516831 research-article2014
SJP42310.1177/1403494813516831T.S. Høj Jørgensen et al.Falls and comorbidity
Scandinavian Journal of Public Health, 2014; 42: 287–294
Original Article
Falls and comorbidity: The pathway to fractures
Terese Sara Høj Jørgensen1, Annette Højmann Hansen2, Marie Sahlberg3, Gunnar H. Gislason4–6, Christian Torp-Pedersen7, Charlotte Andersson4, & Ellen Holm1 1Department of Geriatric Medicine, Nykøbing-Falster Hospital, Denmark, 2Geriatric Section, Department of Internal Medicine, Slagelse Hospital, Denmark, 3Department of Geriatric Medicine, Aalborg Hospital, Denmark, 4Department of Cardiology, Copenhagen University Hospital Gentofte, Denmark, 5Faculty of Health and Medical Sciences, University of Copenhagen, Denmark, 6National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark, and 7Institute of Health, Science and Technology, Aalborg University, Denmark
Abstract Aims: To compare nationwide time trends and mortality in hip and proximal humeral fractures; to explore associations between incidences of falls risk related comorbidities (FRICs) and incidence of fractures. Methods: The study is a retrospective cohort study using nationwide Danish administrative registries from 2000 through 2009. Individuals aged 65 years or older who experienced a hip or a proximal humeral fracture were included. Incidence of hip and of proximal humeral fractures, incidence of FRICs (ischemic heart disease, COPD, dementia, depression, diabetes, heart failure, osteoporosis, Parkinson’s disease and stroke) and incidence rate ratios (IRR) for fractures in patients with FRICs, and all-cause mortality up to 10 years after a hip or a proximal humeral fracture were analysed. Results: A total of 89,150 patients experienced hip fractures and 48,581 proximal humeral fractures. From 2000 through 2009, the incidence of hip fractures per 100,000 individuals declined by 198 (787 to 589, OR = 0.75, CI: 0.72–0.80) among males and by 483 (1758 to 1275, OR = 0.74, CI: 0.72–0.77) among females. Incidences of FRICs decreased. The absolute reduction in fractures was most pronounced for the age group above 75 years (2393 to 1884, OR = 0.81, CI: 0.78–0.83), but the relative reduction was more pronounced in the age group of 65–75 years old (496 to 342, OR = 0.70, CI: 0.66–0.74). IRRs for hip fractures and for proximal humeral fractures were significantly elevated in patients with FRICs. Conclusions: The results suggest that the overall reduction in fractures can be explained by reduction in falls related comorbidity. Key Words: Comorbidity, elderly, falls, fractures, hip fracture, humeral fracture
Introduction Hip fracture is a severe trauma associated with high mortality and morbidity in the elderly population. The worldwide incidence of hip fracture has been expected to rise due to an increasing number of elderly in the population [1]. However, during the last decade, several articles report a falling incidence in hip fracture. It is unclear why the incidence is falling. Anti-osteoporotic medications including vitamin D supplementation could be one
explanation, but does not seem to be enough to explain the decreasing incidence [2]. The effect of increasing focus on geriatric medicine including falls prevention [3], reduction in benzodiazepine use [4], implementation of health promoting programmes, increasing Body Mass Index and increasing general health, may all contribute. Several studies have shown that good health in general is maintained into a higher age [5].
Correspondence: Ellen Holm, Geriatric Department, Nykøbing Falster Hospital, 4800 Nykøbing F, Denmark. E-mail: [email protected] (Accepted 22 November 2013) © 2014 the Nordic Societies of Public Health DOI: 10.1177/1403494813516831
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
288 T.S. Høj Jørgensen et al. Theoretically hip fracture incidence may be declining due to fewer traumas, i.e., falls, or due to increased bone strength. If the reason for declining incidence of hip fracture is fewer falls, we would expect the incidence of falls risk factor related comorbidity (FRIC) to decline in a parallel pattern. Cognitive dysfunction [6,7], somatosensory disorders [8–10], orthostatic hypotension and cardiac arrhythmias [11], movement disorders [12], low muscle strength [13] and stroke [14] are all known risk factors for falls. The primary aim of this study was to compare time trends in hip fracture and proximal humeral fractures and explore the association between FRICs and fractures in order to investigate possible reasons for a falling incidence of hip fractures. A secondary aim was to compare mortality after a hip or proximal humeral fracture. Methods Databases For administrative purposes, every individual with permanent residence in Denmark has since 1968 been given a permanent personal identification number either at the time of birth or immigration through the Danish Civil Registration System. This makes it possible to follow Danish citizens by linkage of several national registries. In this study, we used data from four national registries: the Danish Civil Registration System, the Danish National Patient Registry, the Danish National Prescription Registry and the Income Statistics Registry. The Danish Civil Registration System contains dates of births and deaths since 1968 and full registration of household size since 1977 [15]. In the Danish National Patient Registry, all hospital diagnoses since 1994 are registered according to the 10th revision of the International Classification of Diseases (ICD-10). Diagnoses prior to 1994 were coded according to ICD-8. Since 1994,The Danish National Prescription Registry holds information on all prescription drugs sold in Denmark [16]. Finally, since 1970, the Income Statistics Register holds information on income at individual-level [17]. Study population and comorbidity For the present study, we included all individuals living in Denmark, aged 65 years or older between 2000 and 2009. Individuals were followed from the 1st of January 2000, or at the date of their 65th birthday if younger than 65 years at the 1st of January 2000, and until a hip or proximal humeral fracture, death, emigration, or 31st of December 2009. Comorbidities from 1978 and throughout the observation period
