Abstract
Background
A history of falls represents frailty in older adult patients. However, the influence of a history of falls on clinical outcomes in older adult non-valvular atrial fibrillation (NVAF) patients has not been fully elucidated. This All Nippon AF In the Elderly (ANAFIE) Registry sub-analysis evaluated the relationship between a history of falls within the year prior to the study’s commencement and 2-year clinical outcomes in a large, older adult (aged ≥ 75 years) Japanese population with NVAF.
Methods
The ANAFIE Registry (UMIN000024006) was a multicenter, prospective, observational study with a 24-month follow-up period. Of the 32,275 enrolled patients, 28,664 with available fall history data were divided into two groups: those with and without a history of falls within the previous year (n = 2,347 and 26,317, respectively).
Results
Anticoagulant therapy was administered in 95% and 94% of these patients, respectively. Compared with patients without a history of falls, those with a fall history had higher incidences of stroke/systemic embolic events (2.59 vs. 1.52 per 100 person-years), major bleeding (2.20 vs. 0.96), intracranial hemorrhage (ICH) (1.71 vs. 0.66), cardiovascular death (2.12 vs. 0.97), all-cause death (6.78 vs. 3.34), and falls/fractures (13.52 vs. 5.44) (all, P < 0.001) during the follow-up period. In multivariable models, these associations remained significant for all outcomes. Notably, these clinical events, particularly major bleeding and ICH, were frequently observed after falls/fractures that occurred during the follow-up period. Compared with warfarin, direct oral anticoagulants were associated with significantly lower risks of new falls/fractures during the follow-up period, irrespective of baseline fall history.
Conclusions
A history of falls within the previous year was associated with worse prognosis in older adult Japanese NVAF patients. Major bleeding (particularly ICH) occurred frequently following new occurrences of falls/fractures. Warfarin was associated with higher risks of subsequent falls/fractures.
Trial registration.
ANAFIE Registry was registered in the UMIN Clinical Trials Registry (identifier: UMIN000024006) on September 12, 2016.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12877-026-07095-5.
Keywords: Fall, Atrial fibrillation, Clinical outcomes, Oral anticoagulants
Introduction
Older adult atrial fibrillation (AF) patients requiring anticoagulant therapy face a dilemma: a high risk of thromboembolism on one hand and an elevated risk of bleeding on the other. Advanced age is the most prominent risk factor for both thromboembolism and bleeding [1]. Furthermore, frailty is widely recognized as a risk factor for thromboembolism, bleeding, and mortality [2, 3]. In the warfarin era, frail patients were often reluctant to undergo anticoagulant therapy [4]. Even after the development of direct oral anticoagulants (DOACs), conflicting evidence has been reported regarding anticoagulation therapy in frail AF patients [5]. A meta-analysis focusing on patients aged ≥ 80 or ≥ 90 years found that oral anticoagulants were not significantly associated with either thromboembolism or major bleeding; however, there was substantial heterogeneity across studies [6].
The All Nippon AF In the Elderly (ANAFIE) Registry was established to investigate anticoagulant therapy in older adult patients (aged ≥ 75 years) with non-valvular AF (NVAF) in Japan [7, 8]. Several sub-analyses on frailty have been conducted using this registry. Frailty, as defined by the Kihon Checklist, was significantly associated with increased risk of thromboembolism, major bleeding, and mortality. However, these associations did not remain significant for thromboembolism or major bleeding after multivariable adjustment [2]. In contrast, a history of falls within the past year was significantly associated with increased risk of thromboembolism, major bleeding, and mortality, even after multivariable adjustment [8]. This is consistent with other studies showing a higher risk of death and major bleeding in NVAF patients with a history of falls than in those without [9–11]. A systematic review identified a prior history of falls as the most consistent and strongest risk factor for falls in the older adults [12]. Falls can lead to subsequent falls, which may ultimately result in serious clinical outcomes such as major bleeding or death. These findings suggest that a history of falls may have a stronger influence on anticoagulation strategies for stroke prevention in NVAF patients than other frailty-related factors.
While the safety of warfarin in patients at risk of falls has been extensively discussed [13], data comparing DOACs and warfarin in anticoagulated patients with a history of falls remain limited. A report from the Loire Valley Atrial Fibrillation Project included only 76 patients with a prior history of falls [10]. Two sub-analyses of pivotal clinical trials have also been published: one from the ARISTOTLE trial included 753 NVAF patients with a fall history within the previous year [9], and the other from the ENGAGE-AF trial included 900 NVAF patients with fall risk [11].
Critical knowledge gaps remain regarding the impact of falls on the management of older adult patients with NVAF. First, it remains unclear how the outcomes of patients with a history of falls within the past year compare with outcomes in patients without such a history. Second, the impact of new falls/fractures during follow-up on patient outcomes is currently unknown. Third, it is unclear whether DOAC therapy is more appropriate than warfarin for the prevention of subsequent falls/fractures. To address these knowledge gaps, we conducted a subgroup analysis of the ANAFIE Registry, focusing on the clinical implications of fall history in older adult NVAF patients.
Methods
Ethical issues
The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and the research protocol was approved by local ethics committees [7, 8, 14]. Information regarding the participating institute were shown in supplementary appendix [14]. Written informed consent was obtained from all patients.
Study patient population and data collection
Patient enrollment in the ANAFIE Registry, a multicenter, prospective, observational study, started in October 2016. Details of the study design have been published previously [7]. The study was registered in the UMIN Clinical Trials Registry (identifier: UMIN000024006). The target population included over 30,000 men and women aged 75 years or older with a definitive diagnosis of NVAF confirmed by electrocardiogram who regularly attended scheduled hospital visits. The study plan included following each patient for two years. Exclusion criteria have been described previously [8]. For the present sub-analysis, patients without information on falls within one year prior to enrollment were excluded. Patients were divided into two groups: those with a history of falls within the previous year (n = 2,347) and those without (n = 26,317). Baseline characteristics were compared between the two groups.
Clinical outcomes
The clinical outcomes for this sub-analysis included stroke/systemic embolic events (SEE), ischemic stroke, major bleeding, intracranial hemorrhage (ICH), gastrointestinal bleeding, heart failure hospitalization, cardiovascular death, all-cause death, falls/fractures, and the net clinical outcome (a composite of stroke/SEE, major bleeding, and all-cause death) during the 2-year follow-up period. Falls/fractures were identified from medical records and also from patient or family self-reports.
Statistical analysis
Continuous variables are presented as mean ± standard deviation and categorical variables as numbers (percentages). Group differences were tested using the chi-square test for categorical variables and the unpaired t-test for continuous variables.
