SEARCH

SEARCH BY CITATION

Keywords:

  • risk factors;
  • hip fracture;
  • calcium intake;
  • alcohol consumption;
  • cigarette smoking;
  • physical activity

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

The objectives of the Asian Osteoporosis Study (AOS) were to determine risk factors for hip fracture in men and women in four Asian countries, that is, Singapore, Malaysia, Thailand, and the Philippines. A total of 451 men and 725 women (aged 50 years and over) with hip fractures were compared with an equal number of community controls. A standardized questionnaire was administered by interview. The following relative risks (RRs) were found in women and men by multiple logistic regression: dietary calcium intake < 498 mg/day, 2.0 for women (95% CI, 1.5–2.8) and 1.5 for men (95% CI, 1.0–2.2); no load bearing activity in the immediate past, 2.0 for women (95% CI, 1.4–2.7) and 3.4 for men (95% CI, 2.3–5.1); no vigorous sport activities in young adulthood, 7.2 for women (95% CI, 4.0–13.0) and 2.4 for men (95% CI, 1.6–3.6); cigarette smoking, 1.5 for men (95% CI, 1.0–2.1); alcohol consumption 7 days a week, 2.9 for women (95% CI, 1.0–8.6) and 1.9 for men (95% CI, 1.1–3.2); fell twice or more in the last 12 months, 3.0 for women (95% CI, 1.8–4.8) and 3.4 for men (95% CI, 1.8–6.6); a history of fractures after 50 years of age, 1.8 for women (95% CI, 1.1–2.9) and 3.0 for men (95% CI, 1.6–5.6); a history of stroke, 3.8 for women (95% CI, 2.0–7.1) and 3.6 for men (95% CI, 1.8–7.1); use of sedatives, 2.5 for women (95% CI, 1.0–6.3) and 3.0 for men (95% CI, 1.0–9.7); and use of thyroid drugs, 7.1 for women (95% CI, 2.0–25.9) and 11.8 for men (95% CI, 1.3–106.0). Women who were 1.56 m or taller had an RR of 2.0 (95% CI, 1.3–3.0) for hip fracture and men who were 1.69 m or taller had an RR of 1.9 (95% CI, 1.2–3.1) for hip fracture. Based on these findings, primary preventive programs for hip fracture could be planned in Asia.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

HIP FRACTURE is a major public health problem in Asia. It has been projected that by the next century, 50% of all hip fractures in the world will occur in Asia.(1) Knowledge on risk factors is essential for the prevention of hip fracture. In Chinese people from Hong Kong, a low dietary calcium intake and lack of physical activity were found to be associated with hip fracture.(2) In Japanese people, the risk factors for hip fracture included leanness, regular alcohol intake, and multiparity.(3)

Relatively little is known about the risk factors for hip fracture in other Asian populations. The Asian Osteoporosis Study (AOS) is the first multicenter, case-control study on the risk factor for hip fracture conducted in Asia. The role of a low dietary calcium intake, lack of physical activity, cigarette smoking, alcoholism, falls, and other risk factors in the etiology of hip fracture were studied.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

The study protocol was approved by the Human Research Ethics Committee of the Chinese University of Hong Kong. The study was conducted in Malaysia, the Philippines, Singapore, and Thailand. We strictly adhered to the ethics requirements in the Declaration of Helsinki. For each country, a city or region was selected for the recruitment of cases and controls. These regions were Kuala Lumpur in Malaysia, the National Capital Region in the Philippines, and Bangkok in Thailand. For Singapore, the whole country was involved.

Subject recruitment

Cases were identified by the surveillance of all hospitals in the participating regions in the years of 1997–1998. Men and women who were 50 years and older and who were admitted with a diagnosis of hip fracture were recruited. Subjects whose hip fractures resulted from falling from heights, traffic accidents, and other diseases (e.g., cancer) were excluded from the study. The original plan was to recruit 200 men and 200 women with hip fractures from each country. However, because of financial and time constraints, this target was not attained in some countries.