were identified by admission diagnoses in the Danish National Patient Registry. Diagnoses of diabetes, Parkinson disease, dementia, osteoporosis and depression are often not made in hospital, and we therefore in addition to in-hospital diagnoses also included prescriptions of medication in the Danish National Prescription Registry as a definition of these diagnoses. The ATC codes used were: diabetes A10; osteoporosis M05B; Parkinson’s disease N04; depression N06A; and dementia N06D. Household size and income The Danish Civil Registration System provides information on family type; marriage, registered partnership, cohabiting individuals, couples living together and single people without children. These five categories were dichotomized into two categories; not living alone and living alone. The Income Statistics Register provides information on household income. The income was calculated as an average of 3 years before start of the study, and was included as a categorical variable. Outcomes The primary outcomes comprised a first admission for hip fracture (ICD-10 code DS72.0-DS72.9) compared to a first admission for proximal humeral fractures (ICD-10 code DS42.0-DS42.9). Validity of diagnoses of hip fracture and proximal humeral fracture in the Danish National Patient Registry is high [18]. A secondary outcome was all-cause mortality at different time points after a hip fracture, compared to mortality after proximal humeral fracture. Statistics Multiple logistic regression models were applied to test for trend of yearly incidence in a) hip fracture and b) proximal humeral fractures in individuals aged 65+ years from 2000 to 2009. The association between comorbidity and hip and proximal humeral fractures were tested with Poisson regression models. Directed Acyclic Graphs (DAGs) were used to clarify the underlying causal network and thereby identify potential confounders [19]. For every chronic condition, a Poisson analysis was made and based on the results from DAGs, the analysis was controlled for age, gender, income, calendar year and comorbidity (ischemic heart disease, COPD, dementia, depression, diabetes, osteoporosis and stroke). The Poisson analysis concerning dementia was further controlled for Parkinson’s disease, and the analysis concerning stroke was further controlled for heart failure. The Poisson analysis concerning association between
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
Falls and comorbidity 289 Table I. Baseline table of 784,832 individuals aged 65 years or above in 2000 in Denmark. n (%)
Male
Female
Couple
Living alone
Population IHD Diabetes Dementia Depression Heart failure COPD Osteoporosis PD Stroke
Age in years (%) 65–69
784,832 (100) 74,120 (9.4) 51,509 (6.6) 13,619 (1.7) 124,748 (15.9) 54,167 (6.9) 35,669 (4.5) 27,321 (3.5) 19,187 (2.4) 65,024 (8.3)
326,018 (41.5) 42,567 (13.1) 24,067 (7.4) 4865 (1.5) 36,199 (11.1) 25,360 (7.8) 17,355 (5.3) 2545 (0.8) 7698 (2.4) 30,629 (9.4)
458,814 (58.5) 31,553 (6.9) 27,442 (6) 8754 (1.9) 88,549 (19.3) 28,807 (6.3) 18,314 (4) 24,776 (5.4) 11,489 (2.5) 34,395 (7.5)
391,143 (50.1) 390,074 (49.9) 39.034 (10) 34,788 (8.9) 23,832 (6.1) 27,527 (7.1) 3588 (0.9) 9993 (2.6) 46,439 (11.9) 78,249 (20.1) 22,876 (5.9) 31,104 (8) 16,286 (4.2) 19,272 (4.9) 9460 (2.4) 17,855 (4.6) 7166 (1.8) 11,998 (3.1) 27,364 (7) 37,461 (9.6)
70–79
>80
215,127 (27.4) 360,356 (45.9) 209,349 (26.7) 15,695 (7.3) 34,357 (9.5) 24,068 (11.5) 12,381 (5.8) 24,174 (6.7) 14,954 (7.1) 1181 (0.6) 4866 (1.4) 7572 (3.6) 26,511 (12.3) 55,141 (15.3) 43,096 (20.6) 7682 (3.6) 22,348 (6.2) 24,137 (11.5) 7510 (3.5) 17,984 (5) 10,175 (4.9) 4482 (2.1) 12,542 (3.5) 10,297 (4.9) 3786 (1.8) 8802 (2.4) 6599 (3.2) 10,684 (5) 28,044 (7.8) 26,296 (12.6)
Note: IHD = Ischemic heart disease, COPD = Chronic obstructive pulmonary disease, PD = Parkinson’s disease.
Parkinson’s disease and fractures was only controlled for age and gender. Incidence rate ratios (IRR) for all-cause mortality at different time points after a hip or proximal humeral fracture were analyzed with Poisson regression analysis. This model was adjusted for age, gender, household size, socioeconomic status, calendar year and comorbidity (ischemic heart disease, COPD, dementia, depression, diabetes, heart failure, osteoporosis, Parkinson’s disease and stroke). Prior to analysis, observational time was divided into bands of one-year intervals. Age was updated at the start of each band. To ensure timely correct categorization of comorbidities and exposure to pharmacotherapy, the observational time was also divided into two periods at time of a new diagnosis or change in medication. The Poisson regression models were all tested with goodness off fit test to ensure the applicability of the models with the data. All analyses were tested at a level of 5% significance. Sensitivity analysis was performed by changing the time periods in the Poisson regression model from 1 year to either half a year or 2 years. All analyses were performed with SAS 9.2. Results The study cohort consisted of 1,276,891 individuals with a maximum follow-up of 10 years. Loss to follow-up due to emigration was 0.44%. Baseline characteristics from 2000 are shown in Table I. The incidence of hip fracture and proximal humeral fractures per 100,000 individuals aged +65 years in the period 2000–2009 in Denmark is shown in Figure 1. During the follow-up period, 89,150 individuals had a hip fracture and 48.581 individuals had a proximal humeral fracture. Individuals with a hip fracture had a mean age of 81.8 years, whereas the mean age of individuals with proximal humeral fractures was 79.2 years. More women than men had hip fracture (72.9%
females) and proximal humeral fractures (76.5 % females). From 2000 through 2009, the incidence of hip fractures declined by 198 (787 to 589, OR = 0.75, CI: 0.72–0.80) among males, and by 483 (1758 to 1275, OR = 0.74, CI: 0.72–0.77) among females. The incidence of hip fracture declined significantly by 154 (496 to 342, OR = 0.70, CI: 0.66–0.74) among individuals aged 65–75 years, and by 509 (2393 to 1884, OR = 0.81, CI: 0.78–0.83) among individuals aged above 75 years. Incidence of proximal humeral fractures per 100,000 individuals declined since 2000. The decline was not significant among males (309 to 312, OR = 1.01, CI: 0.93–1.09) but significant among females (776 to 707, OR = 0.92, CI: 0.88-–0.956). The incidence of proximal humeral fractures increased insignificantly by 10 (359 to 369, OR = 1.03, CI: 0.97–1.10) among individuals aged 65–75 years, and decreased significantly by 82 (848 to 766, OR = 0.91, CI: 0.86– 0.95) among individuals aged above 75 years. Comorbidity The adjusted Poisson regression analyses (Figure 2) revealed that all the chronic conditions were associated with higher IRR of hip fracture, and all conditions apart from ischemic heart disease and heart failure were associated with higher IRR of proximal humeral fractures. Prevalence and incidence of comorbidity are shown in Figure 3. For all diseases, incidences are decreasing, and prevalences are increasing since 2000. All-cause mortality after a hip or proximal humeral fracture During the follow-up period, 52.8 % of individuals with a hip fracture and 39.4 % of individuals with proximal humeral fractures died, compared with
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
290 T.S. Høj Jørgensen et al. Age
Gender
2500
Number per 100,000
2000
1500
1000
500
0 2000
2002
2004
2006
2008
2000
2002
2004
Year
2006
2008
Year
Hip fracture age 65-75 Hip fracture age 75+ Proximal humeral fractures age 65-75 Proximal humeral fractures age 75+
Proximal humeral fractures among males Proximal humeral fractures among females Hip fracture among males Hip fracture among females
Figure 1. The figure shows the development of incidence of hip and proximal humeral fractures according to age (A) and sex (B).