Kaplan–Meier curves were used to show the incidence of clinical outcomes during the follow-up period. Group differences were assessed using the log-rank test. Incidences are reported as events per 100 person-years with 95% confidence intervals (CI). Univariable and multivariable Cox regression analyses were used to estimate crude and adjusted hazard ratios (HRs) with corresponding 95% CIs for the effect of a history of falls (with vs. without) on clinical outcomes. Multivariable models were adjusted for relevant baseline characteristics and other frailty-related covariables, including a history of falls within 1 year, age (years), sex, body mass index, hypertension, diabetes mellitus, dyslipidemia, hyperuricemia, chronic kidney disease, severe hepatic disease, myocardial infarction, heart failure, reduced left ventricular ejection fraction, cerebrovascular disease, respiratory disease, digestive diseases, active cancer, dementia, type of AF, history of major bleeding, history of catheter ablation, pacemaker implantation, creatinine clearance, and the use of antiarrhythmic medications, antiplatelet agents, proton pump inhibitors, P-glycoprotein inhibitors, and anticoagulant therapy. Missing data were not included in the multivariable adjustment.
To evaluate the impact of the new occurrence of falls/fractures during follow-up, we analyzed the incidence of adverse events after the occurrence of falls/fractures. In patients who experienced new falls/fractures, the incidence of adverse events was assessed during the period following the event, using an event-based landmark approach. In patients who did not experience falls/fractures during follow-up, the mean time to new falls/fractures in the cohort was used as a fixed landmark, set at day 342, and the incidence of adverse events was evaluated during the subsequent period.
Additionally, to assess the impact of DOACs or no oral anticoagulant (OAC), compared with warfarin, on falls/fractures in the overall study population (n = 28,664) and in subgroups with and without a history of falls (n = 2,347 and 26,317, respectively), the incidence of falls/fractures was illustrated and group differences were evaluated using the same methods described above. Multivariable models were adjusted for the same covariables as described above, except for a history of falls within 1 year and anticoagulant therapy.
All statistical analyses were performed using SPSS version 30.0 (IBM Corp, Armonk, NY, USA) for Windows (Microsoft Corp., Redmond, WA, USA). A two-sided P-value < 0.05 was considered statistically significant.
Results
Patient characteristics
Figure 1 shows the patient population. Among the 32,275 patients enrolled in the ANAFIE registry, 28,664 patients had available information on falls within the year prior to enrollment. Of these, 8.2% (n = 2,347) had a history of falls within the past year at baseline, while the remaining 91.8% (n = 26,317) did not.
Fig. 1.
Patient population
Table 1 summarizes the characteristics of patients with and without a history of falls. Compared with patients without a history of falls, those with a history were older and more frequently female, and had lower body mass index and reduced creatinine clearance. The prevalence of OAC use was slightly higher among patients with a history of falls (94.6%) than among those without (93.6%). Warfarin use was more common in patients with a history of falls (28.1%) than in those without (25.3%), whereas DOAC use was slightly lower in patients with a history of falls (64.7%) than in those without (67.0%).
Table 1.
The baseline patient characteristics and medications
| Total (N = 28,664) |
With Falls (n = 2347) |
Without falls (n = 26,317) |
P-value | |
|---|---|---|---|---|
| Male | 16,410 (57%) | 1149 (49%) | 15,261 (58%) | < 0.001 |
| Age (year) | 81 ± 5 | 83 ± 5 | 81 ± 5 | < 0.001 |
| Age ≥ 85 (years) | 7485 (26%) | 907 (39%) | 6578 (25%) | < 0.001 |
| Height (cm) | 157 ± 9 | 155 ± 10 | 157 ± 9 | < 0.001 |
| Body weight (kg) | 58 ± 11 | 55 ± 11 | 58 ± 11 | < 0.001 |
| Body mass index (kg/m2) | 23.3 ± 3.6 | 22.9 ± 3.8 | 23.4 ± 3.5 | < 0.001 |
| SBP (mmHg) | 127 ± 17 | 126 ± 18 | 128 ± 17 | < 0.001 |
| DBP (mmHg) | 71 ± 12 | 69 ± 12 | 71 ± 12 | < 0.001 |
| Creatinine clearance (mL/min) | 49 ± 18 | 43 ± 18 | 49 ± 17 | < 0.001 |
| CHADS2 score | 2.8 ± 1.2 | 3.1 ± 1.3 | 2.8 ± 1.2 | < 0.001 |
| 1 | 2699 (9%) | 167 (7%) | 2532 (10%) | |
| 2 | 10,175 (36%) | 647 (28%) | 9528 (36%) | |
| 3 | 8407 (29%) | 693 (30%) | 7714 (29%) | |
| ≥ 4 | 7383 (26%) | 840 (36%) | 6543 (25%) | |
| CHA2DS2 VASc score | 4.4 ± 1.4 | 4.9 ± 1.4 | 4.4 ± 1.4 | < 0.001 |
| 2 | 1548 (5%) | 82 (4%) | 1466 (6%) | |
| 3 | 6025 (21%) | 321(14%) | 5704 (22%) | |
| 4 | 8654 (30%) | 612 (26%) | 8042 (31%) | |
| 5 | 6491 (23%) | 598 (26%) | 5893 (22%) | |
| ≥ 6 | 5946 (21%) | 734 (31%) | 5212 (20%) | |
| HAS-BLED score | 1.9 ± 0.9 | 2.0 ± 0.9 | 1.8 ± 0.9 | < 0.001 |
| 1 | 11,320 (40%) | 741 (32%) | 10,579 (40%) | |
| 2 | 11,266 (39%) | 941 (40%) | 10,325 (39%) | |
| ≥ 3 | 6078 (21%) | 665 (28%) | 5413 (21%) | |
| History of major bleeding | 1262 (0%) | 162 (7%) | 1100 (4%) | < 0.001 |
| AF type | 0.01 | |||
| Paroxysmal | 11,981 (42%) | 955 (41%) | 11,026 (42%) | |
| Persistent | 4708 (16%) | 350 (15%) | 4358 (17%) | |
| Long-standing persistent/permanent | 11,975 (42%) | 1042 (44%) | 10,933 (42%) | |
| History of non-pharmacological AF therapy | 4907 (17%) | 404 (17%) | 4503 (17%) | 0.90 |