Community controls were recruited. Different sources of community controls were used in various countries. In Malaysia, subjects presenting in general practice clinics for minor ailments were used. In the Philippines, subjects seen by doctors in outreach health programs were used. In Thailand, neighborhood controls were used. In Singapore, community controls were recruited by household surveys. All community controls were community dwelling and had no history of fractures after the age of 50 years. One community control of the same sex and within 2 years of age was recruited for each case.

Risk factor assessment

A questionnaire was designed to measure the following risk factors: dietary calcium intake, physical activity in young adulthood and in the immediate past, a history of falls in the year before the fracture, a history of fracture after 50 years of age, cigarette smoking, alcohol consumption, medical disorders, use of drugs, and mental function. The questionnaire was first compiled in English and was then translated into the local languages. The questionnaires were then translated back into English to study the validity of the translation. Hip fracture cases were interviewed within 2 weeks of the fracture.

Cognitive function:

Cognitive function was measured by the Short Portable Mental Status questionnaire.(4) Ten questions were asked and a score of 1 was assigned to each correct answer. A total score of 6 and below was used to indicate cognitive impairment. Only subjects who scored 7 and above were recruited into our study.

Dietary calcium intake:

Dietary calcium intake was measured by the food frequency method. A list of 18 food items, which were the major source of calcium in the Asian diet, was compiled. The list has been used in a previous study.(2) In addition, investigators identified other food items to be used in the food frequency questionnaire for their countries. On average, five extra items were added into the food frequency questionnaire. Quartiles of dietary calcium intake were calculated for the whole study population.

Physical activity:

Physical activity in the following phases of life was measured: 18–24 years old, 25–49 years old, 50 years and older, and in the immediate past. For each of the first three periods, subjects were asked whether they had to do most of the housework (for women only), whether they had to perform heavy manual labor (such as farming, fishing, mining, and construction work), and whether they played vigorous sports regularly (defined as activities resulting in sweating).

Subjects also were asked how often they performed the following activities in the immediate past: walking to the shops, walking upstairs, walking uphill, walking with a load, productive housework, and heavy manual labor. The possible responses included: everyday, on a weekly basis, or never. This method of assessment was based on a previous study on hip fracture in Chinese people.(2)

Statistical methods

The sample size was calculated according to published formulae,(5) and was adequate to detect differences with a relatively low exposure (Table 1). Mean values were compared by the Student's t-test. Relative risk (RR) and 95% CI were calculated by logistic regression. RR was first calculated for each risk factor, while adjusting for age and center. Subsequently, risk factors that were found to be associated with hip fracture in both sexes were studied by multiple logistic regressions. A step-down method was adopted and the final model is presented in the Results section.

Table Table 1.. Detectable RR (for a value of p < 0.05 by two-sided test) According to the Prevalence of the Risk Factor Identified Among Controls
Thumbnail image of

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

For most of the questions asked, the response rate was high (>90%). The characteristics of cases and controls are summarized in Table 2. Our results referred only to cases and controls in which men and women passed the cognitive assessment and could not be extrapolated to demented subjects. The relationship between risk factors and hip fractures is described in more detail in the following sections.

Table Table 2.. Characteristics of Cases and Controls
Thumbnail image of

Height, weight, and body mass index

Body weight was not significantly associated with the risk of hip fracture. However, women with hip fracture were significantly taller than controls (p < 0.01) (Table 2). Men and women whose heights were in the highest quartile had an RR of 2 for hip fracture. The body mass index (BMI) in female cases was significantly lower than controls (p < 0.01; Table 2). However, when the BMI was analyzed by quartiles, no significant increase in risk with BMI was observed.

Physical activity

Physical activity in the immediate past was correlated with those in young adulthood. Men and women who performed vigorous sports and manual labor regularly before 50 years of age were more likely to walk uphill, walk with load, and perform manual labor daily or weekly (p < 0.05 by χ2 test for all the associations).

Men and women with hip fracture were found to perform less load-bearing activity in the immediate past than controls. The RR (and 95% CI) for men and women who never performed the following activities in the immediate past were as follows: never walked uphill (in men: RR, 1.6, 95% CI, 1.1–2.2; in women: RR, 1.8, 95% CI, 1.3–2.4); never walked with load (in men: RR, 2.4, 95% CI, 1.6–3.5; in women: RR, 2.4, 95% CI, 1.6–3.4); no manual labor (in men: RR, 1.6, 95% CI, 1.1–2.6; in women: RR, 1.0, 95% CI, 0.7–1.6); no productive housework (in men: RR, 2.1, 95% CI, 1.5–3.0; in women: RR, 2.0, 95% CI, 1.6–2.4).