Hip fracture
Proximal humeral fracture IRR
95% CI
IRR
95% CI
Dementia
1.90
1.86-1.95
1.53
1.47-1.60
Parkinson disease
1.61
1.56-1.66
1.48
1.41-1.55
Osteoporosis Chronic obstrubstructive pulmonary disease
1.52
1.49-1.55
1.61
1.57-1.66
1.43
1.39-1.46
1.24
1.19-1.29
1.41
1.39-1.43
1.39
1.36-1.42
1.30
1.28-1.33
1.21
1.18-1.25
1.12
1.09-1.14
1.54
1.48-1.60 0.96-1.03 0.99-1.07
Depression Stroke Diabetes Heart failure
1.12
1.09-1.14
1.00
Iscaemic heart disease
1.07
1.04-1.09
1.03
1
2
1 Incidence rate ratio
2
Figure 2. Incidence rate ratios for hip fracture and for proximal humeral fracture according to comorbidity.
27.1 % of individuals without a hip fracture and 27.41% of individuals without a proximal humeral fracture. The adjusted multiple Poisson regression analyses (Figure 4) revealed a nearly 12-fold increase in
IRR for mortality 14 days after a hip fracture, and a nearly 5-fold increase in mortality after a proximal humeral fracture. One year after hip fracture and proximal humeral fractures, IRRs for mortality were 1.3 and 1.0, respectively.
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
Falls and comorbidity 291 Prevalence
Number per 100,000 individuals aged 65+ years
Incidence 2500
25000
2000
20000
1500
15000
1000
10000
500
5000
0 2000
2002
2004
2006
2008
0 2000
2002
2004
2006
2008
Year
Year Stroke COPD Heart failure IHD Diabetes Parkinson disease Dementia Depression Osteoporosis
Figure 3. Development of incidence and prevalence of falls risk related comorbidity.
IRR for mortality was still increased 10 years after a fracture of hip or proximal humerus. Sensitivity analyses Changing the time periods in the Poisson regression model from 1 year to either half a year or 2 years did not change the conclusions markedly. Discussion Time trend in hip fracture incidence and comorbidity In this nationwide study, we examined the incidence of hip and proximal humeral fractures during a 10-year period in the Danish population above 65 years of age. The main results of the study were: 1) there was a decrease of hip fractures since 2000; 2) a decline in incidence of a number of falls risk-related chronic conditions in the same period; 3) short-term mortality was increased 12- and 5-fold after hip and proximal humeral fractures, respectively.
Fracture happens when two circumstances are fulfilled: The patient has to experience a trauma, and the force must exceed the strength of the bone. Osteoporosis will increase the risk of fracture after a fall due to lower bone strength. Based on knowledge regarding risk factors for falls in the elderly, a number of chronic conditions could be hypothesized to increase the risk of fractures, and these chronic conditions were chosen for analysis in this study. Our results indicate that the hypothesis is correct. Cognitive dysfunction including dementia and depression is a major risk factor for falls. Mobility disorders and gait disturbance as seen in Parkinson’s disease and stroke constitute a risk for falls. Heart failure could be hypothesized to increase risk of falls due to arrhythmias, orthostatic hypotension and hyponatremia [20]. Diabetes may with the risk of hypoglycemia and polyneuropathy, increase the risk of falling. COPD is a risk factor for osteoporosis and also associated with loss of muscle strength [21]. We found a decreasing incidence but a stable or rising prevalence for most of these chronic conditions. It
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
292 T.S. Høj Jørgensen et al.
Proximal humeral fracture
Hip fracture
IRR 95% CI
IRR 95% CI
10 years
1.90 1.85-1.94
1.41 1.37-1.45
5 years
1.57 1.55-1.60
1.18 1.60-1.21
2 years
1.35 1.31-1.38
1.10 1.06-1.14
1 year
1.32 1.28-1.36
1.02 0.98-1.07
1/2 year
1.89 1.83-1.95
1.28 1.21-1.35
90 days
3.52 3.42-3.62
1.84 1.74-1.95
30 days
6.16 5.92-6.41
3.20 2.95-3.47
14 days
11.9 11.6-12.3
4.77 4.45-5.12
1
2
5
10
1 Incidence rate ratio
2
5
10
Figure 4. Incidence rate ratios for all-cause mortality at different time points after a hip or proximal humeral fracture.
is surprising that the incidence of diabetes seems to be falling. However, another Danish registry based study (using a different register) has indicated similar findings [22]. Since prevalence for the chronic conditions has not been falling, it is not clear how the falling incidence in comorbidity could be part of the explanation of the falling incidence in fractures. A possible explanation could be that living with a disease for several years is a marker of less serious disease. An individual who has had for instance ischemic heart disease during 5 years will contribute to prevalence, but the disease may be stable and have little influence on the individual’s life and overall health status. The decreasing incidence in hip fracture has also been seen in other Western countries [23]. In a Dutch study, a peak was seen in 1995, and afterwards a decrease in incidence [24]. This is also seen in Finland and in the US [25,26]. In Canada, there has been a decline since 1985 [27]. However, the reason for the decline is not clear. To our knowledge, this is the first comparative study of fractures in hip and proximal humerus, and the first study to demonstrate an association between fractures and falls related comorbidity. Time trends in proximal humeral fractures The study showed that the incidence of proximal humeral fractures has been only slightly decreasing
and that the decline was most significant in the oldest persons (75+) and in females. The difference in the development in incidence of hip and proximal humeral fractures suggests that the declining incidence cannot be explained only by increase in strength of bone. If that was the case, the incidence of both types of fractures would be expected to decrease in an identical pattern. This is supported by a study by Abrahamsen et al., in which the association between anti-osteoporotic medications and incidence of hip fracture was explored. The authors concluded that 1.3 % and 3.7 % of the declining incidence of hip fractures in men and women, respectively, could be explained by antiosteoporotic therapy [2]. The possibility cannot be excluded that a rise in the use of anti-osteoporosis medications may have at least in part contributed to the falling fracture incidence in our study. However, the study by Abrahamsen et al. indicates that this would only make a small difference in fracture incidence [2]. The differing trend in incidence development between hip and proximal humeral fractures may be a consequence of differences in the mechanism of falls. A fall on an outstretched arm is the injury mechanism for humeral fractures [28], whereas a hip fracture occurs if the falling person does not manage to break the fall, e.g., with an outstretched arm [29]. Managing to break the fall requires appropriate attention, speed
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