| Catheter ablation | 2549 (9%) | 145 (6%) | 2404 (9%) | < 0.001 |
| Electrical defibrillation | 616 (2%) | 66 (3%) | 550 (2%) | 0.02 |
| ICD | 133 (0%) | 9 (0%) | 124 (0%) | 0.55 |
| Pacemaker | 2052 (7%) | 240 (10%) | 1812 (7%) | < 0.001 |
| Others | 90 (0%) | 11 (0%) | 79 (0%) | 0.16 |
| Comorbidities | ||||
| Hypertension | 21,534 (75%) | 1772 (76%) | 19,762 (75%) | 0.66 |
| Diabetes mellitus | 7616 (27%) | 685 (29%) | 6931 (26%) | 0.003 |
| Chronic kidney disease | 6010 (21%) | 611 (26%) | 5399 (21%) | < 0.001 |
| Myocardial infarction | 1617 (6%) | 149 (6%) | 1468 (6%) | 0.12 |
| Heart failure | 10,564 (37%) | 1032 (44%) | 9532 (36%) | < 0.001 |
| History of cerebrovascular disease | 6455 (23%) | 786 (33%) | 5669 (22%) | < 0.001 |
| Gastrointestinal disease | 8292 (29%) | 775 (33%) | 7517 (29%) | < 0.001 |
| Active cancer | 3041 (11%) | 277 (12%) | 2764 (11%) | 0.05 |
| Dementia | 2245 (8%) | 374 (16%) | 1871 (7%) | < 0.001 |
| Anticoagulant therapy | 26,496 (94%) | 2180 (95%) | 24,316 (94%) | 0.05 |
| Warfarin | 7324 (26%) | 660 (28%) | 6664 (25%) | 0.003 |
| DOAC | 19,160 (67%) | 1518 (65%) | 17,642 (67%) | 0.02 |
| Dabigatran | 2098 (7%) | 117 (5%) | 1981 (8%) | < 0.001 |
| Rivaroxaban | 5750 (20%) | 404 (17%) | 5346 (20%) | < 0.001 |
| Apixaban | 7094 (24%) | 636 (27%) | 6458 (25%) | 0.006 |
| Edoxaban | 4218 (15%) | 359 (15%) | 3857 (15%) | 0.34 |
| Total number of medications | 6.6 ± 3.1 | 7.5 ± 3.4 | 6.5 ± 3.1 | < 0.001 |
| Antihypertensive medications | 20,350 (71%) | 1669 (71%) | 18,681 (71%) | 0.90 |
| Diabetes medications | 4561 (16%) | 406 (17%) | 4155 (16%) | 0.06 |
| Lipid-lowering drugs | 10,745 (37%) | 869 (37%) | 9876 (38%) | 0.63 |
| Antiarrhythmic medications | 16,317 (57%) | 1300 (55%) | 15,017 (57%) | 0.12 |
| Rhythm control | 5648 (20%) | 433 (18%) | 5215 (20%) | 0.11 |
| Rate control | 11,913 (42%) | 973 (41%) | 10,940 (42%) | 0.92 |
| Antiplatelet | 5028 (18%) | 445 (19%) | 4583 (17%) | 0.06 |
| Proton pump inhibitors | 10,507 (37%) | 986 (42%) | 9521 (36%) | < 0.001 |
| P-glycoprotein inhibitors | 460 (2%) | 48 (2%) | 412 (2%) | 0.08 |
Categorical and consecutive data are presented as numbers (%) and means ± standard deviation, respectively
Abbreviations: SBP Systolic blood pressure, DBP Diastolic blood pressure, AF Atrial fibrillation, ICD Implantable cardioverter-defibrillator, DOAC Direct oral anticoagulants
Impact of history of falls on clinical outcomes
The incidence of clinical events in patients with and without a history of falls are shown in Table 2. Clinical events occurred more frequently in patients with a history of falls than in those without. Kaplan–Meier curves for clinical outcomes are presented in Fig. 2A–J. Log-rank testing showed P < 0.001 for all outcomes except for gastrointestinal bleeding (P = 0.004). In multivariable analysis, a history of falls was independently associated with all clinical outcomes, except for gastrointestinal bleeding, after adjustment for other clinical factors (Table 3). Using the Fine–Gray model, the adjusted subdistribution hazard ratio (HR) for a history of falls with respect to Stroke/SSE was 1.43 (95% CI 1.08–1.90; P = 0.013). For comparison, the corresponding HR derived from the Cox model was 1.37 (95% CI 1.08–1.74, P = 0.012). Accounting for competing risks did not materially change the estimates.
Table 2.
Incidence rate of clinical outcomes
| Clinical outcomes | Total (N = 28,664) |
With falls (n = 2,347) |
Without falls (n = 26,317) |
|---|---|---|---|
| Stroke/SEE |
849 1.60 (1.50–1.71) |
108 2.59 (2.15–3.13) |
741 1.52 (1.41–1.63) |
| Ischemic stroke |
656 1.23 (1.14–1.33) |
79 1.89 (1.52–2.35) |
577 1.18 (1.09–1.28) |
| Major bleeding |
563 1.06 (0.97–1.15) |
92 2.20 (1.79–2.69) |
471 0.96 (0.88–1.05) |
| Intracranial hemorrhage |
399 0.75 (0.68–0.82) |
72 1.71 (1.36–2.16) |
327 0.66 (0.60–0.74) |
| GI bleeding |
984 1.87 (1.75–1.99) |
102 2.46 (2.02–2.98) |
882 1.81 (1.70–1.94) |
| Heart failure hospitalization |
2,117 4.06 (3.89–4.23) |
276 6.80 (6.04–7.65) |
1,841 3.82 (3.65–4.00) |
| Cardiovascular death |
569 1.06 (0.98–1.15) |
90 2.12 (1.72–2.61) |
479 0.97 (0.89–1.06) |
| All-cause death |
1,939 3.61 (3.46–3.78) |
288 6.78 (6.04–7.61) |
1,651 3.34 (3.18–3.51) |
| Falls/fractures |
3,068 6.04 (5.83–6.26) |
509 13.52 (12.40–14.75) |
2,559 5.44 (5.23–5.66) |
| Net clinical outcomesa |
2,852 5.40 (5.21–5.61) |
403 9.76 (8.85–10.76) |
2,449 5.03 (4.84–5.24) |
Data are expressed by number of events and events per 100 person-years with 95% confidence intervals
Abbreviations: SEE Systemic embolic events, GI Gastrointestinal
aNet clinical outcomes is a composite of stroke/SEE, major bleeding, and all-cause death
Fig. 2.
Kaplan–Meier curves for clinical outcomes in patients with and without a history of falls. A, Stroke/SSE; B, Ischemic stroke; C, Major bleeding; D, ICH; E, Gastrointestinal bleeding; F, Heart failure hospitalization; G, Cardiovascular death; H, All-cause death; I, Falls/Fractures; J, Net clinical outcomes. Abbreviations: SEE Systemic embolic events, ICH Intracranial hemorrhage. Net clinical outcomes is a composite of stroke/SEE, major bleeding, and all-cause death. Values (%) shown in the figure represent the probability of event occurrence at the end of the observation period; the absolute number of events and incidence rates per 100 person-years are provided in Table 2
Table 3.
