Skip Navigation


American Journal of Epidemiology Advance Access originally published online on December 8, 2008
American Journal of Epidemiology 2009 169(2):143-149; doi:10.1093/aje/kwn379
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
169/2/143    most recent
kwn379v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Related articles in Am. J. Epidemiol.
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Wagner, H.
Right arrow Articles by Michaëlsson, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wagner, H.
Right arrow Articles by Michaëlsson, K.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

American Journal of Epidemiology © The Author 2008. Published by the Johns Hopkins Bloomberg School of Public Health. All rights reserved. For permissions, please e-mail: journals.permissions@oxfordjournals.org.

ORIGINAL CONTRIBUTIONS

Simply Ask Them About Their Balance—Future Fracture Risk in a Nationwide Cohort Study of Twins

Helene Wagner, Håkan Melhus, Rolf Gedeborg, Nancy L. Pedersen and Karl Michaëlsson

Correspondence to Dr. Karl Michaëlsson, Department of Surgical Sciences, Section of Orthopaedics, University Hospital, S751 39 Uppsala, Sweden (e-mail: karl.michaelsson{at}surgsci.uu.se).

Received for publication March 18, 2008. Accepted for publication June 25, 2008.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
The principal causal components of an osteoporotic fracture are a fall and weakened bone strength. While bone quality measures have been frequently studied, the ability of simple measures of impaired balance to predict fracture risk has received less attention. Computer-assisted telephone interviews were conducted between 1998 and 2000 among 24,598 Swedish twins aged 55 years or older. Impaired balance at the time of interview was reported by 2,890 (12%) of the twins. Twin pairs who were discordant with regard to impaired balance were selected for analysis and followed for fractures through 2005. In a pairwise analysis, the odds ratio for hip fracture was 3.13 (95% confidence interval (CI): 1.62, 6.05) among twins with impaired balance as compared with their co-twins with normal balance. When previously recognized clinical risk factors for osteoporotic fracture were considered in the model, the odds ratio for hip fracture with impaired balance was 3.88 (95% CI: 1.40, 10.72). Approximately 40% of all hip fractures were attributable to impaired balance. The odds ratios for any fracture and any osteoporotic fracture for twins with impaired balance were 2.00 (95% CI: 1.29, 3.11) and 2.39 (95% CI: 1.49, 3.82), respectively. These results imply that self-reported impaired balance is a novel and readily assessed risk factor for future fractures in the elderly.

accidental falls; bone and bones; fractures, bone; frail elderly; musculoskeletal equilibrium; osteoporosis; risk factors; twins

Abbreviations: CI, confidence interval; OR, odds ratio; RR, relative risk


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
The principal component causes of an osteoporotic fracture are a fall and a fragile skeleton. Most research on primary prevention of these fractures has focused on bone-related risk factors (1–3). However, the vast majority of all fractures in elderly persons occur after a fall. Therefore, the risk of a future fall, not only bone quality, should be integrated into an individual risk profile for osteoporotic fracture (4).

Attempts to examine the value of variables related to balance and fall risk in assessing the risk of osteoporotic fracture have been few. A measured impaired balance is associated with increased risk of falls (5, 6). However, even if direct measurement of balance performance might be preferable, it is not easily accomplished or reproduced. Thus, self-reports could be used as a good and simple alternative. Self-reported impaired balance is strongly associated with performance-based inferior postural stability (7–12). Measured impaired balance is an independent risk factor for osteoporotic fractures in some (13) but not all (14–16) prospective studies. Self-reported falls have also been reported as a risk factor (15, 16) but not an independent predictor of fracture risk (14).

There is a need for a simple tool with which to identify subjects who are at increased risk of injurious falls. Self-reported impaired balance might be such a readily accessible risk factor. The precise matching on age, together with control of shared genes and shared environmental influences, makes a twin design (17, 18) suitable for the study of an exposure effect on fracture risk. Therefore, we used data from a population-based Swedish twin registry to estimate the future risk of fractures among twins aged 55–88 years who were discordant with regard to self-reported impaired balance.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Subjects were ascertained from the Swedish Twin Registry, currently the largest twin registry in the world (17, 18). Twins aged 55 years or older were considered eligible for this study—a total of 29,276 subjects.