Falls and comorbidity 293 of reaction and muscle strength. A healthier person would therefore be more exposed to a humeral fracture compared to a hip fracture. This may also explain the fact that mean age in patients with humeral fractures was lower than in those with hip fractures. Although dementia and Parkinson’s disease were associated with hip fractures as well as proximal humeral fractures the association was far more pronounced for hip fracture. This is probably due to the influence of falls mechanism on fracture site, since slow reaction time is characteristic for both conditions. Mortality We found mortality after a hip fracture significantly elevated up to 10 years after the fracture. During the first 2 weeks after a fracture, the rise in mortality for hip fracture was more than double compared to proximal humeral fracture. A hip fracture will almost always lead to operation and possible complications. A proximal humeral fracture will in most cases be treated with bandaging. It is therefore easily understood that the short time mortality is more pronounced for hip fracture. The fact that the increase in mortality 1 year after a hip and a proximal humeral fracture are very close, indicates that mortality is raised not due to the actual fracture but rather to underlying causes, i.e., the fracture is a marker of frailty. A study in elderly women with hip fracture estimated that approximately 14% of the deaths after hip fracture were caused directly by the hip fracture and surgery, 17% was caused by chronic diseases, and the rest was a combination, but not directly correlated to the fracture [30]. Conclusion Incidences of hip as well as proximal humeral fractures were decreasing in the period 2000 to 2009. The decline in Hip fractures is far more pronounced than the decline in proximal humeral fractures. The decline in proximal humeral fractures is significant only for women and for the oldest old (>75 years). Dementia, Parkinson’s disease, COPD, Osteoporosis, stroke, depression, heart failure, diabetes, and ischemic heart disease are all associated with significant increase in incidence rate ratio for hip fractures and apart from ischemic heart disease and heart failure also for proximal humeral fractures. There is a decrease in the incidences but not in prevalences of all FRICs. Mortality is increased up to 10 years after a fracture. These results indicate that further reduction in fractures depend on general health promotion with the aim of reducing chronic diseases as well as
falls intervention programs aiming to identify specific falls risk factors related to chronic diseases in elderly patients. Due to the high incidence of fractures in the oldest old, focusing such programs on the oldest old (>75 years) will maximize the benefit. Conflict of interest None declared. Funding This research was supported by the public Regional Research Foundation, Region Zealand. References [1] Gullberg B, Johnell O and Kanis JA. World-wide projections for hip fracture. Osteoporos Int 1997;7:407–13. [2] Abrahamsen B and Vestergaard P. Declining incidence of hip fractures and the extent of use of anti-osteoporotic therapy in Denmark 1997-2006. Osteoporos Int 2010;21:373–80. [3] Tinetti ME, Speechley M and Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med 1988;319:1701–7. [4] Holm E, Fosbol E, Pedersen H, et al. Benzodiazepine use in Denmark 1997–2008. European Geriatric Medicine 2012;3:299–303. [5] Bronnum-Hansen H. Health expectancy in Denmark, 1987-2000. Eur J Public Health 2005;15:20–5. [6] Kearney FC, Harwood RH, Gladman JR, et al. The relationship between executive function and falls and gait abnormalities in older adults: a systematic review. Dement Geriatr Cogn Disord 2013;36:20-35. [7] Iaboni A and Flint AJ. The complex interplay of depression and falls in older adults: a clinical review. Am J Geriatr Psychiatry 2013;21:484–92. [8] Janssens L, Brumagne S, McConnell AK, et al. Proprioceptive changes impair balance control in individuals with chronic obstructive pulmonary disease. PLoS One 2013;8:e57949. [9] Buchman AS, Wilson RS, Leurgans S, et al. Vibratory thresholds and mobility in older persons. Muscle & Nerve 2009;39:754–60. [10] Rao N and Aruin AS. Auxiliary sensory cues improve automatic postural responses in individuals with diabetic neuropathy. Neurorehabilitation and Neural Repair 2011;25:110–7. [11] Moya A, Sutton R, Ammirati F, et al. Guidelines for the diagnosis and management of syncope (version 2009): the task force for the diagnosis and management of syncope of the European Society of Cardiology (ESC). Eur Heart J 2009;30:2631–71. [12] Dennison AC, Noorigian JV, Robinson KM, et al. Falling in Parkinson disease: identifying and prioritizing risk factors in recurrent fallers. Am J Phys Med Rehabil 2007;86:621–32. [13] Xue QL, Walston JD, Fried LP, et al. Prediction of risk of falling, physical disability, and frailty by rate of decline in grip strength: the women’s health and aging study. Arch Intern Med 2011;171:1119–21. [14] Batchelor FA, Mackintosh SF, Said CM, et al. Falls after stroke. Int J Stroke 2012;7:482–90. [15] Pedersen CB. The Danish Civil Registration System. Scand J Publ Health 2011;39:22-5. [16] Kildemoes HW, Sorensen HT and Hallas J.The Danish national prescription registry. Scand J Publ Health 2011;39:38–41.
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
294 T.S. Høj Jørgensen et al. [17] Baadsgaard M and Quitzau J. Danish registers on personal income and transfer payments. Scand J Publ Health 2011;39:103–5. [18] Hundrup YA, Hoidrup S, Obel EB, et al. The valid ity of self-reported fractures among Danish female nurses: comparison with fractures registered in the Danish National Hospital Register. Scand J Public Health 2004;32:136–43. [19] Greenland S, Pearl J and Robins JM. Causal diagrams for epidemiologic research. Epidemiology 1999;10:37–48. [20] Renneboog B, Musch W, Vandemergel X, et al. Mild chronic hyponatremia is associated with falls, unsteadiness, and attention deficits. Am J Med 2006;119:71.e1–8. [21] Corsonello A, Antonelli Incalzi R, Pistelli R, et al. Comorbidities of chronic obstructive pulmonary disease. Curr Opin Pulm Med 2011;17 Suppl 1:S21–8. [22] Carstensen B, Kristensen JK, Ottosen P, et al. The Danish National Diabetes Register: trends in incidence, prevalence and mortality. Diabetologia 2008;51:2187–96. [23] Cooper C, Cole ZA, Holroyd CR, et al. Secular trends in the incidence of hip and other osteoporotic fractures. Osteoporos Int 2011;22:1277–88.
[24] Hartholt KA, Oudshoorn C, Zielinski SM, et al. The epidemic of hip fractures: are we on the right track? PloS one 2011;6:e22227. [25] Kannus P, Niemi S, Parkkari J, et al. Nationwide decline in incidence of hip fracture. J Bone Miner Res 2006;21: 1836–8. [26] Brauer CA, Coca-Perraillon M, Cutler DM, et al. Incidence and mortality of hip fractures in the United States. JAMA 2009;302:1573–9. [27] Leslie WD, O’Donnell S, Jean S, et al. Trends in hip fracture rates in Canada. JAMA 2009;302:883–9. [28] Palvanen M, Kannus P, Parkkari J, et al. The injury mechanisms of osteoporotic upper extremity fractures among older adults: a controlled study of 287 consecutive patients and their 108 controls. Osteoporos Int 2000;11:822–31. [29] Parkkari J, Kannus P, Palvanen M, et al. Majority of hip fractures occur as a result of a fall and impact on the greater trochanter of the femur: a prospective controlled hip fracture study with 206 consecutive patients. Calcif Tissue Int 1999;65:183–7. [30] Browner WS, Pressman AR, Nevitt MC, et al. Mortality following fractures in older women. The study of osteoporotic fractures. Arch Intern Med 1996;156:1521–5.