Impact of a history of falls within 1 year on adverse clinical outcomes
| Clinical outcomes | Univariable models | Multivariable models | ||||
|---|---|---|---|---|---|---|
| Crude HR | 95% confidence intervals | P-value | Adjusted HR | 95% confidence intervals | P-value | |
| Stroke/SEE | 1.71 | 1.40–2.09 | < 0.001 | 1.37 | 1.08–1.74 | 0.01 |
| Ischemic stroke | 1.60 | 1.27–2.03 | < 0.001 | 1.39 | 1.06–1.82 | 0.02 |
| Major bleeding | 2.29 | 1.83–2.87 | < 0.001 | 1.89 | 1.46–2.45 | < 0.001 |
| Intracranial hemorrhage | 2.58 | 2.00–3.34 | < 0.001 | 2.17 | 1.60–2.93 | < 0.001 |
| GI bleeding | 1.35 | 1.10–1.66 | 0.004 | 1.20 | 0.95–1.51 | 0.13 |
| Heart failure hospitalization | 1.78 | 1.57–2.02 | < 0.001 | 1.38 | 1.20–1.59 | < 0.001 |
| Cardiovascular death | 2.20 | 1.76–2.76 | < 0.001 | 1.68 | 1.31–2.15 | < 0.001 |
| All-cause death | 2.04 | 1.80–2.31 | < 0.001 | 1.48 | 1.28–1.70 | < 0.001 |
| Falls/Fractures | 2.48 | 2.26–2.73 | < 0.001 | 2.07 | 1.85–2.31 | < 0.001 |
| Net clinical outcomesa | 1.95 | 1.75–2.16 | < 0.001 | 1.48 | 1.31–1.67 | < 0.001 |
In the multivariable model, fall within 1 year, age (year), sex, body mass index, hypertension, diabetes mellitus, dyslipidemia, hyperuricemia, chronic kidney disease, severe hepatic disease, myocardial infarction, heart failure, reduced left ventricular ejection fraction, cerebrovascular disease, respiratory disease, digestive diseases, active cancer, dementia, type of AF, history of major bleeding, history of catheter ablation, pacemaker implantation, creatinine clearance, and use of antiarrhythmic medications, antiplatelet agents, proton pump inhibitors, P-glycoprotein inhibitors and anticoagulant therapy were forcedly introduced
Abbreviations: HR Hazard ratio, SEE Systemic embolic events, GI Gastrointestinal, AF Atrial fibrillation
aNet clinical outcomes is a composite of stroke/SEE, major bleeding, and all-cause death
To further clarify the impact of adjustment for OAC type (DOAC vs warfarin), we conducted a sensitivity analysis restricted to anticoagulated patients. Kaplan–Meier curves for clinical outcomes in this subgroup are shown in Supplementary Fig. 1. Clinical events occurred more frequently in patients with a history of falls than in those without. Log-rank testing showed P < 0.001 for all outcomes except for gastrointestinal bleeding (P = 0.013). In multivariable analysis, a history of falls was independently associated with all clinical outcomes, except for gastrointestinal bleeding, after adjustment for other clinical factors including OAC type (DOAC vs warfarin) (Supplementary Table 1). Thus, the results of the Kaplan–Meier analyses and Cox regression models in this restricted cohort were generally consistent with those observed in the overall population.
Clinical outcomes after new falls/fractures
Supplementary Table 2 summarizes the baseline characteristics of patients with and without a new occurrence of falls/fractures during the follow-up period (n = 3,068 and n = 25,596, respectively). Compared with patients without a new occurrence of falls/fractures, those who experienced a new occurrence were older and had a lower body mass index and reduced creatinine clearance. Warfarin use was more common among patients with a new occurrence of falls/fractures (29.7%) than among those without (25.1%), whereas DOAC use was slightly lower in patients with a new occurrence of falls/fractures (63.6%) than in those without (67.2%).
Clinical outcomes were compared between patients with and without a new occurrence of falls/fractures (Table 4). Clinical events occurred more frequently in patients with a new occurrence of falls/fractures during follow-up than in those without. Kaplan–Meier curves (Fig. 3 A–F) clearly show that major bleeding, primarily ICH, were more frequently observed at the time of, or shortly after, the occurrence of falls/fractures.
Table 4.
Incidence rate of clinical outcomes after new falls/fractures
| Clinical outcomes | With new falls/fractures during follow-up (n = 3,068) |
Without new falls/fractures during follow-up*1 (n = 25,596) |
|---|---|---|
| Stroke/SEE |
97 3.47 (2.85–4.23) |
344 1.43 (1.29–1.59) |
| Ischemic stroke |
74 2.63 (2.10–3.30) |
276 1.15 (1.02–1.29) |
| Major bleeding |
179 6.52 (5.63–7.54) |
170 0.70 (0.60–0.82) |
| Intracranial hemorrhage |
154 5.58 (4.77–6.54) |
114 0.47 (0.39–0.56) |
| All-cause death |
291 10.13 (9.03–11.37) |
889 3.64 (3.41–3.89) |
| Net clinical outcomes*2 |
463 17.21 (15.71–18.85) |
1206 5.05 (4.77–5.35) |
Data are expressed as the number of events and events per 100 person-years with 95% confidence intervals
*1Clinical outcomes in patients without new falls/fractures during follow-up served as a reference, with the observation period set to begin at 342 days, the average time to the occurrence of falls/fractures events
*2Net clinical outcomes is a composite of stroke/SEE, major bleeding, and all-cause death
Abbreviations: SEE Systemic embolic events
Fig. 3.
Kaplan–Meier curves for clinical outcomes after falls/fractures among patients with and without falls/fractures during follow-up. For patients without falls/fractures during follow-up, the observation period was set to begin at 342 days, which was the average time to the occurrence of falls/fractures events. A Stroke/SSE; B Ischemic stroke, C Major bleeding, D ICH; E All-cause death; F, Net clinical outcomes. Abbreviations: SEE Systemic embolic events, ICH Intracranial hemorrhage. Net clinical outcomes is a composite of stroke/SEE, major bleeding, and all-cause death. Values (%) shown in the figure represent the probability of event occurrence at 1 year; the absolute number of events and incidence rates per 100 person-years are provided in Table 4
To address concerns regarding the time-dependent nature of new falls/fractures and the potential for immortal time or unequal follow-up bias, we conducted additional landmark analyses at baseline (0 months), 6 months, and 12 months among patients without new falls/fractures. From each landmark time point, clinical outcomes were re-estimated using Kaplan–Meier analyses (Supplementary Fig. 2–1 A–F. 2–2 A–F, and 2–3 A–F). The results were consistent across all landmark analyses.
To further clarify the impact of incident falls/fractures, we conducted a sensitivity analysis restricted to patients without a history of falls at baseline, thereby evaluating true incident falls/fractures during follow-up. Kaplan–Meier curves for clinical outcomes in this subgroup are shown in Supplementary Fig. 3. In this restricted cohort, patients who experienced a new occurrence of falls/fractures during follow-up also showed more adverse clinical events compared with those who did not experience falls/fractures.
