
[14] L.A. Ahmed, J.R. Center, A. Bjornerem, D. Bluic, R.M. Joakimsen, L. Jorgensen, H.E. Meyer, N.D. Nguyen, T.V. Nguyen, T.K. Omsland, J. Stormer, G.S. Tell, T.A. van Geel, J.A. Eisman, N. Emaus, Progressively increasing fracture risk with advancing age after initial incident fragility fracture: the Tromso study, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 28(10) (2013) 2214-21.
lP
re
[15] C.M. Klotzbuecher, P.D. Ross, P.B. Landsman, T.A. Abbott, 3rd, M. Berger, Patients with prior fractures have an increased risk of future fractures: a summary of the literature and statistical synthesis, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 15(4) (2000) 721-39.
na
[16] T.A. van Geel, S. van Helden, P.P. Geusens, B. Winkens, G.J. Dinant, Clinical subsequent fractures cluster in time after first fractures, Annals of the rheumatic diseases 68(1) (2009) 99-102.
Jo
ur
[17] A. Balasubramanian, J. Zhang, L. Chen, D. Wenkert, S.G. Daigle, A. Grauer, J.R. Curtis, Risk of subsequent fracture after prior fracture among older women, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 30(1) (2019) 79-92. [18] S. van Helden, J. Cals, F. Kessels, P. Brink, G.J. Dinant, P. Geusens, Risk of new clinical fractures within 2 years following a fracture, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 17(3) (2006) 348-54. [19] T.A. Laurs-van Geel, J.R. Center, P.P. Geusens, G.J. Dinant, J.A. Eisman, Clinical fractures cluster in time after initial fracture, Maturitas 67(4) (2010) 339-42. [20] M. Bonafede, N. Shi, R. Barron, X. Li, D.B. Crittenden, D. Chandler, Predicting imminent risk for fracture in patients aged 50 or older with osteoporosis using US claims data, Archives of osteoporosis 11(1) (2016) 26. [21] C. Roux, K. Briot, Imminent fracture risk, Osteoporosis international : a journal established as result of cooperation between the European Foundation for 23
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Osteoporosis and the National Osteoporosis Foundation of the USA 28(6) (2017) 1765-1769. [22] J.A. Kanis, C. Cooper, R. Rizzoli, B. Abrahamsen, N.M. Al-Daghri, M.L. Brandi, J. Cannata-Andia, B. Cortet, H.P. Dimai, S. Ferrari, P. Hadji, N.C. Harvey, M. Kraenzlin, A. Kurth, E. McCloskey, S. Minisola, T. Thomas, J.Y. Reginster, Identification and management of patients at increased risk of osteoporotic fracture: outcomes of an ESCEO expert consensus meeting, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 28(7) (2017) 2023-2034.
ro of
[23] P. Royston, M.K. Parmar, Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects, Statistics in medicine 21(15) (2002) 2175-97.
re
-p
[24] E.S. Leib, K.G. Saag, J.D. Adachi, P.P. Geusens, N. Binkley, E.V. McCloskey, D.B. Hans, Official Positions for FRAX((R)) clinical regarding glucocorticoids: the impact of the use of glucocorticoids on the estimate by FRAX((R)) of the 10 year risk of fracture from Joint Official Positions Development Conference of the International Society for Clinical Densitometry and International Osteoporosis Foundation on FRAX((R)), Journal of clinical densitometry : the official journal of the International Society for Clinical Densitometry 14(3) (2011) 212-9.
lP
[25] M.E. Charlson, P. Pompei, K.L. Ales, C.R. MacKenzie, A new method of classifying prognostic comorbidity in longitudinal studies: development and validation, Journal of chronic diseases 40(5) (1987) 373-83.
ur
na
[26] O. Johnell, J.A. Kanis, A. Oden, I. Sernbo, I. Redlund-Johnell, C. Petterson, C. De Laet, B. Jonsson, Fracture risk following an osteoporotic fracture, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 15(3) (2004) 175-9.
Jo
[27] S. Gehlbach, K.G. Saag, J.D. Adachi, F.H. Hooven, J. Flahive, S. Boonen, R.D. Chapurlat, J.E. Compston, C. Cooper, A. Diez-Perez, S.L. Greenspan, A.Z. LaCroix, J.C. Netelenbos, J. Pfeilschifter, M. Rossini, C. Roux, P.N. Sambrook, S. Silverman, E.S. Siris, N.B. Watts, R. Lindsay, Previous fractures at multiple sites increase the risk for subsequent fractures: the Global Longitudinal Study of Osteoporosis in Women, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 27(3) (2012) 645-53.
24
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Highlights
Amongst women with one or two fragility fractures, 21–32% will experience a subsequent major osteoporotic fracture within five years.
After adjusting for covariates, women with ≥1 prior fracture are at increased relative risk of subsequent fracture within 24 months.
Vertebral fractures were associated with 2.5–4.1 times the risk of subsequent major osteoporotic fracture within 24 months.
Younger women with a prior fracture are at greatest relative risk of subsequent major osteoporotic fracture within five years.
ro of
This study highlights the need for treatment within two years of fracture,
ur
na
lP
re
-p
targeting patients at highest risk of subsequent fracture.
Jo
25
Emma Söreskog, Oskar Ström, Anna Spångéus, Kristina E. Åkesson, Fredrik Borgström, Jonas Banefelt, Emese Toth, Cesar Libanati, Mata Charokopou PII:
S8756-3282(20)30066-1
DOI:
https://doi.org/10.1016/j.bone.2020.115286
Reference:
BON 115286
To appear in:
Bone
Received date:
15 October 2019
Revised date:
16 January 2020
Accepted date:
14 February 2020
Please cite this article as: E. Söreskog, O. Ström, A. Spångéus, et al., Risk of major osteoporotic fracture after first, second and third fracture in Swedish women aged 50 years and older, Bone(2020), https://doi.org/10.1016/j.bone.2020.115286
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier.