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
SJP42310.1177/1403494813516831T.S. Høj Jørgensen et al.Falls and comorbidity
Scandinavian Journal of Public Health, 2014; 42: 287–294
Original Article
Falls and comorbidity: The pathway to fractures
Terese Sara Høj Jørgensen1, Annette Højmann Hansen2, Marie Sahlberg3, Gunnar H. Gislason4–6, Christian Torp-Pedersen7, Charlotte Andersson4, & Ellen Holm1 1Department of Geriatric Medicine, Nykøbing-Falster Hospital, Denmark, 2Geriatric Section, Department of Internal Medicine, Slagelse Hospital, Denmark, 3Department of Geriatric Medicine, Aalborg Hospital, Denmark, 4Department of Cardiology, Copenhagen University Hospital Gentofte, Denmark, 5Faculty of Health and Medical Sciences, University of Copenhagen, Denmark, 6National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark, and 7Institute of Health, Science and Technology, Aalborg University, Denmark
Abstract Aims: To compare nationwide time trends and mortality in hip and proximal humeral fractures; to explore associations between incidences of falls risk related comorbidities (FRICs) and incidence of fractures. Methods: The study is a retrospective cohort study using nationwide Danish administrative registries from 2000 through 2009. Individuals aged 65 years or older who experienced a hip or a proximal humeral fracture were included. Incidence of hip and of proximal humeral fractures, incidence of FRICs (ischemic heart disease, COPD, dementia, depression, diabetes, heart failure, osteoporosis, Parkinson’s disease and stroke) and incidence rate ratios (IRR) for fractures in patients with FRICs, and all-cause mortality up to 10 years after a hip or a proximal humeral fracture were analysed. Results: A total of 89,150 patients experienced hip fractures and 48,581 proximal humeral fractures. From 2000 through 2009, the incidence of hip fractures per 100,000 individuals declined by 198 (787 to 589, OR = 0.75, CI: 0.72–0.80) among males and by 483 (1758 to 1275, OR = 0.74, CI: 0.72–0.77) among females. Incidences of FRICs decreased. The absolute reduction in fractures was most pronounced for the age group above 75 years (2393 to 1884, OR = 0.81, CI: 0.78–0.83), but the relative reduction was more pronounced in the age group of 65–75 years old (496 to 342, OR = 0.70, CI: 0.66–0.74). IRRs for hip fractures and for proximal humeral fractures were significantly elevated in patients with FRICs. Conclusions: The results suggest that the overall reduction in fractures can be explained by reduction in falls related comorbidity. Key Words: Comorbidity, elderly, falls, fractures, hip fracture, humeral fracture
Introduction Hip fracture is a severe trauma associated with high mortality and morbidity in the elderly population. The worldwide incidence of hip fracture has been expected to rise due to an increasing number of elderly in the population [1]. However, during the last decade, several articles report a falling incidence in hip fracture. It is unclear why the incidence is falling. Anti-osteoporotic medications including vitamin D supplementation could be one
explanation, but does not seem to be enough to explain the decreasing incidence [2]. The effect of increasing focus on geriatric medicine including falls prevention [3], reduction in benzodiazepine use [4], implementation of health promoting programmes, increasing Body Mass Index and increasing general health, may all contribute. Several studies have shown that good health in general is maintained into a higher age [5].
Correspondence: Ellen Holm, Geriatric Department, Nykøbing Falster Hospital, 4800 Nykøbing F, Denmark. E-mail: [email protected] (Accepted 22 November 2013) © 2014 the Nordic Societies of Public Health DOI: 10.1177/1403494813516831
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
288 T.S. Høj Jørgensen et al. Theoretically hip fracture incidence may be declining due to fewer traumas, i.e., falls, or due to increased bone strength. If the reason for declining incidence of hip fracture is fewer falls, we would expect the incidence of falls risk factor related comorbidity (FRIC) to decline in a parallel pattern. Cognitive dysfunction [6,7], somatosensory disorders [8–10], orthostatic hypotension and cardiac arrhythmias [11], movement disorders [12], low muscle strength [13] and stroke [14] are all known risk factors for falls. The primary aim of this study was to compare time trends in hip fracture and proximal humeral fractures and explore the association between FRICs and fractures in order to investigate possible reasons for a falling incidence of hip fractures. A secondary aim was to compare mortality after a hip or proximal humeral fracture. Methods Databases For administrative purposes, every individual with permanent residence in Denmark has since 1968 been given a permanent personal identification number either at the time of birth or immigration through the Danish Civil Registration System. This makes it possible to follow Danish citizens by linkage of several national registries. In this study, we used data from four national registries: the Danish Civil Registration System, the Danish National Patient Registry, the Danish National Prescription Registry and the Income Statistics Registry. The Danish Civil Registration System contains dates of births and deaths since 1968 and full registration of household size since 1977 [15]. In the Danish National Patient Registry, all hospital diagnoses since 1994 are registered according to the 10th revision of the International Classification of Diseases (ICD-10). Diagnoses prior to 1994 were coded according to ICD-8. Since 1994,The Danish National Prescription Registry holds information on all prescription drugs sold in Denmark [16]. Finally, since 1970, the Income Statistics Register holds information on income at individual-level [17]. Study population and comorbidity For the present study, we included all individuals living in Denmark, aged 65 years or older between 2000 and 2009. Individuals were followed from the 1st of January 2000, or at the date of their 65th birthday if younger than 65 years at the 1st of January 2000, and until a hip or proximal humeral fracture, death, emigration, or 31st of December 2009. Comorbidities from 1978 and throughout the observation period