Balance
A comprehensive computer-assisted telephone interview was conducted between 1998 and 2000. The interview included a number of items regarding lifestyle, diseases, and symptoms (17, 18). One such item was whether the twin had impaired balance. The question was "Do you have impaired balance?," with the following response alternatives: yes, no, or uncertain. In total, 24,598 persons (84%) agreed to participate in the interview. Efforts were made to interview members of a pair within 1 month.

Approximately 2.5% of the twins gave the answer "uncertain" to the balance question. Only the 23,487 twins who answered the balance question unequivocally (yes/no) were included. These subjects formed our study base. We restricted the statistical analysis to twins who were discordant with regard to self-reported impaired balance.

For reliability testing, an identical second interview was performed among 105 twins 2 weeks after the first interview. The kappa statistic for the question regarding impaired balance was 0.81 (95% confidence interval (CI): 0.60, 1.00), indicating almost perfect agreement (19).

Zygosity
Information on zygosity was obtained at the time of registry compilation on the basis of questions about childhood resemblance. This information was updated during the most recent screening for persons with previously uncertain zygosity. Four separate validation studies using serologic testing and/or genotyping have shown that, with these items, 95%–98% of twin pairs are classified correctly (17, 18). Twins with uncertain zygosity were excluded from the analysis.

Fractures
Prevalent fractures at study entry (i.e., the time of telephone interview) were identified both by matching to the Swedish National Hospital Discharge Register through December 31, 2005, and by self-reports. Self-reported fractures in this cohort have been found to have high sensitivity and specificity (20). We also matched the study cohort to the National Hospital Discharge Register to identify incident hospital admissions for fractures that occurred after study entry. These incident fractures formed the outcomes in the analysis.

Fractures were grouped into hip fracture, any osteoporotic fracture (21) (hip, pelvis, spine, distal forearm, or proximal humerus—International Classification of Diseases, Tenth Revision, codes S720–S722, S321–S328, S220 and S320, S524–S528, and S422, respectively), and any fracture. For each outcome, fracture cases who had experienced a similar fracture before study entry were excluded from the analysis, since their physical performance or report of balance could have been influenced by their previous fracture. Without this exclusion, there would also have been a risk of including a prevalent case as an incident case. Since hip fractures are less common than any osteoporotic fracture and any fracture, fewer twins were excluded from the hip fracture analysis than from analyses of the other 2 outcomes. In addition, we included only those persons who experienced a fracture after a fall. We used the International Classification of Diseases, Tenth Revision, codes for external causes of injury for identification of the direct mechanism of the fracture. The records of all twins were also matched to the National Hospital Discharge Register and cancer registries for identification of other diagnoses besides fracture.

Statistical analysis
Fracture risk associated with impaired balance was analyzed in pairs of twins who were discordant with regard to impaired balance. For these associations, we used conditional logistic regression to estimate odds ratios and 95% confidence intervals. We wanted each twin in the pair to have a similar time at risk. Since mortality risk was associated with impaired balance, we adjusted for the unequal lengths of follow-up due to death. Thus, twin pairs in which 1 twin died before the end of follow-up (15%) were excluded from the analysis.