The relationship between past physical activity and the risk of hip fracture also was studied in the following periods: 18–24 years old, 25–49 years old, and 50 years and older. Activity levels in the three study periods were highly correlated, and the RR for the period of 25–49 years of age is presented in Table 2. Men and women who did not perform vigorous sports or manual labor regularly and women who did not have to do most of the housework were at significantly increased risk of hip fracture. Such results seemed to suggest an association between past activity and the risk of hip fracture.

Dietary calcium intake

The mean calcium intake was lower in women with hip fracture than controls (p < 0.01; Table 2). Men and women whose dietary calcium intake was in the lowest three quartiles (i.e., lower than 498 mg) had an increase in the RR of hip fracture ranging from 1.6–3 (Table 3).

Table Table 3.. Prevalence (%) of Risk Factors in the Control Population and the RR and 95% CI of Hip Fracture, Adjusted for Age and Center
Thumbnail image of

Cigarette smoking and alcohol consumption

Current smokers seemed to have a lower risk of hip fracture (Table 3; in men: RR, 0.7, 95% CI, 0.5–1.0; in women: RR, 0.5, 95% CI, 0.3–0.7). However, ex-smokers had an increased risk of hip fracture (in men: RR, 2.1, 5% CI, .5–2.9; in women: RR, 1.4, 95% CI, 0.9–2.0). Occasional alcohol consumption was protective against hip fracture in women (RR, 0.5; 95% CI, 0.3–0.9). However, men and women who consumed alcohol 7 days a week were at a high risk of hip fracture (in men: RR, 2.0, 95% CI, 1.3–3.1; in women: RR, 2.1, 95% CI, 1.0–4.7). The RR was significantly increased for men who had been drinking for 25 years or more and for men who consumed more than 14 alcoholic drinks per week.

Falls and fracture

Men and women who had falls in the previous year were at an increased risk of fracture (Table 3). For men with two or more falls in the last year, the RR was 2.6 (95% CI, 1.6–4.4). For women with two or more falls in the last year, the RR was 3.0 (95% CI, 2.1–4.5).

Men and women who had a history of fracture after 50 years of age also were at a higher risk of hip fracture. The RR was 2.7 (95% CI, 1.6–4.6) in men and 2.3 (95% CI, 1.6–3.4) in women.

Medical disorders and drug history

Men and women with stroke (in men: RR, 4.4, 95% CI, 2.5–7.7; in women: RR, 3.9, 95% CI, 2.4–6.4) and parkinsonism (in men: RR, 9.1, 95% CI, 1.2–71.3; in women: RR, 2.6, 95% CI, 0.7–9.8) had significantly higher risk of hip fracture. There was an association between a history of stroke and a of history fracture. In men with a history of stroke, 31% had one or more falls in the last year as compared with 17% of controls. Similarly, 40% of women with stroke had one or more falls in the last year as compared with 25% of controls (p < 0.05).

As compared with stroke and parkinsonism, the relationship between heart disease, hypertension, diabetes mellitus, thyroid disorders, kidney disorders, cancer, and hip fractures were much weaker. Men and women who were on sedatives, thyroid drugs, and diuretics had a higher risk of hip fractures. The form of diuretics used was not recorded.

Reproductive factors

Women with hip fracture had an earlier menopause than controls (p < 0.05; Table 2). Women whose menopause occurred at 45 years and younger had an RR of 1.5 (95% CI, 1.1–2.0) for hip fracture. There was no significant difference in the average number of live births between cases and controls.

Among all study subjects, only 5 women had ever taken hormonal replacement therapy. Hence, the sample size was too small to study the relationship between hormonal replacement therapy and hip fracture.