were identified by admission diagnoses in the Danish National Patient Registry. Diagnoses of diabetes, Parkinson disease, dementia, osteoporosis and depression are often not made in hospital, and we therefore in addition to in-hospital diagnoses also included prescriptions of medication in the Danish National Prescription Registry as a definition of these diagnoses. The ATC codes used were: diabetes A10; osteoporosis M05B; Parkinson’s disease N04; depression N06A; and dementia N06D. Household size and income The Danish Civil Registration System provides information on family type; marriage, registered partnership, cohabiting individuals, couples living together and single people without children. These five categories were dichotomized into two categories; not living alone and living alone. The Income Statistics Register provides information on household income. The income was calculated as an average of 3 years before start of the study, and was included as a categorical variable. Outcomes The primary outcomes comprised a first admission for hip fracture (ICD-10 code DS72.0-DS72.9) compared to a first admission for proximal humeral fractures (ICD-10 code DS42.0-DS42.9). Validity of diagnoses of hip fracture and proximal humeral fracture in the Danish National Patient Registry is high [18]. A secondary outcome was all-cause mortality at different time points after a hip fracture, compared to mortality after proximal humeral fracture. Statistics Multiple logistic regression models were applied to test for trend of yearly incidence in a) hip fracture and b) proximal humeral fractures in individuals aged 65+ years from 2000 to 2009. The association between comorbidity and hip and proximal humeral fractures were tested with Poisson regression models. Directed Acyclic Graphs (DAGs) were used to clarify the underlying causal network and thereby identify potential confounders [19]. For every chronic condition, a Poisson analysis was made and based on the results from DAGs, the analysis was controlled for age, gender, income, calendar year and comorbidity (ischemic heart disease, COPD, dementia, depression, diabetes, osteoporosis and stroke). The Poisson analysis concerning dementia was further controlled for Parkinson’s disease, and the analysis concerning stroke was further controlled for heart failure. The Poisson analysis concerning association between
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
Falls and comorbidity 289 Table I. Baseline table of 784,832 individuals aged 65 years or above in 2000 in Denmark. n (%)
Male
Female
Couple
Living alone
Population IHD Diabetes Dementia Depression Heart failure COPD Osteoporosis PD Stroke
Age in years (%) 65–69
784,832 (100) 74,120 (9.4) 51,509 (6.6) 13,619 (1.7) 124,748 (15.9) 54,167 (6.9) 35,669 (4.5) 27,321 (3.5) 19,187 (2.4) 65,024 (8.3)
326,018 (41.5) 42,567 (13.1) 24,067 (7.4) 4865 (1.5) 36,199 (11.1) 25,360 (7.8) 17,355 (5.3) 2545 (0.8) 7698 (2.4) 30,629 (9.4)
458,814 (58.5) 31,553 (6.9) 27,442 (6) 8754 (1.9) 88,549 (19.3) 28,807 (6.3) 18,314 (4) 24,776 (5.4) 11,489 (2.5) 34,395 (7.5)
391,143 (50.1) 390,074 (49.9) 39.034 (10) 34,788 (8.9) 23,832 (6.1) 27,527 (7.1) 3588 (0.9) 9993 (2.6) 46,439 (11.9) 78,249 (20.1) 22,876 (5.9) 31,104 (8) 16,286 (4.2) 19,272 (4.9) 9460 (2.4) 17,855 (4.6) 7166 (1.8) 11,998 (3.1) 27,364 (7) 37,461 (9.6)
70–79
>80
215,127 (27.4) 360,356 (45.9) 209,349 (26.7) 15,695 (7.3) 34,357 (9.5) 24,068 (11.5) 12,381 (5.8) 24,174 (6.7) 14,954 (7.1) 1181 (0.6) 4866 (1.4) 7572 (3.6) 26,511 (12.3) 55,141 (15.3) 43,096 (20.6) 7682 (3.6) 22,348 (6.2) 24,137 (11.5) 7510 (3.5) 17,984 (5) 10,175 (4.9) 4482 (2.1) 12,542 (3.5) 10,297 (4.9) 3786 (1.8) 8802 (2.4) 6599 (3.2) 10,684 (5) 28,044 (7.8) 26,296 (12.6)
Note: IHD = Ischemic heart disease, COPD = Chronic obstructive pulmonary disease, PD = Parkinson’s disease.
Parkinson’s disease and fractures was only controlled for age and gender. Incidence rate ratios (IRR) for all-cause mortality at different time points after a hip or proximal humeral fracture were analyzed with Poisson regression analysis. This model was adjusted for age, gender, household size, socioeconomic status, calendar year and comorbidity (ischemic heart disease, COPD, dementia, depression, diabetes, heart failure, osteoporosis, Parkinson’s disease and stroke). Prior to analysis, observational time was divided into bands of one-year intervals. Age was updated at the start of each band. To ensure timely correct categorization of comorbidities and exposure to pharmacotherapy, the observational time was also divided into two periods at time of a new diagnosis or change in medication. The Poisson regression models were all tested with goodness off fit test to ensure the applicability of the models with the data. All analyses were tested at a level of 5% significance. Sensitivity analysis was performed by changing the time periods in the Poisson regression model from 1 year to either half a year or 2 years. All analyses were performed with SAS 9.2. Results The study cohort consisted of 1,276,891 individuals with a maximum follow-up of 10 years. Loss to follow-up due to emigration was 0.44%. Baseline characteristics from 2000 are shown in Table I. The incidence of hip fracture and proximal humeral fractures per 100,000 individuals aged +65 years in the period 2000–2009 in Denmark is shown in Figure 1. During the follow-up period, 89,150 individuals had a hip fracture and 48.581 individuals had a proximal humeral fracture. Individuals with a hip fracture had a mean age of 81.8 years, whereas the mean age of individuals with proximal humeral fractures was 79.2 years. More women than men had hip fracture (72.9%
females) and proximal humeral fractures (76.5 % females). From 2000 through 2009, the incidence of hip fractures declined by 198 (787 to 589, OR = 0.75, CI: 0.72–0.80) among males, and by 483 (1758 to 1275, OR = 0.74, CI: 0.72–0.77) among females. The incidence of hip fracture declined significantly by 154 (496 to 342, OR = 0.70, CI: 0.66–0.74) among individuals aged 65–75 years, and by 509 (2393 to 1884, OR = 0.81, CI: 0.78–0.83) among individuals aged above 75 years. Incidence of proximal humeral fractures per 100,000 individuals declined since 2000. The decline was not significant among males (309 to 312, OR = 1.01, CI: 0.93–1.09) but significant among females (776 to 707, OR = 0.92, CI: 0.88-–0.956). The incidence of proximal humeral fractures increased insignificantly by 10 (359 to 369, OR = 1.03, CI: 0.97–1.10) among individuals aged 65–75 years, and decreased significantly by 82 (848 to 766, OR = 0.91, CI: 0.86– 0.95) among individuals aged above 75 years. Comorbidity The adjusted Poisson regression analyses (Figure 2) revealed that all the chronic conditions were associated with higher IRR of hip fracture, and all conditions apart from ischemic heart disease and heart failure were associated with higher IRR of proximal humeral fractures. Prevalence and incidence of comorbidity are shown in Figure 3. For all diseases, incidences are decreasing, and prevalences are increasing since 2000. All-cause mortality after a hip or proximal humeral fracture During the follow-up period, 52.8 % of individuals with a hip fracture and 39.4 % of individuals with proximal humeral fractures died, compared with
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
290 T.S. Høj Jørgensen et al. Age
Gender
2500
Number per 100,000
2000
1500
1000
500
0 2000
2002
2004
2006
2008
2000
2002
2004
Year
2006
2008
Year
Hip fracture age 65-75 Hip fracture age 75+ Proximal humeral fractures age 65-75 Proximal humeral fractures age 75+
Proximal humeral fractures among males Proximal humeral fractures among females Hip fracture among males Hip fracture among females
Figure 1. The figure shows the development of incidence of hip and proximal humeral fractures according to age (A) and sex (B).