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
Risk of Major Osteoporotic Fracture After First, Second and Third Fracture in Swedish Women Aged 50 Years and Older Emma Söreskog,a Oskar Ström,b Anna Spångéus,c Kristina E Åkesson,d Fredrik Borgström,e Jonas Banefelt,f Emese Toth,g Cesar Libanati,h Mata Charokopoui a
Quantify Research, Hantverkargatan 8, SE-112 21 Stockholm, Sweden,
[email protected]; bQuantify Research, Hantverkargatan 8, SE-
ro of
112 21 Stockholm, Sweden and Karolinska Institutet, Medical Management Centre, SE-171 77, Stockholm, Sweden, [email protected]; cLinköping University, Sandbäcksgatan 7, SE-581 83, Linköping, Sweden,
-p
[email protected]; dLund University, Skåne University Hospital, Inga Marie
re
Nilssons gata 22, SE-205 02 Malmö, Sweden, [email protected]; eQuantify
lP
Research, Hantverkargatan 8, SE-112 21 Stockholm, Sweden and Karolinska Institutet, Medical Management Centre, SE-171 77, Stockholm, Sweden,
na
[email protected]; fQuantify Research, Hantverkargatan 8, SE-112 21 Stockholm, Sweden, [email protected]; gUCB Pharma,
ur
Allée de la Recherche, 60 1070 Brussels, Belgium, [email protected]; hUCB
i
Jo
Pharma, Allée de la Recherche, 60 1070 Brussels, Belgium, [email protected]; UCB Pharma, Allée de la Recherche, 60 1070 Brussels, Belgium,
[email protected]
Correspondence to: Emma Söreskog; [email protected] Short title: Fracture after Fracture Funding: UCB Pharma and Amgen Inc. Abbreviations: Anatomical Therapeutic Chemical Classification System (ATC); bone mineral density (BMD); confidence interval (CI); hazard ratio (HR); International 1
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
Classification of Diseases, 10th Revision (ICD-10); major osteoporotic fracture
Jo
ur
na
lP
re
-p
ro of
(MOF); standard deviation (SD)
2
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
Abstract
Background Osteoporosis affects approximately one in five European women and leads to fragility fractures, which result in poor health, social and economic consequences. Fragility fractures are a strong risk factor for subsequent major osteoporotic fracture (MOF), with risk of MOF being elevated in the 1–2 years following an earlier fracture, a concept described as “imminent risk”. This study examines risk of subsequent MOF in patients with one, two or three prior fractures by age and type of fracture.
ro of
Methods
In this retrospective, observational cohort study, Swedish women aged ≥50 years with ≥1 any clinical fragility fracture between July 1, 2006–December 31, 2012 were
-p
identified from Sweden’s National Patient Register. Each patient was age- and sexmatched to three controls without history of fracture. Group 1 women included those
re
with one fragility fracture during the study period; Group 2 included those with two fragility fractures; and Group 3 included those with three fragility fractures. “Index
lP
fracture” was defined as the first fracture during the study period for Group 1; the second for Group 2; and the third for Group 3. Patients in each cohort and matched
na
controls were followed for up to 60 months or until subsequent MOF (hip, vertebra,
Results
ur
forearm, humerus), death or end of data availability.
Jo
231,769 women with at least one fracture were included in the study and therefore constituted Group 1; of these, 39,524 constituted Group 2 and of those, 7,656 constituted Group 3. At five years, cumulative incidence of subsequent MOF was higher in patients with a history of fracture as compared to controls (Group 1: 20.7% vs 12.3%; Group 2: 32.0% vs 15.3%). Three-year cumulative incidence for Group 3 was 12.1% (vs 10.7% for controls). After adjusting for baseline covariates, risk of subsequent MOF was highest within 0–24 months following an index fracture, then decreased but remained elevated as compared to controls. Having two prior fractures, vertebral fractures and younger age at time of index fracture were associated with greater relative risk.
3
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
Conclusions Women with a history of osteoporotic fracture are at increased risk of subsequent fracture, which is highest during the first 24 months following a fracture. Younger women and those with vertebral fractures are at greatest relative risk, suggesting that treatment should target these patients and be timely enough to impact the period of imminent risk.
Jo
ur
na
lP
re
-p
ro of
Key words: fracture incidence; fragility fracture; imminent risk; osteoporosis
4
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
1. INTRODUCTION Osteoporosis, as defined by densitometric measurement, is a major cause of fragility fractures, and affects approximately 21% of women aged 50–84 in the EU and 15.4% of women aged 50 and older in the US [1, 2]. Fragility fractures not only lead to poor health and social outcomes such as pain and disability [3-5], but are also associated with a significant economic burden. For example, in the EU, fragility fractures accounted for €26 billion and 26,300 life-years lost in 2010 alone [1]. Fragility fractures occur as a consequence of insufficient bone strength due to
ro of
deficits in bone mass and microstructure [6]. Risk factors for fracture include increased age, low bone mineral density (BMD), female sex and glucocorticoid use, among others [7-10]. Additionally, a history of fragility fracture has been identified as a strong driver of subsequent fracture risk [7, 9-14]. The risk of any subsequent
-p
fracture was reported to be twice that of a first fracture using pooled data from a large meta-analysis of studies published between 1966 and 1999 [15], and was even
re
higher (approximately four-fold) for subsequent vertebral fractures in women with a
lP
prior vertebral fracture. Similar findings were observed by others; Ahmed et al. (2013) found that an initial fracture was associated with 1.3–2.0 times the risk (for women and men, respectively) as compared to patients without an earlier fracture,
na
with that risk concentrated within the first five years post-fracture [14]. Additionally, van Geel et al. (2009) estimated that risk of subsequent fracture within 15 years of a
ur
first fracture is approximately double that of patients without a first fracture [16].
Jo
It is now clear that the increase in subsequent fracture risk following an initial fracture is not linear over time. While van Geel et al. demonstrated a doubling of risk resulting from a prior fracture, the same study found a five-fold risk of subsequent major osteoporotic fracture (MOF) in the year following the first fracture [16]. Other studies have reported that up to one-fourth of patients will sustain a second fracture during the 1–2 years following an earlier fracture, depending on the type of first fracture [17, 18]. This pattern of elevated subsequent fracture risk in the period immediately following a fragility fracture has been described as “imminent fracture risk” [12, 19-22]. What is less clear is how the risk of a subsequent fracture changes following a second or third fracture, and how a patient’s age and prior fracture type influence the level of imminent risk.
5
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
This study examines the risk of subsequent MOF, defined as those at the hip, vertebra, forearm or humerus, in patients with a history of one, two or three prior fractures of any type, compared with age- and sex-matched controls without history of fracture. 2. METHODS 2.1. Patients and study design In this retrospective, observational cohort study, Swedish females aged ≥50 years with at least one fragility fracture occurring during the study period of July 1, 2006–
ro of
December 31, 2012 were identified from the Swedish National Patient Register (Figure 1). Women with clinical fragility fractures at the hip, vertebra, shoulder, upper arm, pelvis, femur, lower leg, wrist, hand, rib and/or sternum were included.