Results of multiple logistic regression

The final multiple logistic regression models are presented in Table 4. The lifestyle factors that were statistically significantly associated with hip fracture were a low dietary calcium intake, a lack of regular load-bearing activity in the immediate past, no vigorous sports activity at 25–49 years, cigarette smoking (in men only), daily alcohol consumption, a history of fracture after 50 years of age, a history of falls in the year before the fracture, and a history of stroke. The use of sedatives and drugs for thyroid disorders also were associated with an increased risk of hip fracture in women. Subjects whose height was in the highest quartile had a higher risk of hip fracture. The coefficient of determination for the multiple logistic regression models was 0.22 in men and 0.20 in women.

Table Table 4.. Results of Multiple Logistic Regression (by the Step-Down Method) in Men and Women
Thumbnail image of

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

Hip fracture is a major public health problem in Asia and knowledge about risk factors is important for its prevention. Although many large-scale epidemiological studies have been conducted in white populations, knowledge about risk factors in Asians remains sparse. The AOS is the first large-scale, multicenter, case-control study on risk factors for hip fracture in Asia. Risk factors were assessed by interview, and we did not validate the reproducibility and accuracy of the information obtained separately. However, meticulous attention was paid to the standardization of the questionnaire and training of interviewers. Although all study subjects had passed mental assessment, the recall of subjects could have been affected slightly by impaired mental status during the postoperative period. However, we did not consider that our observation could be attributed largely to recall bias.

We found a low dietary calcium intake, a lack of physical activity, alcoholism, and cigarette smoking to be risk factors for hip fracture. Moreover, subjects with a history of falls and fractures were at increased risk of hip fracture. Stroke, parkinsonism, and use of sedatives and drugs for thyroid disorders predisposed to hip fracture.

Our results could be compared with other epidemiological studies conducted in white and Asian subjects. In the Mediterranean Osteoporosis Study (MEDOS) conducted in Europe, Johnell et al.(6) found late menarche, poor mental function, low BMI, lack of physical activity, low exposure to sunlight, and low consumption of calcium and tea to be significantly associated with the risk of hip fracture. In the United States, Cummings et al.(7) found the risk of hip fracture to be higher in white women with the following characteristics: had a fracture history, were tall at 25 years, rated their own health as poor, had previous hyperthyroidism, had been treated with long-acting benzodiazepines or anticonvulsant drugs, had a maternal history of hip fracture, and ingested greater amounts of anticonvulsants and caffeine. In Japan, Fujiwara et al.(3) found a low BMI, regular alcohol intake, prevalent vertebral fracture, having five or more children, a low milk intake, and later age at menarche to be associated with the risk of hip fracture.

Although our findings were largely in agreement with the previously mentioned studies, the relative importance of risk factors was different. These will be further discussed.

We have previously indicated that a low dietary calcium intake was associated with a significant increase in hip fracture risk in Chinese men and women from Hong Kong.(2) In the current study, the same observations were made in other southeast Asian countries. The RR was 2 in subjects whose calcium intake was below 500 mg/day. Moreover, the average dietary calcium intake in our study population was much lower than in white subjects, and it is likely that calcium intake has the largest effect in populations with low intakes. Holbrook et al.(8) found that women whose dietary calcium intake was below 440 mg/1000 kcal per day had an RR of 2.5 for hip fracture. In Europe, Johnell et al.(6) also found the risk of hip fracture to increase with diminishing calcium intake, in subjects whose dietary calcium intake was less than 500 mg/day. The effectiveness of calcium supplementation for the prevention of osteoporotic fractures had been reviewed by Cummings and Nevitt recently.(9) In their study, calcium supplements of 800–1200 mg were associated with RR reductions of 25–70%.(10–13) Given that the dietary calcium intake in most Asian countries is low, calcium supplement should have a considerable impact on the reduction of fracture risk.

In addition to calcium intake, physical activity was an important protective factor against hip fracture in our study. Reduction in physical activity level is the factor that can best account for the epidemic of hip fracture occurring in Chinese men and women from Hong Kong.(2) The results of previous studies in white subjects also suggested a protective effect of exercise on fractures. This included findings from the Leisure World Study,(14) the MEDOS,(6) the Study of Osteoporotic Fractures,(7) and recent studies by Gregg et al.(15) and Jaglal et al.(16) Because Asian subjects seldom perform recreational exercise, enhancing activity in everyday life might be particularly important.