Hip fracture
Proximal humeral fracture IRR
95% CI
IRR
95% CI
Dementia
1.90
1.86-1.95
1.53
1.47-1.60
Parkinson disease
1.61
1.56-1.66
1.48
1.41-1.55
Osteoporosis Chronic obstrubstructive pulmonary disease
1.52
1.49-1.55
1.61
1.57-1.66
1.43
1.39-1.46
1.24
1.19-1.29
1.41
1.39-1.43
1.39
1.36-1.42
1.30
1.28-1.33
1.21
1.18-1.25
1.12
1.09-1.14
1.54
1.48-1.60 0.96-1.03 0.99-1.07
Depression Stroke Diabetes Heart failure
1.12
1.09-1.14
1.00
Iscaemic heart disease
1.07
1.04-1.09
1.03
1
2
1 Incidence rate ratio
2
Figure 2. Incidence rate ratios for hip fracture and for proximal humeral fracture according to comorbidity.
27.1 % of individuals without a hip fracture and 27.41% of individuals without a proximal humeral fracture. The adjusted multiple Poisson regression analyses (Figure 4) revealed a nearly 12-fold increase in
IRR for mortality 14 days after a hip fracture, and a nearly 5-fold increase in mortality after a proximal humeral fracture. One year after hip fracture and proximal humeral fractures, IRRs for mortality were 1.3 and 1.0, respectively.
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
Falls and comorbidity 291 Prevalence
Number per 100,000 individuals aged 65+ years
Incidence 2500
25000
2000
20000
1500
15000
1000
10000
500
5000
0 2000
2002
2004
2006
2008
0 2000
2002
2004
2006
2008
Year
Year Stroke COPD Heart failure IHD Diabetes Parkinson disease Dementia Depression Osteoporosis
Figure 3. Development of incidence and prevalence of falls risk related comorbidity.
IRR for mortality was still increased 10 years after a fracture of hip or proximal humerus. Sensitivity analyses Changing the time periods in the Poisson regression model from 1 year to either half a year or 2 years did not change the conclusions markedly. Discussion Time trend in hip fracture incidence and comorbidity In this nationwide study, we examined the incidence of hip and proximal humeral fractures during a 10-year period in the Danish population above 65 years of age. The main results of the study were: 1) there was a decrease of hip fractures since 2000; 2) a decline in incidence of a number of falls risk-related chronic conditions in the same period; 3) short-term mortality was increased 12- and 5-fold after hip and proximal humeral fractures, respectively.
Fracture happens when two circumstances are fulfilled: The patient has to experience a trauma, and the force must exceed the strength of the bone. Osteoporosis will increase the risk of fracture after a fall due to lower bone strength. Based on knowledge regarding risk factors for falls in the elderly, a number of chronic conditions could be hypothesized to increase the risk of fractures, and these chronic conditions were chosen for analysis in this study. Our results indicate that the hypothesis is correct. Cognitive dysfunction including dementia and depression is a major risk factor for falls. Mobility disorders and gait disturbance as seen in Parkinson’s disease and stroke constitute a risk for falls. Heart failure could be hypothesized to increase risk of falls due to arrhythmias, orthostatic hypotension and hyponatremia [20]. Diabetes may with the risk of hypoglycemia and polyneuropathy, increase the risk of falling. COPD is a risk factor for osteoporosis and also associated with loss of muscle strength [21]. We found a decreasing incidence but a stable or rising prevalence for most of these chronic conditions. It
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
292 T.S. Høj Jørgensen et al.
Proximal humeral fracture
Hip fracture
IRR 95% CI
IRR 95% CI
10 years
1.90 1.85-1.94
1.41 1.37-1.45
5 years
1.57 1.55-1.60
1.18 1.60-1.21
2 years
1.35 1.31-1.38
1.10 1.06-1.14
1 year
1.32 1.28-1.36
1.02 0.98-1.07
1/2 year
1.89 1.83-1.95
1.28 1.21-1.35
90 days
3.52 3.42-3.62
1.84 1.74-1.95
30 days
6.16 5.92-6.41
3.20 2.95-3.47
14 days
11.9 11.6-12.3
4.77 4.45-5.12
1
2
5
10
1 Incidence rate ratio
2
5
10
Figure 4. Incidence rate ratios for all-cause mortality at different time points after a hip or proximal humeral fracture.
is surprising that the incidence of diabetes seems to be falling. However, another Danish registry based study (using a different register) has indicated similar findings [22]. Since prevalence for the chronic conditions has not been falling, it is not clear how the falling incidence in comorbidity could be part of the explanation of the falling incidence in fractures. A possible explanation could be that living with a disease for several years is a marker of less serious disease. An individual who has had for instance ischemic heart disease during 5 years will contribute to prevalence, but the disease may be stable and have little influence on the individual’s life and overall health status. The decreasing incidence in hip fracture has also been seen in other Western countries [23]. In a Dutch study, a peak was seen in 1995, and afterwards a decrease in incidence [24]. This is also seen in Finland and in the US [25,26]. In Canada, there has been a decline since 1985 [27]. However, the reason for the decline is not clear. To our knowledge, this is the first comparative study of fractures in hip and proximal humerus, and the first study to demonstrate an association between fractures and falls related comorbidity. Time trends in proximal humeral fractures The study showed that the incidence of proximal humeral fractures has been only slightly decreasing
and that the decline was most significant in the oldest persons (75+) and in females. The difference in the development in incidence of hip and proximal humeral fractures suggests that the declining incidence cannot be explained only by increase in strength of bone. If that was the case, the incidence of both types of fractures would be expected to decrease in an identical pattern. This is supported by a study by Abrahamsen et al., in which the association between anti-osteoporotic medications and incidence of hip fracture was explored. The authors concluded that 1.3 % and 3.7 % of the declining incidence of hip fractures in men and women, respectively, could be explained by antiosteoporotic therapy [2]. The possibility cannot be excluded that a rise in the use of anti-osteoporosis medications may have at least in part contributed to the falling fracture incidence in our study. However, the study by Abrahamsen et al. indicates that this would only make a small difference in fracture incidence [2]. The differing trend in incidence development between hip and proximal humeral fractures may be a consequence of differences in the mechanism of falls. A fall on an outstretched arm is the injury mechanism for humeral fractures [28], whereas a hip fracture occurs if the falling person does not manage to break the fall, e.g., with an outstretched arm [29]. Managing to break the fall requires appropriate attention, speed
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