-p
Women identified to have a fracture during the period of January 1, 2001–June 30, 2006 were excluded. Women identified to have Paget’s disease or any malignancy
re
(other than basal cell carcinoma) at any time during the study were excluded from
lP
the analysis.
Three separate analysis cohorts were identified. Group 1 included all women with at least one fragility fracture during the study period; Group 2 was constituted of the
na
women in Group 1 who sustained a second fragility fracture during the study period; and Group 3 was constituted of the women in Group 2 who sustained a third fragility
ur
fracture during the study period. “Index fracture” was defined as the fragility fracture
Jo
that classified a patient in a given Group. Therefore, the index fracture for Group 1 was their first fragility fracture to occur during the study period. For Group 2, the index fracture was their second to occur during the study period, and for Group 3, the index fracture was their third fracture during the study period. The date of the index fracture was defined as the index date. The baseline period during which patient characteristics data were retrieved was defined as the 12-month period before the index fracture date. For each group, each woman was age- and sex-matched with replacement to three controls without a history of fracture on, or before (dating back to January 1, 2001), the index fracture date. Controls were randomly selected from a population of 260,401 women ≥50 years old included in the Swedish Total Population Register and were assigned the index date of their matched counterpart. Patients in each cohort and their matched 6
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
controls were followed for up to 60 months or until subsequent MOF, death or end of data availability. Index fracture was defined as any fragility fracture because all types are associated with increased fracture risk [15]. For the outcome fracture, we selected MOFs as they represent the group of fractures with the greatest burden. 2.2. Data sources The National Patient Register was used to identify patients for inclusion; extracted data included sex, patient identification numbers, diagnosis codes, dates of diagnosis, duration of hospitalization and specialist visits. Diagnosis codes for each index and
ro of
subsequent fracture type are listed in Supplementary 1. Patient dispensed drug data were obtained from Sweden’s Prescribed Drug Register; extracted data included patient identification numbers, Anatomical Therapeutic Chemical Classification System (ATC) codes, prescription dates, dispensing dates, defined daily doses per
-p
prescribed package and number of pills/injections, etc. Sweden’s Cause of Death Register was used to establish patients’ dates of death using the national person
re
identification numbers based on birthdate. Clinical and demographic characteristics
lP
were collected at the date of each index fracture or in the 12-month period prior to the index fracture date. Matched data were retrieved for controls from the Swedish
na
Total Population Register.
2.3. Identifying fractures
ur
To avoid counting the same index fracture more than once, fractures at the same
Jo
site in the body and identified during an inpatient hospitalization were counted only if they occurred at least six months after the index fracture. Fractures at the same site of an earlier fracture and identified during an outpatient visit were counted only if they occurred at least 12 months after the index fracture. Hip fractures required an inpatient hospital admission. 2.4. Data analysis MOFs included hip, vertebral, forearm and humerus fractures. Incidence of subsequent MOF was calculated for Groups 1, 2 and 3 and compared with the respective control group. The relationship between having an index fracture and risk of subsequent fracture was assessed via the Royston-Parmar parametric spline model with MOF as a single failure event [23]. Data were right-censored at time of death or end of follow-up. Hazard ratios were assessed in separate time windows: 0– 7
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
6 months, 7–12 months, 13–18 months, 19–24 months, 25–36 months, 37–48 months and 49–60 months after index fracture. The time windows were chosen after iteratively testing the proportional hazard assumption by different windows. These analyses were performed on each group overall and by age (50–59, 60–69, 70–79 and ≥80 years). The risk of subsequent MOF was adjusted for several baseline covariates known to be associated with increased risk of fragility fracture: attained age; use of osteoporosis medication in the past 12 months; use of drugs known to increase the risk of falls in the past 12 months; glucocorticoid use in the past 12 months (defined
ro of
according to the FRAX® algorithm as past or present exposure of ≥5 mg daily dosage of prednisolone or equivalent for ≥3 months [24]); number of different medications prescribed in the past 12 months; secondary osteoporosis; rheumatoid arthritis;
-p
Charlson comorbidity index measured two years prior to index date [25], and prepackaged drug dispensing (defined as use of the ApoDos system within the past 12
re
months), which was used as a proxy for being “dependent”.
lP
3. RESULTS
3.1. Baseline characteristics
na
A total of 231,769 women with at least one index fracture during the study period were included in the study and constituted Group 1. Of these, 39,524 women
ur
experienced at least two fractures during the study period and constituted Group 2;
Jo
of those, 7,656 had at least three fractures during the study period and constituted Group 3 (Figure 1). The mean ages for women in Groups 1, 2 and 3 were 74.1, 79.3 and 81.8 years, respectively (Table 1). In Group 1, 22.8% of index fractures were at the hip and 14.7% were vertebral. In Group 2, 28.0% of index (second) fractures were at the hip and 24.5% were vertebral (Table 1). Group 3 index (third) fractures were not analyzed by fracture type due to small sample size. At least threequarters of women were not receiving any osteoporosis treatment in the 12 months prior to index fracture, even when they had suffered three prior fractures (Table 1). The proportions of patients with fracture risk factors (e.g. comorbidities, glucocorticoid use, recent hospitalization, outpatient specialist visits) were larger in Groups 1, 2 and 3 than in controls. These proportions increased with each additional prior fracture. For example, the proportions of patients with glucocorticoid use within 8
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
the last 12 months was 3.6%, 5.1% and 6.2% in Groups 1, 2 and 3, respectively (Table 1). Generally, the proportion of patients with fracture risk factors increased with age across all Groups and their respective controls (Supplementary 2).
a
Jo
ur
na
lP
re
-p
ro of
Figure 1. Patient population
Group 3 index (third) fractures were not analyzed by fracture type due to small sample size.