We used 2 different analytical models. The first comprised simple sex- and age-adjusted pairwise comparisons, and the second was a multivariable model that additionally included previously recognized clinical risk factors for osteoporotic fracture. As such risk factors, we included difficulty rising from a chair (self-report, yes/no); a previous fracture after the age of 50 years, based on self-reports and on data from the National Hospital Discharge Register (yes/no); any registered fall from a standing height within the past 10 years (yes/no); smoking status (never, former, current); physical activity level (low, medium, high); present use of corticosteroids (yes/no); body mass index (weight (kg)/height (m)2, continuous); body weight (kg, continuous); height (cm, continuous); alcohol intake (abstainer, normal intake, suspected abuse); diabetes mellitus (yes/no); and an index for activities of daily living (continuous). Physical activity was assessed at the interview by means of the following question: "How much exercise do you perform?" The response alternatives were 1) mostly sedentary activities; 2) light exercise such as walking or light gardening; 3) regular recreational sports; and 4) regular hard physical training. Based on the highest positive response to these questions, 3 leisure physical activity categories were constructed: low (category 1), moderate (categories 2 and 3), and high (category 4). Similar simple physical activity alternatives have been previously found to be positively associated with physical fitness in 70-year-old men, measured both as capillary density of vastus lateralis muscle fibers and as physical performance in a symptom-limited exercise test carried out on an ergometer cycle (22). The activities of daily living index, with a possible range of 0–25 points, was constructed from self-reported responses to the following questions: 1) "Do you have help to cook your food and do your shopping?"; 2) "Do you have help with dressing and bathing?"; 3) "Do you have help to remind you to take your medicine?"; 4) "Do you have daily help with household work?"; 5) "Do you need any other daily living support?" Negative answers to all 5 questions produced a score of 25. This sum was reduced by 5 points for each positive answer to the questions.

Subgroup analyses were performed with stratification for age (≤75 years vs. >75 years) and, in twins of the same sex only, additional stratification for zygosity and sex. The men had few fractures, which contributed to unstable estimates. Furthermore, since some subgroups contained few cases, we were not able to fit the full multivariable model. We therefore present only the sex- and age-adjusted results for some subgroup analyses.

The population attributable risk of fracture due to impaired balance was calculated as p(OR – 1)/[p(OR – 1) + 1], where p is the prevalence of impaired balance among the fracture cases and OR is the odds ratio. Therefore, the estimated population attributable risk was dependent on the prevalence of impaired balance. The proportion of twins with impaired balance was based on all twins, not only discordant pairs.

Finally, we estimated odds ratios for hip fracture in twin pairs who were discordant with regard to smoking status at study entry, current systemic cortisone use, low physical activity, or difficulty rising from a chair.

We performed all analyses using the PHREG procedure in SAS software (version 9.1; SAS Institute Inc., Cary, North Carolina).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
A total of 2,890 twins (12%) reported impaired balance at study entry. Impaired balance was more common among women (15%) than among men (9%). In the whole cohort, the proportion of subjects who reported impaired balance increased with increasing age: Among persons younger than 70 years, 8% stated that they had impaired balance; among persons aged 70–79 years, the proportion was 18%; and among persons aged 80 years or older, 31% stated that they had impaired balance.

Descriptive characteristics of balance-discordant female, male, and unlike-sexed twin pairs are displayed in Table 1. These pairs were included in the hip fracture analysis. The twins were aged 55–88 years. Impaired balance was associated with low physical activity and several disease categories.


View this table:
[in this window]
[in a new window]

  		Table 1. Selected Characteristics of Female and Male Like-Sexed and Unlike-Sexed Twins Discordant With Regard to Self-Reported Impaired Balance, Swedish Twin Registry, 1998–2000

 
There were 714 twin pairs who were discordant for impaired balance in the overall fracture analysis. The within-pair comparison (Table 2) revealed that twins with self-reported impaired balance had twice the risk of suffering any fracture as their co-twins without impaired balance (odds ratio (OR) = 2.00, 95% CI: 1.29, 3.11). This estimate was not attenuated when we also considered previously recognized clinical risk factors in the multivariable model. Approximately one-fifth of fractures of any type were attributable to impaired balance. An indication that the increased risk of any fracture was independent of common familial effects was the odds ratio of 2.67 (95% CI: 1.04, 6.82) in monozygotic twins. The risk of fractures associated with impaired balance was also doubled when analysis was carried out in women only, but there were too few fracture cases for precise estimation of risk in men. The increased risk of fracture seemed to persist in persons older than age 75 years (OR = 3.58, 95% CI: 0.98, 13.07).