Difference in subsequent falls/fractures among patients receiving warfarin, DOAC, or no OAC therapy
Supplementary Table 3 summarizes the characteristics of patients stratified by anticoagulant strategy (warfarin, DOAC, or no OAC) (n = 7,324, n = 19,160 and n = 1,798, respectively), with standard mean differences (SMDs) calculated for DOAC versus warfarin and no OAC versus warfarin. Compared with patients receiving warfarin, those receiving DOACs had higher creatinine clearance, a higher prevalence of paroxysmal AF, and a lower prevalence of long-standing persistent/permanent AF (absolute SMD > 0.2). Compared with patients receiving warfarin, those receiving no OAC were older, had lower body height and body weight, higher systolic blood pressure, lower CHADS2 scores, a higher prevalence of prior catheter ablation, a lower prevalence of heart failure, higher use of rhythm control drugs, lower use of rate control drugs, and higher use of antiplatelet therapy (absolute SMD > 0.2).
Kaplan–Meier curves for falls/fractures are shown in Fig. 4A–C. Incidences of falls/fractures during the follow-up period in patients receiving warfarin, DOAC, or no OAC therapy are shown in Table 5. Falls/fractures occurred less frequently in patients treated with DOACs or no OAC than in those treated with warfarin. In multivariable analysis, both DOAC and no OAC therapy were associated with significantly lower risks of subsequent falls/fractures than warfarin therapy (DOACs: HR 0.82, 95% CI 0.75–0.90, P < 0.001; no OAC: HR 0.64, 95% CI 0.52–0.79, P < 0.001). This association was observed in patients with and without a history of falls at baseline. Among patients with a history of falls at baseline, the use of DOACs or no OAC was associated with significantly lower risks of subsequent falls/fractures than warfarin use (DOACs: HR 0.71, 95% CI 0.57–0.87, P = 0.001; no OAC: HR 0.43, 95% CI 0.25–0.75, P = 0.003). Among those without a history of falls, the corresponding HRs were 0.87 (95% CI 0.78–0.96, P = 0.008) for DOACs and 0.72 (95% CI 0.58–0.89, P = 0.003) for no OAC (Table 6).
Fig. 4.
Kaplan–Meier curves for subsequent falls/fractures in patients receiving warfarin, DOACs, or no OACs. A Total population; B Patients with falls within the previous year at baseline; C Patients without falls within the previous year at baseline. Abbreviations: DOACs Direct oral anticoagulants, OAC Oral anticoagulant. Values (%) shown in the figure represent the probability of event occurrence at the end of the observation period; the absolute number of events and incidence rates per 100 person-years are provided in Table 5
Table 5.
Incidence rate of new falls/fractures during the follow-up period according to anticoagulation therapy
| Warfarin | DOACs | No OAC | |
|---|---|---|---|
| All |
(n = 7,324) 912 7.15 (6.70–7.63) |
(n = 19,160) 1951 5.70 (5.45–5.96) |
(n = 1,798) 155 4.92 (4.20–5.76) |
| Patients with a fall history |
(n = 660) 177 17.08 (14.74–19.79) |
(n = 1,518) 303 12.29 (10.98–13.76) |
(n = 124) 18 9.16 (5.80–14.49) |
| Patients without a fall history |
(n = 6,664) 735 6.27 (5.83–6.74) |
(n = 17,642) 1648 5.19 (4.94–5.45) |
(n = 1,674) 137 4.64 (3.92–5.48) |
Data are expressed by number of events and events per 100 person-years with 95% confidence intervals
Abbreviations: DOACs Direct oral anticoagulants, OAC Oral anticoagulant
Table 6.
Impact of DOAC and no OAC use versus warfarin on new falls/fractures, stratified by history of falls within 1 year
| DOACs vs. warfarin | No OAC vs. warfarin | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Crude HR (95% CI) |
P-value | Adjusted HR (95% CI) |
P-value | Crude HR (95% CI) |
P-value | Adjusted HR (95% CI) |
P-value | ||
| All population | 0.80 (0.74–0.86) | < 0.001 | 0.82 (0.75–0.90) | < 0.001 | 0.69 (0.58–0.82) | < 0.001 | 0.64 (0.52–0.79) | < 0.001 | |
| Patients with a history of falls within 1 year | 0.72 (0.60–0.87) | < 0.001 | 0.71 (0.57–0.87) | 0.001 | 0.54 (0.33–0.87) | 0.01 | 0.43 (0.25–0.75) | 0.003 | |
| Patients without a history of falls within 1 year | 0.83 (0.76–0.90) | < 0.001 | 0.87 (0.78–0.96) | 0.008 | 0.74 (0.62–0.89) | 0.001 | 0.72 (0.58–0.89) | 0.003 | |
In the multivariable model, the same variables as in Table 3, except for history of falls within 1 year and anticoagulant therapy, were forcedly introduced
Abbreviations: DOACs direct oral anticoagulants, OAC oral anticoagulant, HR hazard ratio
Discussion
Major findings
The main findings of this study are: (1) Approximately 8.2% of patients had a history of falls within the year prior to enrollment. Compared with patients without a history of falls, those with a history were older and had lower body mass index and creatinine clearance, and higher risk scores for stroke and bleeding. The majority of patients in both groups were receiving anticoagulant therapy (93.6%), predominantly DOACs (66.8%). (2) The two-year incidence of all clinical outcomes was significantly higher in patients with a history of falls than in those without. Multivariable analysis found that patients with previous falls had approximately twice the risk of major bleeding and ICH, 1.5 times the risk of stroke/SEE, ischemic stroke, heart failure hospitalization, all-cause death, and net clinical outcomes, and more than twice the risk of subsequent falls/fractures than patients without a history of falls. (3) All adverse events were more frequent observed after the occurrence of falls/fractures during the follow-up period, and major bleeding (particularly ICH) was more frequently observed at the time of, or shortly after, the occurrence of falls/fractures. (4) DOAC and no OAC therapies were associated with significantly lower risks of subsequent falls/fractures during the follow-up period than warfarin therapy, regardless of baseline fall history.
History of falls and clinical outcomes
Approximately 30% of the general population aged 65 years and older experiences at least one fall per year [15]. However, the frequency of falls varies significantly depending on whether older individuals live in the community or reside in care facilities. In a single-center study from Japan (Shinken Database), the one-year incidence of falls/fractures was reported as 2.7% and 5.0% among patients aged 75 years and older with paroxysmal AF and persistent AF, respectively [16]. A study using a health insurance database involving DOAC users (presumed to have NVAF) reported that 7.4% had experienced falls within the past year (mean age: 80.9 years for those with a fall history vs. 75.4 years for those without) [17]. In another report that followed older adult NVAF patients for one year, patients classified as high fall-risk (defined as those who had experienced a fall within the previous year or who were aware of problems with walking, balance, or mobility) had an annual fall incidence of 8.0%, compared with 6.8% in the low-risk group (P = 0.64) [18]. This was largely consistent with the incidence observed in our study (6.0%).