9
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
Journal Pre-proof
15 February 2020
Table 1. Baseline patient characteristics Group 1
Group 2
1 fracture during study period (n=231,769)
Controlsb (n=231,769)
2 fractures during study period (n=39,524)
74.1 (74.1–74.2)
74.1 (74.1–74.2)
79.3 (79.2–79.4)
Secondary osteoporosis, % (95% CI)
9.4 (9.3–9.5)
6.3 (6.2–6.3)
11.9 (11.6–12.2)
Rheumatoid arthritis, % (95% CI)
2.4 (2.4–2.5)
1.6 (1.6–1.7)
Charlson comorbidity index, mean (95% CI)
0.4 (0.4–0.4)
0.3 (0.3–0.3)
Group 3
Controlsb (n=39,524)
3 fractures during study period (n=7,656)
Controlsb (n=7,656)
79.3 (79.2–79.4)
81.8 (81.5–82.0)
81.8 (81.5–82.0)
6.3 (6.1–6.4)
13.8 (13.0–14.6)
6.1 (5.8–6.4)
2.9 (2.7–3.0)
1.6 (1.6–1.7)
3.2 (2.8–3.6)
1.6 (1.4–1.7)
0.6 (0.6–0.6)
0.4 (0.4–0.4)
0.7 (0.6–0.7)
0.4 (0.4–0.4)
n r u
1.4 (1.4–1.4)
7.2 (7.0–7.3)
1.7 (1.6–1.7)
10.8 (10.4–11.3)
1.8 (1.7–1.9)
1.5 (1.5–1.6)
1.1 (1.1–1.1)
2.5 (2.4–2.5)
1.2 (1.2–1.2)
3.1 (3.0–3.3)
1.2 (1.2–1.3)
No osteoporosis medications during the 12 months before index fracture date, % (95% CI)
81.8 (81.6–81.9)
82.4 (82.4–82.5)
77.8 (77.4–78.3)
83.4 (83.2–83.6)
74.9 (73.9–75.9)
83.3 (82.8–83.8)
Glucocorticoid use during the 12 months before index fracture date, % (95% CI)c
3.6 (3.6–3.7)
2.5 (2.5–2.6)
5.1 (4.8–5.3)
2.9 (2.8-3.0)
6.2 (5.7–6.8)
3.3 (3.1–3.5)
73.7 (73.6–73.9)
68.2 (68.1–68.3)
85.7 (85.4–86.1)
75.1 (74.9–75.4)
90.8 (90.1–91.4)
78.6 (78.1–79.1)
Patient characteristicsa
Mean age at index fracture date, years (95% CI)
Days of all-cause hospitalization during the 12 months before index fracture date, mean (95% CI)
o J
Number of outpatient physician specialist visits during the 12 months before index fracture date (all-cause), mean (95% CI)
Exposure to drugs increasing the risk of falls during the 12 months before index fracture date, % (95% CI)
2.7 (2.7–2.7)
l a
o r p
e
r P
f o
10
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
15 February 2020
15.3 (15.2–15.5)
9.7 (9.6–9.8)
30.0 (29.5–30.4)
14.2 (14.0–14.4)
42.6 (41.4–43.7)
16.6 (16.1–17.1)
6.2 (6.2–6.2)
5.3 (5.3–5.3)
8.2 (8.2–8.3)
5.9 (5.9–5.9)
9.6 (9.5–9.7)
6.3 (6.2–6.3)
Hip, n (%)
52,381 (22.6)
NA
11,054 (28.0)
NA
NRd
NA
Vertebral, n (%)
33,965 (14.7)
NA
9,675 (24.5)
NA
NRd
NA
Pre-packaged drug dispensing, % (95% CI) Number of different medications prescribed within the 12 months before index fracture date, mean (95% CI) Index fracture type
Journal Pre-proof
a
o r p
f o
Index fracture refers to the first fragility fracture during the study period for Group 1; the second for Group 2; and the third for Group 3. Index fracture date was defined as
the date of index fracture. bSampled from a pool of 260,401 women from the general population without a history of fracture at the index date. cGlucocorticoid use defined by
e
the FRAX® algorithm. dNot shown due to insufficient sample size. CI: confidence interval; NA: not applicable; NR: not reported.
l a
r P
n r u
o J
11
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) 3.2. Cumulative fracture incidence These descriptive data demonstrate that at all time points, cumulative incidence of subsequent MOF was higher in patients with a history of fracture as compared to controls with no history of fracture (Figure 2). One-year cumulative incidence of subsequent MOF was 5.1% for Group 1 as compared to 2.5% in matched controls. For Group 2 and matched controls, cumulative incidence was 8.6% and 3.3%, respectively. For Group 3 and matched controls, cumulative incidence was 5.4% and 3.5%, respectively.
ro of
Two-year cumulative incidence of subsequent MOF was 9.7% for Group 1 as compared to 5.0% in matched controls. For Group 2 and matched controls, cumulative incidence was 15.5% and 6.4%, respectively. For Group 3 and matched
-p
controls, cumulative incidence was 9.0% and 7.0%, respectively.
The five-year cumulative incidence of any subsequent MOF in Group 1 was 20.7% as
re
compared to 12.3% in matched controls. In Group 2, five-year cumulative incidence was twice as high compared to controls (32.0% vs 15.3%). Cumulative incidence of
lP
subsequent MOF at five years for Group 3 is not reported due to limited follow-up for
na
this group.
Cumulative incidence of subsequent MOF differed based on index fracture type. In
ur
Group 1, five-year cumulative incidence of subsequent MOF in women with an index vertebral fracture was 44.2% versus 27.3% in women with an index hip fracture. In
Jo
Group 2, five-year cumulative incidence of subsequent MOF was 51.8% in women with an index vertebral fracture versus 29.1% in women with an index hip fracture (not shown).
12
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Figure 2. Cumulative incidence of subsequent MOF after one, two and
Jo
ur
na
lP
re
-p
ro of
three fractures during study period compared to no-fracture controlsa
a
Shaded areas show 95% confidence intervals. MOF: major osteoporotic fracture.
13
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) 3.3. Risk of subsequent MOF Risk of subsequent MOF in patients with an index fracture, after adjusting for baseline covariates and as compared to controls, are stratified by age and index fracture type in Table 2. Excess relative risk was highest in the year following the index fracture and decreased thereafter in all Groups. In Groups 1 and 2, risk of subsequent MOF for most age groups and time periods was elevated as compared to controls. In patients younger than 80 years in Group 3, risk was generally elevated for 0–24 months as compared to controls, while for patients ≥80 years it was lower
ro of
than in controls at 7–12 months and beyond. When comparing index fracture types, vertebral fractures were associated with the greatest amount of excess risk relative to controls, particularly during Months 0–24 following the index fracture. The excess risk then decreased over time, but risk
-p
remained elevated throughout the five-year follow-up period (Table 2). The excess risk of MOF following an index vertebral fracture was higher in the unadjusted model
re
throughout the five-year follow-up period than in the adjusted model which
lP
accounted for covariates, though adjusted risk remained elevated as compared to any other index fracture (Figure 3). In patients with an index hip fracture, excess
Jo
ur
fracture (Table 2).