View this table:
[in this window]
[in a new window]

  		Table 2. Odds Ratios for Fracture in Persons With Impaired Balance as Compared With Those With Normal Balance Among Twin Pairs Discordant With Regard to Self-Reported Balance Status, Swedish Twin Registry, 1998–2005a

 
Impaired balance was associated with an increased risk of any type of osteoporotic fracture, with an odds ratio of 2.39 (95% CI: 1.49, 3.82) that was not attenuated after adjustment for clinical risk factors (OR = 2.99, 95% CI: 1.56, 5.73). The increase in risk also remained when the analysis was restricted to non-hip-fracture cases (OR = 2.00, 95% CI: 1.08, 3.72). If we restricted the analysis to women among like-sexed pairs, those with impaired balance had an odds ratio of 3.00 (95% CI: 1.60, 5.62) for any osteoporotic fracture in comparison with their co-twins with reported normal balance. The increase in fracture risk was observed among persons aged 75 years or younger (OR = 2.05, 95% CI: 1.23, 3.41) as well as in those older than age 75 years (OR = 5.00, 95% CI: 1.45, 17.27).

In the hip fracture analysis, there were 864 discordant pairs. Twins with reported impaired balance had 3 times the risk of incurring a hip fracture as their co-twins (OR = 3.13, 95% CI: 1.62, 6.05; multivariable-adjusted OR = 3.88, 95% CI: 1.40, 10.72). Approximately 40% of all hip fractures were attributable to impaired balance. We found an increased risk of hip fracture among women with impaired balance and among both persons aged 75 years or younger (OR = 2.98, 95% CI: 1.31, 6.79) and persons older than age 75 years (OR = 3.50, 95% CI: 1.15, 10.63).

As a comparison with established risk factors for hip fracture, we also evaluated the risk of hip fracture in 1,515 twin pairs who were discordant with regard to current smoking. The odds ratio for hip fracture for current smoking as compared with nonsmoking was 2.67 (95% CI: 1.12, 6.36). In 247 pairs discordant for cortisone use, the odds ratio for hip fracture with use was 2.23 (0.48, 10.34). A low level of physical activity conferred an odds ratio of 1.78 (95% CI: 1.01, 3.13) for hip fracture. This analysis was confined to 2,265 twin pairs who were discordant for physical activity, where 1 twin had reported a low physical activity level and the co-twin reported a medium or high physical activity level. Difficulty rising from a chair, however, was not associated with increased hip fracture risk in an analysis carried out among 1,323 discordant twin pairs. After adjustment for impaired balance, the odds ratio for hip fracture was 1.02 (95% CI: 0.48, 2.18), whereas the crude odds ratio was 1.47 (95% CI: 0.76, 2.83).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Results from this nationwide, prospective, population-based cohort study of twins imply that elderly persons who answer yes to a simple question regarding impaired balance have a substantially increased risk of fracture. This easily accessible risk factor was responsible for 20%–50% of the variation in fracture risk, depending on site and age. The highest population attributable risks were obtained among persons older than 75 years of age.

Several authors have concluded that bone mineral density alone does not predict future osteoporotic fracture risk sufficiently (4, 13, 14), especially in the oldest subjects (4), but few previous prospective cohort studies have addressed the impact of falls or balance performance on fracture risk. Results have been inconsistent. Dargent-Molina et al. (13) found that measured fall- and balance-related factors, such as gait speed (per standard-deviation decrease, relative risk (RR) = 1.4) and tandem walking (per 1-point increase in score, RR = 1.2), were associated with hip fracture risk in women aged 75 years or older. Previous falls or fractures were not considered in the analysis. Cummings et al. (14), however, found that a fall in the previous year, gait speed, and neuromuscular testing were not able to discriminate female hip fracture patients independently of quadriceps muscle strength, measured as inability to rise from a chair (RR = 2.1). Previous fractures were not considered in the analysis. Nevertheless, Nguyen et al. (23) identified a fall in the past year (RR = 1.4 in women and RR = 1.2 in men), prior fractures (RR = 2.2 in women and RR = 1.8 in men), and quadriceps strength as independent risk factors for hip fracture in the Dubbo Study, even though the last variable contributed minimally to the model. Balance, measured as postural sway, was not associated with fracture risk after adjustment for the other variables. Finally, in the Os des Femmes de Lyon (OFELY) Study, a history of fragility fracture (OR = 3.3), low grip strength (for a value ≤0.6 bar (~60 kPa), OR = 2.0), and past falls (OR = 1.8) predicted fragility fractures in women (15). To our knowledge, no investigators have prospectively evaluated self-reported impaired balance for the prediction of future fracture risk.