A history of falls within the previous year, which we identified in 8.2% of NVAF patients in our study population, is relatively common in real-world clinical settings and serves as a useful and easily assessable predictor of future falls, as shown in previous studies [9, 10, 17]. In our study, the annual incidence of new falls during follow-up was 13.5% among patients with a history of falls and only 5.4% in those without, corresponding to nearly twice the risk of new falls after multivariable adjustment. Falls are strongly associated with head trauma and ICH. The EPICA study, which focused on patients aged ≥ 80 years, identified a history of bleeding and a history of falls as predictors of subsequent major bleeding events [19]. In our cohort, a history of falls was a general prognostic factor for adverse outcomes in NVAF patients, and was associated with increased risk of all-cause death, major bleeding, and ICH, which is consistent with previous studies [9–11, 20].
The association between a history of falls and the risk of stroke/SEE remains unclear. The ARISTOTLE study reported no significant relationship between a history of falls and stroke/SEE [9]. In contrast, the Loire Valley Atrial Fibrillation Project identified a history of falls as an independent predictor of thromboembolism [10]. However, a direct causal relationship between falls and subsequent stroke/SEE appears implausible. In our study, a history of falls at baseline was significantly associated with subsequent falls/fractures. In clinical practice, anticoagulant therapy is often discontinued and not resumed in older adult patients who experience falls/fractures, which may lead to subsequent ischemic stroke. A study from the United States reported that among NVAF patients who sustained injuries while on anticoagulant therapy, the resumption rate of anticoagulants was 41% following traumatic brain injury, indicating that anticoagulants were not resumed in nearly half of these cases [21]. Therefore, the discontinuation of OAC therapy after falls/fractures may partly explain the observed association between falls and stroke/SEE. Patients who experienced falls/fractures, particularly those accompanied by sustained injuries, require close clinical monitoring. Additionally, modifiable bleeding risk factors should be addressed[22] to facilitate the resumption of anticoagulation therapy whenever possible. The rate of resuming therapy may be relatively higher among patients enrolled in clinical trials such as ARISTOTLE [9].
Impact of DOAC, no OAC therapy, or warfarin on falls/fractures
Surprisingly, in our population, the use of DOACs or no OAC therapy was associated with significantly lower risks of new occurrence of falls/fractures during the follow-up period than warfarin use, regardless of baseline fall history. Lower falls/fractures with DOACs may partly reflect confounding by indication, despite possible biological mechanisms. Although the mechanisms are unclear, these findings may be partially explained by differences in the effects of these therapies on bone and muscle metabolism.
Vitamin K plays a crucial role in bone and muscle metabolism [23]. Warfarin, a vitamin K antagonist, inhibits the carboxylation of vitamin K-dependent bone mineralization proteins, including osteocalcin, matrix Gla protein, and periostin, thereby increasing the risk of fractures [24–26]. In contrast, DOACs have not been reported to interfere with bone metabolism-related proteins [27].
Supporting these mechanistic insights, recent systematic reviews and meta-analyses have demonstrated that DOACs users have a lower risk of fractures than warfarin users [28, 29]. Furthermore, several studies in AF populations have reported an increased risk of osteoporotic fragility fractures in patients treated with warfarin [30, 31]. A Medicare population-based study raised particular concern, showing that AF patients who used warfarin for more than one year had a significantly higher risk of fragility fractures than those who did not receive warfarin therapy [30]. This finding is consistent with our results.
With regard to muscle strength, observational studies have shown that better physical performance is associated with higher plasma concentrations of vitamin K [32, 33]. In particular, individuals with plasma vitamin K1 concentrations ≥ 1 nmol/L demonstrated superior lower-extremity physical performance both cross-sectionally and over a 5-year follow-up period [32]. Although a causal relationship has not been confirmed in interventional studies, inhibition of vitamin K activity may adversely affect muscle mass and function, potentially leading to falls.
Several studies have shown that warfarin-treated patients commonly reduce their intake of vitamin K–rich vegetables.[34] At the same time, observational data indicate that insufficient vegetable consumption is associated with poorer physical function and an increased risk of falls or injurious falls in older adults [35]. Taken together, these findings raise the possibility that dietary restriction practices among warfarin-treated patients—particularly reduced consumption of vitamin K–rich vegetables—may inadvertently increase vulnerability to falls, potentially through effects on musculoskeletal function, nutritional status, or frailty. Further research is warranted to clarify this potential pathway.
In our population, falls/fractures was the most frequently observed clinical event. Notably, major bleeding events (particularly ICH) were more frequently observed following the occurrence of falls/fractures. Given that both stroke/SEE and major bleeding (especially ICH) frequently occurred shortly after the new occurrence of falls/fractures, preventing falls/fractures may help reduce these subsequent adverse outcomes in older adult patients with NVAF.
In this context, DOACs may be considered a reasonable option, as they were not associated with an increased incidence of falls/fractures in our cohort; however, causal conclusions cannot be drawn from these observational data. These findings highlight the favorable safety profile of DOACs compared with warfarin observed in this cohort, in NVAF patients with a history of falls in real-world clinical settings. Given that DOACs are also associated with a lower incidence of ICH than warfarin [36], DOACs may be considered a reasonable option in patients for whom both agents are clinically appropriate.
Limitations
Our study has several limitations. First, it was a post hoc evaluation. Second, all falls may not have been reported to attending physicians [37], and therefore, our data may have underestimated the prevalence of fall history at baseline and new occurrence of falls during follow-up. Third, although we adjusted for potential confounders known to be associated with clinical events in patients with NVAF, residual confounding cannot be ruled out. In addition, detailed information on frailty and activities of daily living was not available in the present analysis. Therefore, fall history may have served, at least in part, as a surrogate marker for unmeasured frailty, and residual confounding related to frailty cannot be fully excluded. Fourth, the study population was primarily hospital-based and may not be representative of all patients with NVAF, such as those cared for by home-visiting physicians or those residing in nursing homes. Additional limitations of the ANAFIE Registry have been reported previously [8], including enrollment of only Japanese NVAF patients, the lack of restrictions on including patients already receiving anticoagulant therapy at enrollment, and the exclusion of certain variables, such as changes in OAC agents during follow-up. The registry population consisted predominantly of older Japanese patients, and therefore the results may not be directly generalized to younger or Western populations. The incidence of major bleeding in the main ANAFIE Registry was lower than expected, which may have influenced the results of this analysis.
Conclusions
A history of falls within the previous year was identified as an independent risk factor for adverse clinical outcomes in patients with NVAF. DOACs were associated with significantly lower risk of new occurrence of falls/fractures during the follow-up period than warfarin therapy, regardless of baseline fall history. Among patients who experienced a new occurrence of falls/fractures, major bleeding, particularly ICH, was frequently observed following the event.
Supplementary Material
Acknowledgements
The authors thank ANAFIE Publication Committee (Hirotsugu Atarashi, MD, PhD; Ken Okumura, MD, PhD; Takenori Yamaguchi, MD, PhD) and Satoshi Teramukai, PhD for supporting statistical analysis. We also thank Leah Cannon, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.