na
risk was also elevated but was more stable than in patients with an index vertebral
14
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
Journal Pre-proof
15 February 2020
Table 2. Adjusted risk of subsequent MOF after one, two and three fractures compared with no fractures, by index fracture site and time since index fracturea Index fracture site
Months after index fracture Age (years) 0–6
7–12
13–18
f o
19–24
37–48
4960
3.0 (2.3–3.7)
2.9 (2.2–3.6)
2.8 (2.0–3.6)
2.5 (2.2–2.8)
2.4 (2.1–2.8)
2.3 (1.9–2.7)
2.3 (1.8–2.8)
1.8 (1.6–1.9)
1.7 (1.5–1.9)
1.7 (1.4–1.9)
1.6 (1.3–2.0)
1.2 (1.2–1.3)
1.2 (1.1–1.3)
1.2 (1.0–1.3)
1.2 (0.9–1.4)
ro
25–36
Group 1: Risk of MOF after one fracture during study period compared to no fractures, HR (95% CI)
Hip
Vertebral
Any
50
3.4 (2.7–4.2)
3.9 (3.1–4.8)
3.7 (2.9–4.5)
60
2.8 (2.5–3.1)
3.2 (2.8–3.6)
3.0 (2.6–3.4)
70
2.0 (1.8–2.1)
2.3 (2.1–2.5)
2.1 (1.9–2.4)
80
1.4 (1.3–1.5)
1.6 (1.5–1.7)
1.5 (1.4–1.6)
All ages (n=52,831)
1.6 (1.4–1.9)
1.8 (1.6–2.1)
1.7 (1.5–2.0)
1.4 (1.2–1.7)
1.4 (1.2–1.6)
1.3 (1.1–1.6)
1.3 (1.0–1.7)
50
8.2 (6.5–9.8)
5.9 (4.7–7.2)
5.5 (4.3–6.7)
4.8 (3.8–5.9)
4.3 (3.3–5.3)
3.6 (2.6–4.6)
2.4 (1.5–3.3)
60
6.5 (5.6–7.4)
4.8 (4.0–5.5)
4.4 (3.7–5.1)
3.9 (3.3–4.5)
3.5 (2.8–4.1)
2.9 (2.2–3.5)
1.9 (1.3–2.6)
70
5.1 (4.6–5.6)
3.7 (3.2–4.2)
3.4 (3.0–3.9)
3.0 (2.6–3.4)
2.7 (2.3–3.1)
2.2 (1.8–2.7)
1.5 (1.0–2.0)
80
3.9 (3.6–4.1)
2.8 (2.5–3.1)
2.6 (2.3–2.9)
2.3 (2.0–2.5)
2.0 (1.7–2.3)
1.7 (1.4–2.0)
1.1 (0.8–1.5)
All ages (n=9,675)
4.1 (3.4–5.0)
3.0 (2.5–3.7)
2.8 (2.3–3.4)
2.5 (2.0–3.0)
2.2 (1.8–2.7)
1.8 (1.4–2.4)
1.2 (0.8–1.7)
50
2.8 (2.7–3.0)
2.8 (2.6–3.0)
2.6 (2.4–2.7)
2.3 (2.2–2.4)
2.2 (2.0–2.3)
2.0 (1.9–2.2)
2.0 (1.8–2.2)
60
2.4 (2.3–2.5)
2.3 (2.2–2.4)
2.1 (2.0–2.2)
1.9 (1.8–2.0)
1.8 (1.7–1.9)
1.7 (1.6–1.8)
1.7 (1.5–1.8)
70
2.2 (2.1–2.2)
2.1 (2.0–2.2)
2.0 (1.9–2.1)
1.8 (1.7–1.8)
1.6 (1.6–1.7)
1.5 (1.4–1.6)
1.5 (1.4–1.7)
80
1.9 (1.8–1.9)
1.9 (1.8–1.9)
1.7 (1.6–1.8)
1.5 (1.5–1.6)
1.4 (1.4–1.5)
1.3 (1.2–1.4)
1.3 (1.2–1.4)
All ages (n=231,769)
2.1 (1.9–2.2)
2.0 (1.9–2.2)
1.9 (1.7–2.0)
1.7 (1.6–1.8)
1.6 (1.5–1.7)
1.5 (1.3–1.6)
1.5 (1.3–1.6)
l a
o J
n r u
3.1 (2.4–3.8)
e r P
-p
15
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
Index fracture site
Journal Pre-proof
15 February 2020
Months after index fracture Age (years)
0–6
7–12
13–18
19–24
25–36
37–48
4960
3.2 (1.1–5.4)
2.3 (0.5–4.1)
2.2 (-0.4–4.8)
3.0 (1.9–4.1)
2.1 (0.9–3.3)
2.0 (-0.1–4.2)
2.1 (1.4–2.7)
1.5 (0.7–2.2)
1.4 (-0.1–2.8)
1.3 (0.9–1.6)
0.9 (0.4–1.3)
0.8 (0.0–1.7)
Group 2: Risk of MOF after two fractures during study period compared to no fractures, HR (95% CI)
Hip
Vertebral
Any
50
2.6 (0.7–4.5)
3.9 (1.4–6.4)
3.0 (1.1–5.0)
3.1 (1.1–5.1)
60
3.9 (2.8–5.1)
3.6 (2.5–4.8)
2.8 (1.8–3.7)
2.9 (1.9–3.8)
70
2.7 (2.2–3.2)
2.5 (1.9–3.1)
1.9 (1.4–2.4)
2.0 (1.5–2.5)
80
1.7 (1.5–1.8)
1.5 (1.3–1.8)
1.2 (0.9–1.4)
1.2 (1.0–1.4)
All ages (n=11,054)
1.6 (1.1–2.5)
1.5 (1.0–2.3)
1.2 (0.7–1.8)
1.2 (0.8–1.8)
1.2 (0.8–2.0)
0.9 (0.5–2.6)
0.9 (0.3–2.5)
50
7.4 (4.2–10.7)
6.3 (3.4–9.2)
4.3 (2.3–6.3)
4.9 (2.5–7.4)
5.0 (2.1–7.8)
2.8 (0.1–5.5)
60
6.2 (4.7–7.8)
5.3 (3.8–6.8)
3.6 (2.5–4.7)
3.0 (2.1–3.9)
4.2 (2.7–5.6)
4.2 (2.2–6.1)
2.3 (0.2–4.5)
70
4.7 (3.8–5.5)
4.0 (3.1–4.9)
2.7 (2.0–3.4)
2.3 (1.7–2.8)
3.1 (2.1–4.1)
3.1 (1.8–4.5)
1.8 (0.2–3.3)
80
3.5 (3.1–3.9)
3.0 (2.4–3.5)
r P
3.6 (1.9–5.3)
2.0 (1.6–2.5)
1.7 (1.3–2.1)
2.3 (1.7–3.0)
2.3 (1.4–3.3)
1.3 (0.2–2.5)
All ages (n=9,675)
4.2 (3.0–5.9)
3.6 (2.5–5.2)
2.4 (1.7–3.6)
2.0 (1.4–3.0)
2.9 (1.8–4.3)
2.8 (1.7–4.7)
1.6 (0.6–4.0)
50
4.6 (3.9–5.3)
4.1 (3.4–4.8)
3.5 (2.9–4.1)
3.3 (2.7–3.8)
3.2 (2.6–3.8)
2.9 (2.2–3.7)
3.0 (1.7–4.2)
60
3.8 (3.4–4.2)
3.3 (2.9–3.8)
2.9 (2.5–3.3)
2.7 (2.3–3.0)
2.6 (2.2–3.0)
2.4 (1.9–2.9)
2.4 (1.4–3.4)
70
3.2 (2.9–3.4)
2.8 (2.5–3.1)
2.4 (2.2–2.7)
2.3 (2.0–2.5)
2.2 (1.9–2.5)
2.0 (1.6–2.5)
2.0 (1.2–2.9)
80
2.2 (2.1–2.4)
2.0 (1.8–2.2)
1.7 (1.6–1.9)
1.6 (1.4–1.7)
1.6 (1.3–1.8)
1.4 (1.1–1.7)
1.4 (0.9–2.0)
All ages (n=39,524)
2.4 (2.0–2.8)
2.1 (1.8–2.6)
1.8 (1.5–2.2)
1.7 (1.4–2.0)
1.7 (1.3–2.0)
1.5 (1.2–2.0)
1.5 (1.0–2.4)
l a
o J
n r u
o r p
f o
e
16
Submission Revisions – Fracture after Fracture (MOF 1,2,3)
Index fracture site
Journal Pre-proof
15 February 2020
Months after index fracture Age (years)
0–6
7–12
13–18
19–24
25–36
37–48
4960
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
NRb
Group 3: Risk of MOF after three fractures during study period compared to no fractures, HR (95% CI)
Any
50
4.7 (2.4–7.0)
3.3 (1.5–5.1)
2.8 (1.2–4.4)
2.3 (1.0–3.6)
60
4.0 (2.9–5.2)
2.8 (1.8–3.8)
2.5 (1.5–3.4)
2.0 (1.2–2.8)
70
2.6 (2.0–3.3)
1.8 (1.3–2.4)