Well-founded evidence shows that pharmacologic interventions can improve bone quality and reduce osteoporotic fracture risk in women with low bone density. In contrast, attempts to reduce the risk of injurious falls have received less attention. Nevertheless, efforts to improve balance with exercise programs have been successful (24–28). Today, fall-preventive interventions that are likely to be effective are available (4, 29, 30). These include muscle strengthening and balance training, home hazard modification, and withdrawal of medications that increase the risk of falls, including psychotropic drugs. There are indications that nursing home patients with falling tendencies may reduce their likelihood of hip fracture by wearing hip pads, although lack of compliance is a shortcoming with present devices (26, 31, 32). These previous results, together with our present findings, imply that interventional efforts to improve balance and reduce the likelihood or impact of injurious falls (4) could be specifically directed towards elderly persons who state that they have impaired balance.

The strengths of our study were the prospective twin cohort design, the inclusion of only first fractures in the analysis, identification of fractures by means of a complete national inpatient register, and the possibility of adjusting our estimates for several covariates. The fractures and other diseases were identified from registers using the unique personal registration number assigned to every resident of Sweden (33). We were also able to confirm an approximately doubled hip fracture risk with previously recognized lifestyle factors, such as current smoking, low physical activity, and cortisone use.

Our study had several conceivable limitations. Even though our study base was large, the matched analytical approach included only balance-discordant twins with few fractures, which produced broad confidence intervals. This was an even greater problem in the subgroup analyses. We regard this type of analysis as preferential, since it reduces the likelihood of bias and residual confounding in comparison with an ordinary nonmatched analysis. If there is high heritability of impaired balance, with a common tendency to incur fracture, a twin design can underestimate (especially in monozygotic twins) the risk of fracture due to balance. However, only approximately one-fourth of the variance in self-reported impaired balance was explained by genetic factors (34). We were not able to control our estimates for bone mineral density, which is often included in fracture prediction models. Nonetheless, we did adjust our estimates for several variables known to affect bone mineral density, most importantly age (included as part of the design) and body size (35). The association between body mass index and impaired balance was weak: R2 < 0.002. The additionally weak associations between other regular risk factors for osteoporotic fracture, such as physical activity, smoking, and cortisone use, and impaired balance can explain why our estimates were not attenuated in the multivariable model. We excluded from the analyses non-fall-related fracture cases and fracture cases who had experienced a similar fracture before study entry; that is, only first fractures were analyzed. Nevertheless, when these cases were retained in the analyses, our results were only slightly changed (data not shown).

We conclude that elderly persons who state that their balance is impaired have a substantially increased risk of osteoporotic fracture, independently of previously recognized risk factors. Simply asking a patient about his or her balance is a straightforward, simple tool for fracture risk assessment.


   ACKNOWLEDGMENTS
 
Author affiliations: Department of Surgical Sciences, Section of Orthopaedics, University Hospital, Uppsala, Sweden (Helene Wagner, Karl Michaëlsson); Department of Medical Sciences, Section of Clinical Pharmacology, University Hospital, Uppsala, Sweden (Håkan Melhus); Uppsala Clinical Research Center, University Hospital, Uppsala, Sweden (Håkan Melhus, Rolf Gedeborg, Karl Michaëlsson); Department of Surgical Sciences, Section of Anaesthesiology and Intensive Care, University Hospital, Uppsala, Sweden (Rolf Gedeborg); and Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden (Nancy L. Pedersen).

This study was supported by the Swedish Research Council (grants K2006-73P-20267-01-4 and K2006-73x-13511-07-3 to Karl Michaëlsson) and the US National Institutes of Health (grant AG 08724 to Margaret Gatz).

Conflict of interest: none declared.