Falls and comorbidity 293 of reaction and muscle strength. A healthier person would therefore be more exposed to a humeral fracture compared to a hip fracture. This may also explain the fact that mean age in patients with humeral fractures was lower than in those with hip fractures. Although dementia and Parkinson’s disease were associated with hip fractures as well as proximal humeral fractures the association was far more pronounced for hip fracture. This is probably due to the influence of falls mechanism on fracture site, since slow reaction time is characteristic for both conditions. Mortality We found mortality after a hip fracture significantly elevated up to 10 years after the fracture. During the first 2 weeks after a fracture, the rise in mortality for hip fracture was more than double compared to proximal humeral fracture. A hip fracture will almost always lead to operation and possible complications. A proximal humeral fracture will in most cases be treated with bandaging. It is therefore easily understood that the short time mortality is more pronounced for hip fracture. The fact that the increase in mortality 1 year after a hip and a proximal humeral fracture are very close, indicates that mortality is raised not due to the actual fracture but rather to underlying causes, i.e., the fracture is a marker of frailty. A study in elderly women with hip fracture estimated that approximately 14% of the deaths after hip fracture were caused directly by the hip fracture and surgery, 17% was caused by chronic diseases, and the rest was a combination, but not directly correlated to the fracture [30]. Conclusion Incidences of hip as well as proximal humeral fractures were decreasing in the period 2000 to 2009. The decline in Hip fractures is far more pronounced than the decline in proximal humeral fractures. The decline in proximal humeral fractures is significant only for women and for the oldest old (>75 years). Dementia, Parkinson’s disease, COPD, Osteoporosis, stroke, depression, heart failure, diabetes, and ischemic heart disease are all associated with significant increase in incidence rate ratio for hip fractures and apart from ischemic heart disease and heart failure also for proximal humeral fractures. There is a decrease in the incidences but not in prevalences of all FRICs. Mortality is increased up to 10 years after a fracture. These results indicate that further reduction in fractures depend on general health promotion with the aim of reducing chronic diseases as well as
falls intervention programs aiming to identify specific falls risk factors related to chronic diseases in elderly patients. Due to the high incidence of fractures in the oldest old, focusing such programs on the oldest old (>75 years) will maximize the benefit. Conflict of interest None declared. Funding This research was supported by the public Regional Research Foundation, Region Zealand. References [1] Gullberg B, Johnell O and Kanis JA. World-wide projections for hip fracture. Osteoporos Int 1997;7:407–13. [2] Abrahamsen B and Vestergaard P. Declining incidence of hip fractures and the extent of use of anti-osteoporotic therapy in Denmark 1997-2006. Osteoporos Int 2010;21:373–80. [3] Tinetti ME, Speechley M and Ginter SF. Risk factors for falls among elderly persons living in the community. N Engl J Med 1988;319:1701–7. [4] Holm E, Fosbol E, Pedersen H, et al. Benzodiazepine use in Denmark 1997–2008. European Geriatric Medicine 2012;3:299–303. [5] Bronnum-Hansen H. Health expectancy in Denmark, 1987-2000. Eur J Public Health 2005;15:20–5. [6] Kearney FC, Harwood RH, Gladman JR, et al. The relationship between executive function and falls and gait abnormalities in older adults: a systematic review. Dement Geriatr Cogn Disord 2013;36:20-35. [7] Iaboni A and Flint AJ. The complex interplay of depression and falls in older adults: a clinical review. Am J Geriatr Psychiatry 2013;21:484–92. [8] Janssens L, Brumagne S, McConnell AK, et al. Proprioceptive changes impair balance control in individuals with chronic obstructive pulmonary disease. PLoS One 2013;8:e57949. [9] Buchman AS, Wilson RS, Leurgans S, et al. Vibratory thresholds and mobility in older persons. Muscle & Nerve 2009;39:754–60. [10] Rao N and Aruin AS. Auxiliary sensory cues improve automatic postural responses in individuals with diabetic neuropathy. Neurorehabilitation and Neural Repair 2011;25:110–7. [11] Moya A, Sutton R, Ammirati F, et al. Guidelines for the diagnosis and management of syncope (version 2009): the task force for the diagnosis and management of syncope of the European Society of Cardiology (ESC). Eur Heart J 2009;30:2631–71. [12] Dennison AC, Noorigian JV, Robinson KM, et al. Falling in Parkinson disease: identifying and prioritizing risk factors in recurrent fallers. Am J Phys Med Rehabil 2007;86:621–32. [13] Xue QL, Walston JD, Fried LP, et al. Prediction of risk of falling, physical disability, and frailty by rate of decline in grip strength: the women’s health and aging study. Arch Intern Med 2011;171:1119–21. [14] Batchelor FA, Mackintosh SF, Said CM, et al. Falls after stroke. Int J Stroke 2012;7:482–90. [15] Pedersen CB. The Danish Civil Registration System. Scand J Publ Health 2011;39:22-5. [16] Kildemoes HW, Sorensen HT and Hallas J.The Danish national prescription registry. Scand J Publ Health 2011;39:38–41.
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
294 T.S. Høj Jørgensen et al. [17] Baadsgaard M and Quitzau J. Danish registers on personal income and transfer payments. Scand J Publ Health 2011;39:103–5. [18] Hundrup YA, Hoidrup S, Obel EB, et al. The valid ity of self-reported fractures among Danish female nurses: comparison with fractures registered in the Danish National Hospital Register. Scand J Public Health 2004;32:136–43. [19] Greenland S, Pearl J and Robins JM. Causal diagrams for epidemiologic research. Epidemiology 1999;10:37–48. [20] Renneboog B, Musch W, Vandemergel X, et al. Mild chronic hyponatremia is associated with falls, unsteadiness, and attention deficits. Am J Med 2006;119:71.e1–8. [21] Corsonello A, Antonelli Incalzi R, Pistelli R, et al. Comorbidities of chronic obstructive pulmonary disease. Curr Opin Pulm Med 2011;17 Suppl 1:S21–8. [22] Carstensen B, Kristensen JK, Ottosen P, et al. The Danish National Diabetes Register: trends in incidence, prevalence and mortality. Diabetologia 2008;51:2187–96. [23] Cooper C, Cole ZA, Holroyd CR, et al. Secular trends in the incidence of hip and other osteoporotic fractures. Osteoporos Int 2011;22:1277–88.
[24] Hartholt KA, Oudshoorn C, Zielinski SM, et al. The epidemic of hip fractures: are we on the right track? PloS one 2011;6:e22227. [25] Kannus P, Niemi S, Parkkari J, et al. Nationwide decline in incidence of hip fracture. J Bone Miner Res 2006;21: 1836–8. [26] Brauer CA, Coca-Perraillon M, Cutler DM, et al. Incidence and mortality of hip fractures in the United States. JAMA 2009;302:1573–9. [27] Leslie WD, O’Donnell S, Jean S, et al. Trends in hip fracture rates in Canada. JAMA 2009;302:883–9. [28] Palvanen M, Kannus P, Parkkari J, et al. The injury mechanisms of osteoporotic upper extremity fractures among older adults: a controlled study of 287 consecutive patients and their 108 controls. Osteoporos Int 2000;11:822–31. [29] Parkkari J, Kannus P, Palvanen M, et al. Majority of hip fractures occur as a result of a fall and impact on the greater trochanter of the femur: a prospective controlled hip fracture study with 206 consecutive patients. Calcif Tissue Int 1999;65:183–7. [30] Browner WS, Pressman AR, Nevitt MC, et al. Mortality following fractures in older women. The study of osteoporotic fractures. Arch Intern Med 1996;156:1521–5.
Downloaded from sjp.sagepub.com at HOWARD UNIV UNDERGRAD LIBRARY on February 18, 2015