1.6 (1.0–2.2)
1.3 (0.8–1.7)
80
1.1 (0.9–1.3)
0.8 (0.6–1.0)
0.7 (0.5–0.9)
0.5 (0.4–0.7)
All ages (n=7,656)
1.8 (1.0–3.1)
1.2 (0.7–2.2)
1.1 (0.6–2.0)
0.9 (0.5–1.6)
a
p e
ro
f o
Table shows the predicted HRs (95% CI) for a woman aged 50, 60, 70 or 80 years old based on parametric survival models adjusted for baseline covariates. All women in the
r P
study were included in the models. Index fracture refers to the first fragility fracture during the study period for Group 1; the second for Group 2; and the third for Group 3. Index fracture date was defined as the date of index fracture. bNot shown due to insufficient sample size. CI: confidence interval; HR: hazard ratio; MOF: major osteoporotic
l a
fracture; NR: not reported.
n r u
o J
17
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Figure 3. Risk of subsequent MOF after index vertebral fracture by time
lP
re
-p
ro of
since index fracturea
a
Models adjusted and unadjusted for potentially confounding demographic and clinical covariates are
shown. Hazard ratios are calculated versus non-fracture controls. Shaded areas show 95% confidence
ur
4. DISCUSSION
na
intervals. MOF: major osteoporotic fracture.
Jo
The descriptive data reported in this study suggest that amongst women with one fragility fracture, one in five will sustain a subsequent MOF within five years; amongst women with two fragility fractures, this increases to one in three. Cumulative incidence of subsequent MOF was higher following an index vertebral fracture than following an index hip fracture or any other index fracture. In women with three fragility fractures, 12.1% sustained subsequent MOF within three years. After adjusting for covariates, relative risk of subsequent fracture following a vertebral fracture for a woman aged 70 years or older with two previous fractures is approximately 50% greater as compared to a woman of the same age but with one previous fracture. This suggests that prevention of the first and/or second fracture could alleviate the recurrence of fractures that are indicated by these data and emphasizes the importance of early treatment after a fracture. 18
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) 4.1. Fracture history and age impact subsequent fracture risk Women <80 years of age with a history of one or more fractures demonstrated increased risk of subsequent MOF as compared to controls at all time points. However, in women aged ≥80 years with a history of three fractures, relative risk of subsequent MOF between 7 and 24 months was lower than in women with no prior fractures. It was not possible in this study to determine the reason for the reduced risk of subsequent MOF in women aged 80 years and older and three fractures, but a possible reason is limited mobility (and thereby reduced exposure to falls); 42% of women in this group utilized assisted drug dispensing, indicating some level of
ro of
dependence. Additionally, older patients are inherently at greater risk of death. Therefore, competing mortality risk may have reduced the time at risk of subsequent MOF for this group, leading to fewer subsequent MOFs experienced by this age
-p
group and contributing to the lower risk seen for women ≥80 years, whereas younger women did not exhibit a reduction in relative risk.
re
This study accounts for several factors that increase fracture risk, yet still the excess
lP
risk in women with a history of fracture remains, indicating that prior fracture represents an independent risk factor for future fracture for at least five years. However, while our results echo earlier findings that show increased subsequent
na
fracture risk in women with a history of fracture [15, 26], and specifically in the two years immediately following a fracture, the magnitude of this increase was not as
ur
dramatic as that reported elsewhere [12, 16]. This may be a result of other studies only adjusting for age and not for other potential risk factors, which may have led to
Jo
higher estimated risk in those analyses. 4.2. Strengths and limitations This study compared the risk of subsequent fractures in a large sample of patients with a history of fracture versus non-fractured controls of the same age while adjusting for key clinical and demographic characteristics. Gehlbach et al. also compared groups of patients with one or more previous fractures, controlling for age and type of previous fracture; however, unlike our study, the Gehlbach study was questionnaire-based and excluded patients who declined, or were unable, to complete the survey [27].