   NOTES
 
Editor's note: An invited commentary on this article appears on page 150.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Kanis JA, Oden A, Johnell O, et al. The use of clinical risk factors enhances the performance of BMD in the prediction of hip and osteoporotic fractures in men and women. Osteoporos Int (2007) 18(8):1033–1046.[CrossRef][Web of Science][Medline]
  2. Ralston SH. Genetics of osteoporosis. Proc Nutr Soc. (2007) 66(2):158–165.[Medline]
  3. Kannus P, Uusi-Rasi K, Palvanen M, et al. Non-pharmacological means to prevent fractures among older adults. Ann Med (2005) 37(4):303–310.[CrossRef][Web of Science][Medline]
  4. Järvinen TL, Sievänen H, Khan KM, et al. Shifting the focus in fracture prevention from osteoporosis to falls. BMJ (2008) 336(7636):124–126.[Free Full Text]
  5. Austin N, Devine A, Dick I, et al. Fear of falling in older women: a longitudinal study of incidence, persistence, and predictors. J Am Geriatr Soc. (2007) 55(10):1598–1603.[Medline]
  6. Chu LW, Chiu AY, Chi I. Impact of falls on the balance, gait, and activities of daily living functioning in community-dwelling Chinese older adults. J Gerontol A Biol Sci Med Sci. (2006) 61(4):399–404.[Abstract/Free Full Text]
  7. Baloh RW, Spain S, Socotch TM, et al. Posturography and balance problems in older people. J Am Geriatr Soc. (1995) 43(6):638–644.[Web of Science][Medline]
  8. Vellas BJ, Wayne SJ, Romero LJ, et al. Fear of falling and restriction of mobility in elderly fallers. Age Ageing (1997) 26(3):189–193.[Abstract/Free Full Text]
  9. Alexander NB, Guire KE, Thelen DG, et al. Self-reported walking ability predicts functional mobility performance in frail older adults. J Am Geriatr Soc. (2000) 48(11):1408–1413.[Web of Science][Medline]
  10. Rathouz PJ, Kasper JD, Zeger SL, et al. Short-term consistency in self-reported physical functioning among elderly women: The Women's Health and Aging Study. Am J Epidemiol (1998) 147(8):764–773.[Abstract/Free Full Text]
  11. Guralnik JM, Simonsick EM, Ferrucci L, et al. A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol (1994) 49(2):M85–M94.[Abstract]
  12. Ferrer M, Lamarca R, Orfila F, et al. Comparison of performance-based and self-rated functional capacity in Spanish elderly. Am J Epidemiol (1999) 149(3):228–235.[Abstract/Free Full Text]
  13. Dargent-Molina P, Favier F, Grandjean H, et al. Fall-related factors and risk of hip fracture: the EPIDOS prospective study. Lancet (1996) 348(9021):145–149.[CrossRef][Web of Science][Medline]
  14. Cummings SR, Nevitt MC, Browner WS, et al. Risk factors for hip fracture in white women. Study of Osteoporotic Fractures Research Group. N Engl J Med (1995) 332(12):767–773.[Abstract/Free Full Text]
  15. Albrand G, Munoz F, Sornay-Rendu E, et al. Independent predictors of all osteoporosis-related fractures in healthy postmenopausal women: The OFELY Study. Bone (2003) 32(1):78–85.
  16. Nguyen ND, Pongchaiyakul C, Center JR, et al. Identification of high-risk individuals for hip fracture: a 14-year prospective study. J Bone Miner Res. (2005) 20(11):1921–1928.[CrossRef][Web of Science][Medline]
  17. Pedersen NL, Lichtenstein P, Svedberg P. The Swedish Twin Registry in the third millennium. Twin Res. (2002) 5(5):427–432.[CrossRef][Web of Science][Medline]
  18. Lichtenstein P, De Faire U, Floderus B, et al. The Swedish Twin Registry: a unique resource for clinical, epidemiological and genetic studies. J Intern Med (2002) 252(3):184–205.[CrossRef][Web of Science][Medline]
  19. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics (1977) 33(1):159–174.[CrossRef][Web of Science][Medline]
  20. Michaëlsson K, Melhus H, Ferm H, et al. Genetic liability to fractures in the elderly. Arch Intern Med (2005) 165(16):1825–1830.[Abstract/Free Full Text]