Abbreviations
- AF
Atrial fibrillation
- DOAC
Direct oral anticoagulant
- ANAFIE
All Nippon AF In the Elderly
- NVAF
Non-valvular atrial fibrillation
- SEE
Systemic embolic events
- ICH
Intracranial hemorrhage
- CI
Confidence intervals
- OAC
Oral anticoagulant
- HR
Hazard ratio
- SMD
Standard mean difference
Authors’ contributions
Author contributions of the present study are as follows: Conception and design of the study; Dr Arita, Dr Suzuki, Dr Yamashita. Analysis and interpretation of data; Dr Arita, Dr Suzuki, Dr Hirota, Dr Yagi, Dr Yamashita. Collection and assembly of data; Dr Arita, Dr Suzuki, Dr Hirota, Dr Yagi, Dr Otsuka, Mr. Kobayashi. Drafting of the article; Dr Arita, Dr Suzuki, Dr Yamashita, Dr Inoue. Critical revision of the article for important intellectual content; Dr Suzuki, Dr Yamashita, Dr Inoue.
Funding
This research was supported by Daiichi Sankyo Co, Ltd. The sponsor contributed to the study design, data analysis planning, data interpretation, and the decision to submit the manuscript for publication. However, the sponsor was not directly involved in data management, direct data access, or statistical analysis.
Data availability
The individual deidentified participant data and study protocol will be shared for up to 36 months after publication of the article. Access criteria for data sharing (including requests) will be decided on by a committee led by Daiichi-Sankyo. To gain access, those requesting data access will need to sign a data access agreement. Requests should be directed to yamt-tky@umin.ac.jp.
Declarations
Ethics approval and consent to participate
The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki and the research protocol was approved by local ethics committees [7, 8, 14]. Information regarding the participating institute was shown in supplementary appendix [14]. Written informed consent was obtained from all patients. The study was registered in the UMIN Clinical Trials Registry (identifier: UMIN000024006).
Consent for publication
Not applicable.
Competing interests
Dr Suzuki has received remuneration from Bristol-Myers Squibb and Daiichi Sankyo. Dr Yamashita has received research funding from Daiichi Sankyo Co, Ltd, manuscript fees from Daiichi Sankyo Co, Ltd and Bristol Myers Squibb K.K., and remuneration from Daiichi Sankyo Co, Ltd, Bayer Yakuhin Ltd, Bristol Myers Squibb K.K., Nippon Boehringer Ingelheim Co, Ltd, Novartis, and Ohtsuka Pharmaceutical, Co, Ltd. Dr Inoue has received consultancy fees from Daiichi Sankyo.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Takuto Arita, Email: arita@cvi.or.jp.
ANAFIE investigators:
References
- 1.Granziera S, Cohen AT, Nante G, Manzato E, Sergi G. Thromboembolic prevention in frail elderly patients with atrial fibrillation: a practical algorithm. J Am Med Dir Assoc. 2015;16(5):358–64. [DOI] [PubMed] [Google Scholar]
- 2.Akishita M, Suzuki S, Inoue H, Akao M, Atarashi H, Ikeda T, et al. Frailty and outcomes in older adults with non-valvular atrial fibrillation from the ANAFIE registry. Arch Gerontol Geriatr. 2022;101:104661. [DOI] [PubMed] [Google Scholar]
- 3.Wilkinson C, Wu J, Searle SD, Todd O, Hall M, Kunadian V, et al. Clinical outcomes in patients with atrial fibrillation and frailty: insights from the ENGAGE AF-TIMI 48 trial. BMC Med. 2020;18(1):401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Lefebvre MC, St-Onge M, Glazer-Cavanagh M, Bell L, Kha Nguyen JN, Viet-Quoc Nguyen P, et al. The effect of bleeding risk and frailty status on anticoagulation patterns in octogenarians with atrial fibrillation: the FRAIL-AF study. Can J Cardiol. 2016;32(2):169–76. [DOI] [PubMed] [Google Scholar]
- 5.Wilkinson C, Todd O, Clegg A, Gale CP, Hall M. Management of atrial fibrillation for older people with frailty: a systematic review and meta-analysis. Age Ageing. 2019;48(2):196–203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Barssoum K, Kumar A, Thakkar S, Sheth AR, Kharsa A, Ibrahim M, et al. Meta-analysis of safety and efficacy of anticoagulation versus no anticoagulation in octogenarians and nonagenarians with atrial fibrillation. High Blood Press Cardiovasc Prev. 2021;28(3):271–82. [DOI] [PubMed] [Google Scholar]
- 7.Inoue H, Yamashita T, Akao M, Atarashi H, Ikeda T, Okumura K, et al. Prospective observational study in elderly patients with non-valvular atrial fibrillation: rationale and design of the All Nippon AF In the Elderly (ANAFIE) registry. J Cardiol. 2018;72(4):300–6. [DOI] [PubMed] [Google Scholar]
- 8.Yamashita T, Suzuki S, Inoue H, Akao M, Atarashi H, Ikeda T, et al. Two-year outcomes of more than 30 000 elderly patients with atrial fibrillation: results from the All Nippon AF In the Elderly (ANAFIE) Registry. Eur Heart J Qual Care Clin Outcomes. 2022;8(2):202–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rao MP, Vinereanu D, Wojdyla DM, Alexander JH, Atar D, Hylek EM, Hanna M, Wallentin L, Lopes RD, Gersh BJ et al: Clinical outcomes and history of fall in patients with atrial fibrillation treated with oral anticoagulation: insights from the ARISTOTLE Trial. Am J Med. 2018;131(3):269–275 e262. [DOI] [PubMed]
- 10.Banerjee A, Clementy N, Haguenoer K, Fauchier L, Lip GY. Prior history of falls and risk of outcomes in atrial fibrillation: the Loire Valley Atrial Fibrillation Project. Am J Med. 2014;127(10):972–8. [DOI] [PubMed] [Google Scholar]
- 11.Steffel J, Giugliano RP, Braunwald E, Murphy SA, Mercuri M, Choi Y, et al. Edoxaban versus warfarin in atrial fibrillation patients at risk of falling: ENGAGE AF-TIMI 48 analysis. J Am Coll Cardiol. 2016;68(11):1169–78. [DOI] [PubMed] [Google Scholar]
- 12.Tinetti ME, Kumar C. The patient who falls: “It’s always a trade-off.” JAMA. 2010;303(3):258–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Man-Son-Hing M, Nichol G, Lau A, Laupacis A. Choosing antithrombotic therapy for elderly patients with atrial fibrillation who are at risk for falls. Arch Intern Med. 1999;159(7):677–85. [DOI] [PubMed] [Google Scholar]
- 14.Koretsune Y, Yamashita T, Akao M, Atarashi H, Ikeda T, Okumura K, et al. Baseline demographics and clinical characteristics in the All Nippon AF in the Elderly (ANAFIE) registry. Circ J. 2019;83(7):1538–45. [DOI] [PubMed] [Google Scholar]