19
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Additional strengths of our study derive from the large datasets, the robustness and completeness of the registries, and the fact that all inpatient and specialized outpatient visits associated with fractures occurring in Sweden between 2006 and 2012 were included, improving the reliability of the fracture counts. This study also compared data with a large random sample from the total population. Furthermore, unlike previous studies [12, 16], adjustments were made in this study for several potentially confounding clinical and demographic variables, though not all potential risk variables were universally available (e.g. BMD, body mass index). In this study, the statistical adjustments had a clear impact on the results, suggesting that
ro of
adjusting for covariates is important for these types of analyses and should be included in future research.
This study also has limitations. Despite a large initial study sample, fracture data for
-p
women with three prior fractures were only reportable up to three years due to limited follow-up. In addition, fracture data were based on diagnostic coding and
re
were not adjudicated. To avoid counting the same incident fracture more than once, fractures at the same site in the body were counted only if they occurred at least six
lP
months (inpatient) or 12 months (outpatient) after the index fracture. This may have resulted in either underestimation or overestimation of fracture incidence but is
na
difficult to determine with the available data. In addition, asymptomatic vertebral fractures are often undiagnosed and therefore may have been undercounted. This
ur
may have led to either an overestimate or underestimate of the hazard ratios, depending on the population (fractured patients versus controls) that sustained an
Jo
undiagnosed fracture. Lastly, high-trauma and pathological fractures were not specifically excluded, though all included fractures were associated with osteoporosis, so this is unlikely to have had a large impact on the results. 4.3. Conclusions The results from this study indicate that women with a recent fracture are at increased risk of MOF as compared to age- and sex-matched women without history of fracture. The relative risk increased with each subsequent fracture and was highest in the two years following the fracture; for younger women, it remained high within 2–5 years. These results support treatment initiation early after a fracture and beyond, suggesting that treatment timing should be accounted for in health economic assessments of fracture prevention. 20
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) ACKNOWLEDGEMENTS The authors thank the patients, the investigators and their teams who took part in this study. The authors also acknowledge Helen Chambers, PhD, Costello Medical, UK for publication coordination and Kristian Clausen, MPH, and Simon Foulcer, PhD from Costello Medical, UK, for medical writing and editorial assistance based on the authors’ input and direction. This study was funded by UCB Pharma and Amgen Inc. FUNDING This study was sponsored by UCB Pharma and Amgen Inc. Support for third-party
ro of
writing assistance for this article, provided by Kristian Clausen, MPH, Costello Medical, UK was funded by UCB Pharma and Amgen Inc. in accordance with Good Publication
AUTHORS’ CONTRIBUTIONS
-p
Practice (GPP3) guidelines (http://www.ismpp.org/gpp3).
re
Substantial contributions to study conception and design: ES, OS, AS, KEA, FB, JB, ET, MC, CL; substantial contributions to analysis and interpretation of the data: ES,
lP
OS, AS, KEA, FB, JB, ET, MC, CL; drafting the article or revising it critically for important intellectual content: ES, OS, AS, KEA, FB, JB, ET, MC, CL; final approval of
ur
DISCLOSURES
na
the version of the article to be published: ES, OS, AS, KEA, FB, JB, ET, MC, CL.
ES, OS, FB, JB: employed by Quantify Research and funded by UCB Pharma and
Jo
Amgen Inc. to conduct this study; AS: received lecture fees from Amgen Inc., Eli Lilly and Mylan; KA: received lecture fees from Amgen Inc., Eli Lilly, Merck and UCB Pharma; ET, CL, MT: employed by and stockholders of UCB Pharma.
21
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) REFERENCES [1] E. Hernlund, A. Svedbom, M. Ivergard, J. Compston, C. Cooper, J. Stenmark, E.V. McCloskey, B. Jonsson, J.A. Kanis, Osteoporosis in the European Union: medical management, epidemiology and economic burden. A report prepared in collaboration with the International Osteoporosis Foundation (IOF) and the European Federation of Pharmaceutical Industry Associations (EFPIA), Archives of osteoporosis 8 (2013) 136. [2] N.C. Wright, A.C. Looker, K.G. Saag, J.R. Curtis, E.S. Delzell, S. Randall, B. Dawson-Hughes, The recent prevalence of osteoporosis and low bone mass in the United States based on bone mineral density at the femoral neck or lumbar spine, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 29(11) (2014) 2520-2526.
ro of
[3] A. Svedbom, E. Hernlund, M. Ivergard, J. Compston, C. Cooper, J. Stenmark, E.V. McCloskey, B. Jonsson, J.A. Kanis, Osteoporosis in the European Union: a compendium of country-specific reports, Archives of osteoporosis 8 (2013) 137.
re
-p
[4] C. Kerr, C. Bottomley, S. Shingler, L. Giangregorio, H.M. de Freitas, C. Patel, S. Randall, D.T. Gold, The importance of physical function to people with osteoporosis, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 28(5) (2017) 1597-1607.
na
lP
[5] S.L. Silverman, M.E. Minshall, W. Shen, K.D. Harper, S. Xie, The relationship of health-related quality of life to prevalent and incident vertebral fractures in postmenopausal women with osteoporosis: results from the Multiple Outcomes of Raloxifene Evaluation Study, Arthritis and rheumatism 44(11) (2001) 2611-9.
ur
[6] E. Seeman, P.D. Delmas, Bone quality--the material and structural basis of bone strength and fragility, The New England journal of medicine 354(21) (2006) 2250-61.