Cigarette smoking and heavy alcohol consumption were found to be associated with the risk of hip fracture in our study. In a recent m-analysis by Law and Hackshaw,(17) it was estimated that one hip fracture in eight in England was attributable to smoking. Because of the small number of female smokers in Asia, the attributable fraction of smoking for hip fracture would be much smaller. For alcoholism, the RR of 2 for hip fracture in our study concurred with findings in Japanese men and women,(3) as well as in white men and women.(18–21) We also observed occasional alcohol consumption to be protective against hip fracture. In white men and women, moderate alcohol drinkers had higher bone mineral density (BMD)(22,23) but the underlying mechanisms were unknown. Because the proportion of Asian women who were alcoholic was low, the prevention of heavy alcohol consumption will have a relatively small effect on fracture risk.

In this study, a history of stroke, use of sedatives, and a history of falls were found to be important risk factors for hip fracture. Such observations are similar to the results of previous studies. In the Study of Osteoporotic Fractures (SOF),(24) the tendency and type of falls were found to be associated with the risk of hip fracture. Similarly, in the Epidémiologie de l'Osteoporose (EPIDOS),(25) fall-related factors also were found to be important determinants of hip fracture in Europeans. These observations raised the question of whether falls are preventable. Indeed, recent randomized controlled clinical trials(25–27) proved that this was so.

In our study, body weight was not associated with hip fracture, although taller subjects were at higher risk of hip fracture. The lack of association between body weight and fracture risk could be because of the relative leanness of Asian subjects. Although several studies in white subjects also had shown that taller subjects were at a higher risk of hip fracture.(6,7,26,27) Such findings have little practical applications in fracture prevention.

In the AOS, a combination of risk factors accounted for around 20% of the variance in hip fracture risk. Such associations are modest but they suggested that the hip fractures are, to a certain extent, preventable. Maintaining a high calcium intake and adequate physical activity may be particularly important in preventing hip fractures in our populations. Cigarette smoking and alcohol consumption are associated with relatively small effects on hip fracture risk. However, their control and prevention are important, for they have other adverse health effects. Last, the removal of risk factors associated with falls may contribute substantially in reducing hip fracture risk. Based on our findings, comprehensive preventive programs for osteoporosis and hip fracture could be initiated and evaluated.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References