  21. Seeley DG, Browner WS, Nevitt MC, et al. Which fractures are associated with low appendicular bone mass in elderly women? The Study of Osteoporotic Fractures Research Group. Ann Intern Med (1991) 115(11):837–842.[Abstract/Free Full Text]
  22. Michaëlsson K, Olofsson H, Jensevik K, et al. Leisure physical activity and the risk of fracture in men [electronic article]. PLoS Med (2007) 4(6):e199.[CrossRef][Medline]
  23. Nguyen ND, Frost SA, Center JR, et al. Development of a nomogram for individualizing hip fracture risk in men and women. Osteoporos Int (2007) 18(8):1109–1117.[Medline]
  24. Fiatarone MA, O'Neill EF, Ryan ND, et al. Exercise training and nutritional supplementation for physical frailty in very elderly people. N Engl J Med (1994) 330(25):1769–1775.[Abstract/Free Full Text]
  25. Province MA, Hadley EC, Hornbrook MC, et al. The effects of exercise on falls in elderly patients. A preplanned meta-analysis of the FICSIT Trials. JAMA (1995) 273(17):1341–1347.[Abstract/Free Full Text]
  26. Sinaki M. Falls, fractures, and hip pads. Curr Osteoporos Rep (2004) 2(4):131–137.[CrossRef][Medline]
  27. Howe TE, Rochester L, Jackson A, et al. Exercise for improving balance in older people [electronic article]. Cochrane Database Syst Rev. (2007) (4):CD004963.
  28. Kita K, Hujino K, Nasu T, et al. A simple protocol for preventing falls and fractures in elderly individuals with musculoskeletal disease. Osteoporos Int (2007) 18(5):611–619.[Medline]
  29. Gillespie LD, Gillespie WJ, Robertson MC, et al. Interventions for preventing falls in elderly people [electronic article]. Cochrane Database Syst Rev. (2003) (4):CD000340.
  30. Voukelatos A, Cumming RG, Lord SR, et al. A randomized, controlled trial of tai chi for the prevention of falls: The Central Sydney Tai Chi Trial. J Am Geriatr Soc. (2007) 55(8):1185–1191.[CrossRef][Web of Science][Medline]
  31. Kannus P, Parkkari J. Prevention of hip fracture with hip protectors. Age Ageing (2006) 35(suppl 2):ii51–ii54.[Abstract/Free Full Text]
  32. Sawka AM, Boulos P, Beattie K, et al. Hip protectors decrease hip fracture risk in elderly nursing home residents: a Bayesian meta-analysis. J Clin Epidemiol (2007) 60(4):336–344.[CrossRef][Web of Science][Medline]
  33. Calltorp J, Adami HO, Astrom H, et al. Country profile: Sweden. Lancet (1996) 347(9001):587–594.[Web of Science][Medline]
  34. Wagner H, Melhus H, Pedersen NL, Michaëlsson K. Heritability of impaired balance: a nationwide cohort study in twins. Osteoporos Int. (2008) Sep 19 [Epub ahead of print].
  35. De Laet C, Kanis JA, Odén A, et al. Body mass index as a predictor of fracture risk: a meta-analysis. Osteoporos Int (2005) 16(11):1330–1338.[CrossRef][Web of Science][Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?

Related articles in Am. J. Epidemiol.:

Invited Commentary: A Fine Balance—Weighing Risk Factors Against Risk
Michael G. Walsh
Am. J. Epidemiol. 2009 169: 150-152. [Abstract] [FREE Full Text]  



This article has been cited by other articles:


Home page
 JWatch Women's HealthHome page
Assessing Fracture Risk Might Be as Easy as Asking About Balance
Journal Watch Women's Health, February 19, 2009; 2009(219): 3 - 3.
[Full Text]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
169/2/143    most recent
kwn379v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Related articles in Am. J. Epidemiol.
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Wagner, H.
Right arrow Articles by Michaëlsson, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Wagner, H.
Right arrow Articles by Michaëlsson, K.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?