Jo
[7] J.A. Kanis, F. Borgstrom, C. De Laet, H. Johansson, O. Johnell, B. Jonsson, A. Oden, N. Zethraeus, B. Pfleger, N. Khaltaev, Assessment of fracture risk, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 16(6) (2005) 581-9. [8] J.R. Center, D. Bliuc, T.V. Nguyen, J.A. Eisman, Risk of subsequent fracture after low-trauma fracture in men and women, JAMA 297(4) (2007) 387-94. [9] F. Cosman, S.J. de Beur, M.S. LeBoff, E.M. Lewiecki, B. Tanner, S. Randall, R. Lindsay, Clinician's guide to prevention and treatment of osteoporosis, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 25(10) (2014) 2359-81. [10] J.A. Kanis, Diagnosis of osteoporosis and assessment of fracture risk, Lancet (London, England) 359(9321) (2002) 1929-36. [11] J.A. Eisman, E.R. Bogoch, R. Dell, J.T. Harrington, R.E. McKinney, Jr., A. McLellan, P.J. Mitchell, S. Silverman, R. Singleton, E. Siris, Making the first fracture 22
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) the last fracture: ASBMR task force report on secondary fracture prevention, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 27(10) (2012) 2039-46. [12] H. Johansson, K. Siggeirsdottir, N.C. Harvey, A. Oden, V. Gudnason, E. McCloskey, G. Sigurdsson, J.A. Kanis, Imminent risk of fracture after fracture, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 28(3) (2017) 775-780. [13] J.A. Kanis, O. Johnell, C. De Laet, H. Johansson, A. Oden, P. Delmas, J. Eisman, S. Fujiwara, P. Garnero, H. Kroger, E.V. McCloskey, D. Mellstrom, L.J. Melton, H. Pols, J. Reeve, A. Silman, A. Tenenhouse, A meta-analysis of previous fracture and subsequent fracture risk, Bone 35(2) (2004) 375-82.
-p
ro of
[14] L.A. Ahmed, J.R. Center, A. Bjornerem, D. Bluic, R.M. Joakimsen, L. Jorgensen, H.E. Meyer, N.D. Nguyen, T.V. Nguyen, T.K. Omsland, J. Stormer, G.S. Tell, T.A. van Geel, J.A. Eisman, N. Emaus, Progressively increasing fracture risk with advancing age after initial incident fragility fracture: the Tromso study, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 28(10) (2013) 2214-21.
lP
re
[15] C.M. Klotzbuecher, P.D. Ross, P.B. Landsman, T.A. Abbott, 3rd, M. Berger, Patients with prior fractures have an increased risk of future fractures: a summary of the literature and statistical synthesis, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 15(4) (2000) 721-39.
na
[16] T.A. van Geel, S. van Helden, P.P. Geusens, B. Winkens, G.J. Dinant, Clinical subsequent fractures cluster in time after first fractures, Annals of the rheumatic diseases 68(1) (2009) 99-102.
Jo
ur
[17] A. Balasubramanian, J. Zhang, L. Chen, D. Wenkert, S.G. Daigle, A. Grauer, J.R. Curtis, Risk of subsequent fracture after prior fracture among older women, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 30(1) (2019) 79-92. [18] S. van Helden, J. Cals, F. Kessels, P. Brink, G.J. Dinant, P. Geusens, Risk of new clinical fractures within 2 years following a fracture, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 17(3) (2006) 348-54. [19] T.A. Laurs-van Geel, J.R. Center, P.P. Geusens, G.J. Dinant, J.A. Eisman, Clinical fractures cluster in time after initial fracture, Maturitas 67(4) (2010) 339-42. [20] M. Bonafede, N. Shi, R. Barron, X. Li, D.B. Crittenden, D. Chandler, Predicting imminent risk for fracture in patients aged 50 or older with osteoporosis using US claims data, Archives of osteoporosis 11(1) (2016) 26. [21] C. Roux, K. Briot, Imminent fracture risk, Osteoporosis international : a journal established as result of cooperation between the European Foundation for 23
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Osteoporosis and the National Osteoporosis Foundation of the USA 28(6) (2017) 1765-1769. [22] J.A. Kanis, C. Cooper, R. Rizzoli, B. Abrahamsen, N.M. Al-Daghri, M.L. Brandi, J. Cannata-Andia, B. Cortet, H.P. Dimai, S. Ferrari, P. Hadji, N.C. Harvey, M. Kraenzlin, A. Kurth, E. McCloskey, S. Minisola, T. Thomas, J.Y. Reginster, Identification and management of patients at increased risk of osteoporotic fracture: outcomes of an ESCEO expert consensus meeting, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 28(7) (2017) 2023-2034.
ro of
[23] P. Royston, M.K. Parmar, Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects, Statistics in medicine 21(15) (2002) 2175-97.
re
-p
[24] E.S. Leib, K.G. Saag, J.D. Adachi, P.P. Geusens, N. Binkley, E.V. McCloskey, D.B. Hans, Official Positions for FRAX((R)) clinical regarding glucocorticoids: the impact of the use of glucocorticoids on the estimate by FRAX((R)) of the 10 year risk of fracture from Joint Official Positions Development Conference of the International Society for Clinical Densitometry and International Osteoporosis Foundation on FRAX((R)), Journal of clinical densitometry : the official journal of the International Society for Clinical Densitometry 14(3) (2011) 212-9.
lP
[25] M.E. Charlson, P. Pompei, K.L. Ales, C.R. MacKenzie, A new method of classifying prognostic comorbidity in longitudinal studies: development and validation, Journal of chronic diseases 40(5) (1987) 373-83.
ur
na
[26] O. Johnell, J.A. Kanis, A. Oden, I. Sernbo, I. Redlund-Johnell, C. Petterson, C. De Laet, B. Jonsson, Fracture risk following an osteoporotic fracture, Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 15(3) (2004) 175-9.
Jo
[27] S. Gehlbach, K.G. Saag, J.D. Adachi, F.H. Hooven, J. Flahive, S. Boonen, R.D. Chapurlat, J.E. Compston, C. Cooper, A. Diez-Perez, S.L. Greenspan, A.Z. LaCroix, J.C. Netelenbos, J. Pfeilschifter, M. Rossini, C. Roux, P.N. Sambrook, S. Silverman, E.S. Siris, N.B. Watts, R. Lindsay, Previous fractures at multiple sites increase the risk for subsequent fractures: the Global Longitudinal Study of Osteoporosis in Women, Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 27(3) (2012) 645-53.
24
Journal Pre-proof Submission Revisions – Fracture after Fracture (MOF 1,2,3) Highlights
Amongst women with one or two fragility fractures, 21–32% will experience a subsequent major osteoporotic fracture within five years.
After adjusting for covariates, women with ≥1 prior fracture are at increased relative risk of subsequent fracture within 24 months.
Vertebral fractures were associated with 2.5–4.1 times the risk of subsequent major osteoporotic fracture within 24 months.
Younger women with a prior fracture are at greatest relative risk of subsequent major osteoporotic fracture within five years.
ro of
This study highlights the need for treatment within two years of fracture,
ur
na
lP
re
-p
targeting patients at highest risk of subsequent fracture.
Jo
25