- 15.Batey M, Hecht J, Callahan C, Wahl W. Direct oral anticoagulants do not worsen traumatic brain injury after low-level falls in the elderly. Surgery. 2018;164(4):814–9. [DOI] [PubMed] [Google Scholar]
- 16.Arita T, Suzuki S, Yagi N, Otsuka T, Semba H, Kano H, et al. Impact of atrial fibrillation on falls in older patients: which is a problem, existence or persistence? J Am Med Dir Assoc. 2019;20(6):765–9. [DOI] [PubMed] [Google Scholar]
- 17.Grymonprez M, Petrovic M, De Backer TL, Steurbaut S, Lahousse L. Clinical outcomes in patients with atrial fibrillation and a history of falls using non-vitamin K antagonist oral anticoagulants: a nationwide cohort study. IJC Heart Vasc. 2023;47:101223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Donze J, Clair C, Hug B, Rodondi N, Waeber G, Cornuz J, et al. Risk of falls and major bleeds in patients on oral anticoagulation therapy. Am J Med. 2012;125(8):773–8. [DOI] [PubMed] [Google Scholar]
- 19.Poli D, Antonucci E, Testa S, Tosetto A, Ageno W, Palareti G, et al. Bleeding risk in very old patients on vitamin K antagonist treatment: results of a prospective collaborative study on elderly patients followed by Italian centres for anticoagulation. Circulation. 2011;124(7):824–9. [DOI] [PubMed] [Google Scholar]
- 20.Gage BF, Birman-Deych E, Kerzner R, Radford MJ, Nilasena DS, Rich MW. Incidence of intracranial hemorrhage in patients with atrial fibrillation who are prone to fall. Am J Med. 2005;118(6):612–7. [DOI] [PubMed] [Google Scholar]
- 21.Liu X, Baumgarten M, Smith G, Gambert S, Gottlieb S, Rattinger G, et al. Warfarin usage among elderly atrial fibrillation patients with traumatic injury, an analysis of United States Medicare fee-for-service enrollees. J Clin Pharmacol. 2015;55(1):25–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Grymonprez M, Steurbaut S, De Backer TL, Petrovic M, Lahousse L. Effectiveness and safety of oral anticoagulants in older patients with atrial fibrillation: a systematic review and meta-analysis. Front Pharmacol. 2020;11:583311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Alonso N, Meinitzer A, Fritz-Petrin E, Enko D, Herrmann M. Role of vitamin K in bone and muscle metabolism. Calcif Tissue Int. 2023;112(2):178–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fusaro M, Dalle Carbonare L, Dusso A, Arcidiacono MV, Valenti MT, Aghi A, et al. Differential effects of dabigatran and warfarin on bone volume and structure in rats with normal renal function. PLoS ONE. 2015;10(8):e0133847. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sugiyama T, Takaki T, Sakanaka K, Sadamaru H, Mori K, Kato Y, et al. Warfarin-induced impairment of cortical bone material quality and compensatory adaptation of cortical bone structure to mechanical stimuli. J Endocrinol. 2007;194(1):213–22. [DOI] [PubMed] [Google Scholar]
- 26.Morishima Y, Kamisato C, Honda Y, Furugohri T, Shibano T. The effects of warfarin and edoxaban, an oral direct factor Xa inhibitor, on gammacarboxylated (Gla-osteocalcin) and undercarboxylated osteocalcin (uc-osteocalcin) in rats. Thromb Res. 2013;131(1):59–63. [DOI] [PubMed] [Google Scholar]
- 27.Tufano A, Coppola A, Contaldi P, Franchini M, Minno GD. Oral anticoagulant drugs and the risk of osteoporosis: new anticoagulants better than old? Semin Thromb Hemost. 2015;41(4):382–8. [DOI] [PubMed] [Google Scholar]
- 28.Tsai SHL, Hu CW, Shao SC, Tischler EH, Obisesan OH, Vervoort D, et al. Comparative risks of fracture among direct oral anticoagulants and warfarin: a systematic review and network meta-analysis. Front Cardiovasc Med. 2022;9:896952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Xie X, Liu Y, Li J, Gu F, Zhang K, Sui Z, et al. Fracture risks in patients with atrial fibrillation treated with different oral anticoagulants: a meta-analysis and systematic review. Age Ageing. 2022;51(1):afab264. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Gage BF. Warfarin-induced fractures in atrial fibrillation? J Am Coll Cardiol. 2019;74(17):2159–61. [DOI] [PubMed] [Google Scholar]
- 31.Binding C, Bjerring Olesen J, Abrahamsen B, Staerk L, Gislason G, Nissen Bonde A. Osteoporotic fractures in patients with atrial fibrillation treated with conventional versus direct anticoagulants. J Am Coll Cardiol. 2019;74(17):2150–8. [DOI] [PubMed] [Google Scholar]
- 32.Shea MK, Loeser RF, Hsu FC, Booth SL, Nevitt M, Simonsick EM, et al. Vitamin K status and lower extremity function in older adults: the health aging and body composition study. J Gerontol A Biol Sci Med Sci. 2016;71(10):1348–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Beaudart C, Locquet M, Touvier M, Reginster JY, Bruyere O. Association between dietary nutrient intake and sarcopenia in the sarcophage study. Aging Clin Exp Res. 2019;31(6):815–24. [DOI] [PubMed] [Google Scholar]
- 34.Leblanc C, Dube MP, Presse N, Dumas S, Nguyen M, Rouleau-Mailloux E, et al. Avoidance of vitamin K-rich foods is common among warfarin users and translates into lower usual vitamin K intakes. J Acad Nutr Diet. 2016;116(6):1000–7. [DOI] [PubMed] [Google Scholar]
- 35.Sim M, Blekkenhorst LC, Lewis JR, Bondonno CP, Devine A, Zhu K, et al. Vegetable and fruit intake and injurious falls risk in older women: a prospective cohort study. Br J Nutr. 2018;120(8):925–34. [DOI] [PubMed] [Google Scholar]
- 36.Ruff CT, Giugliano RP, Braunwald E, Hoffman EB, Deenadayalu N, Ezekowitz MD, et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials. Lancet. 2014;383(9921):955–62. [DOI] [PubMed] [Google Scholar]
- 37.Wild D, Nayak US, Isaacs B. How dangerous are falls in old people at home? Br Med J (Clin Res Ed). 1981;282(6260):266–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The individual deidentified participant data and study protocol will be shared for up to 36 months after publication of the article. Access criteria for data sharing (including requests) will be decided on by a committee led by Daiichi-Sankyo. To gain access, those requesting data access will need to sign a data access agreement. Requests should be directed to yamt-tky@umin.ac.jp.