This project was supported by MSD Pharmaceuticals. We thank Dr. Clement Tsang for coordinating the study and Ms. Vanessa Lam for preparing the manuscript.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. References
  • 1
    Cooper C, Campion G, Melton LJ 1992 Hip fractures in the elderly: A world-wide projection. Osteoporos Int 2:285289.
  • 2
    Lau EMC, Donnan SPB, Barker DJ, Cooper C 1998 Physical activity and calcium intake in fractured proximal femur in Hong Kong. BMJ 297:144153.
  • 3
    Fujiwara S, Kasagi F, Yanada M, Kodama K 1997 Risk factors for hip fracture in a Japanese cohort. J Bone Miner Res 12:9981004.
  • 4
    Pfeiffer E 1975 A short mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc 23:440.
  • 5
    Fletcher RH, Fletcher SW, Wagner EH 1996 Clinical Epidemiology: The Essentials, 3rd Ed. Williams and Wilkins, Baltimore, MD, USA, pp. 105.
  • 6
    Johnell O, Gullberg B, Kanis JA, Allander E, Elffors L, Dequeker J, Dilsen G, Gennari C, Vaz AL, Lyritis G, Mazzuoli G, Miravet L, Passeri M, Cano RP, Rapado A, Ribot C 1995 Risk factors for hip fracture in European women: The MEDOS Study. J Bone Miner Res 10:18021815.
  • 7
    Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE, Cauley J, Black D, Vogt TM 1995 Risk factors for hip fracture in white women. N Engl J Med 332:767773.
  • 8
    Holbrook TL, Barrett-Connor E, Wingard DL 1988 Dietary calcium and risk of hip fracture: 14-year prospective population study. Lancet 2:10461049.
  • 9
    Cumming RG, Nevitt MC 1997 Calcium for prevention of osteoporotic fractures in postmenopausal women. J Bone Miner Res 12:13211329.
  • 10
    Chapuy MC, Ariot ME, Duboeuf F, Brun J, Crouzer B, Arnaud S, Delmas PD, Meunier PJ 1992 Vitamin D and calcium to prevent hip fractures in elderly women. N Engl J Med 327:16371642.
  • 11
    Chevalley T, Rizzoli R, Nydegger V, Slossman D, Rapin CH, Michel JP, Vasey H, Bonjour JP 1994 Effects of calcium supplements on femoral bone mineral density and vertebral fracture rate in vitamin-D-replete elderly patients. Osteoporos Int 4:245252.
  • 12
    Reid IR, Ames RW, Evans MC, Gamble GD, Sharpe SJ 1995 Long-term effects of calcium supplementation on bone loss and fractures in postmenopausal women: A randomised controlled trial. Am J Med 98:331335.
  • 13
    Recker R, Hinders S, Davies KM, Heaney RP, Stegman MR, Lappe JM, Kimmel DB 1996 Correcting calcium nutritional deficiency prevents spine fractures in elderly women. J Bone Miner Res 11:19611966.
  • 14
    Paganin-Hill A, Chao A, Ross RK, Henderson BE 1991 Exercise and other factors in the prevention of hip fracture: The Leisure World Study. Epidemiology 2:1625.
  • 15
    Gregg EW, Cauley JA, Seeley DG, Ensrud KE, Bauer DC 1998 Physical activity and osteoporotic fracture risk in older women. Ann Intern Med 129:8188.
  • 16
    Jaglal SB, Kreiger N, Darlington G 1993 Past and recent physical activity and risk of hip fracture. Am J Epidemiol 138:107118.
  • 17
    Law MR, Hackshaw AK 1997 A meta-analysis of cigarette smoking, bone mineral density and risk of hip fracture: Recognition of a major effect. J Bone Miner Res 315:841846.
  • 18
    Nelson HD, Nevitt MC, Scott JC, Stone KL, Cummings SR 1994 Smoking, alcohol, and neuromuscular and physical function of older women. JAMA 272:18251831.
  • 19
    Johnell O, Kristenson H, Redlund-Johnell I 1985 Lower limb fractures and registration for alcoholism. Scand J Soc Med 13:9597.
  • 20
    Felson DT, Kiel DP, Anderson JJ, Kannel WB 1988 Alcohol consumption and hip fractures: The Framingham study. Am J Epidemiol 128:11021110.
  • 21
    Hernandez-Avila M, Colditz GA, Stampfer MJ, Rosner B, Speizer FE, Willett WC 1991 Caffeine, moderate alcohol intake, and risk of fractures of the hip and forearm in middle-aged women. Am J Clin Nutr 54:157163.
  • 22
    Holbrook TL, Barrett-Connor E 1993 A prospective study of alcohol consumption and bone mineral density. J Bone Miner Res 306:15061509.
  • 23
    Felson DT, Zhang YQ, Hannan MT, Kannel WB, Kiel DP 1995 Alcohol intake and bone mineral density in elderly men and women. Am J Epidemiol 142:485492.
  • 24
    Nevitt MC, Cummings SR 1997 Type of fall and risk of hip and wrist fractures: The study of osteoporotic fractures. J Am Geriatr Soc 41:12261234.
  • 25
    Dargent-Molina P, Favier F, Grandjean H, Baudoin C, Schott AM, Hausherr E, Meunier PJ, Breat G 1996 Fall-related factors and risk of hip fracture: The EPIDOS prospective study. Lancet 348:145149.
  • 26
    Tinetti ME, Baker DI, McAvav G, Claus E, Garrett P, Gottschalk M, Koch ML, Trainor K, Horwitz RI 1994 A multi factorial intervention to reduce the risk of falling among elderly people living in the community. N Engl J Med 331:821827.
  • 27
    Close J, Ellis M, Hooper R, Glucksman E, Jackson S, Swift C 1999 Prevention of falls in the elderly trial (PROFET): A randomised controlled trial. Lancet 353:9397.