Osteoporos Int (2006) 17: 143–149 DOI 10.1007/s00198-005-1947-8 O R I GI N A L A R T IC L E Risk factors for fracture of the shafts of the tibia and fibula in older individuals Jennifer L. Kelsey Æ Theresa H.M. Keegan Mila M. Prill Æ Charles P. Quesenberry Jr. Stephen Sidney Received: 13 October 2004 / Accepted: 10 May 2005 / Published online: 9 August 2005 International Osteoporosis Foundation and National Osteoporosis Foundation 2005 Abstract A case-control study to identify risk factors for fracture of the shafts of the tibia and fibula among persons 45 years of age and older was undertaken in five Northern California Kaiser Permanente Medical Centers during 1996–2001. One hundred seventy-nine cases of newly diagnosed fracture of the tibia/fibula shaft and 2,399 controls sampled from the membership lists of the same five medical centers were included. Information on potential risk factors was obtained by a standardized questionnaire administered by trained interviewers. The number of previous fractures was associated with an increased risk [adjusted odds ratio (OR) (95% confidence interval) =1.49 (1.09–2.03) per previous fracture]. Attributes known or thought to be associated with protection against loss of bone mass, including high body mass index [adjusted OR =0.82 (0.69–0.97) per 5 kg/m2 increase], having ever used thiazide diuretics or water pills for at least 1 year [adjusted OR =0.62 (0.38– 1.02)], and current use of menopausal hormone therapy among females [adjusted OR =0.84 (0.53–1.32)] tended to show decreased risks. Factors generally associated with lower bone mass, such as current cigarette smoking [OR =1.55 (1.01–2.39)] and, to some extent, lack of physical activity [OR =1.31 (0.87–1.96) for the lowest quartile compared to the upper three quartiles], tended to demonstrate increased risks. The number of falls in the past year and risk factors for falls were not associJ.L. Kelsey Æ T.H.M. Keegan Æ M.M. Prill Department of Health Research and Policy, Stanford University School of Medicine, Stanford, CA, USA J.L. Kelsey (&) Division of Preventive and Behavioral Medicine, University of Massachusetts Medical School, 55 Lake Avenue North, Shaw Building, Worcester, MA 01655, USA E-mail: Jennifer.Kelsey@umassmed.edu Tel.: +1-508-8568549 Fax: +1-508-8562022 C.P. Quesenberry Jr. Æ S. Sidney Division of Research, Kaiser Permanente Medical Care Program, Oakland, CA, USA ated with tibia/fibula shaft fractures, and indicators of health status were weakly and inconsistently associated with risk. Thus, this study suggests that risk factors for low bone mass, but not health status or risk factors for falls, may be important in the etiology of fracture of the shaft of the tibia/fibula in older individuals. Keywords Epidemiology Æ Fibula shaft fracture Æ Osteoporosis Æ Risk factors Æ Tibia shaft fracture Introduction Fractures of the shafts of the tibia and fibula in older people are not common, but can be quite disabling when they do occur. Based on data from the U.S. Medicare population, Barrett et al. [1] estimated that among persons at an age of 65 years, 1.3% of White females, 0.5% of White males and 0.9% of Black females will fracture the shaft or proximal end of the tibia or fibula by the time they reach age 85. Among older people, females are at greater risk than males, and Whites are at slightly higher risk than Blacks [1, 2, 3, 4, 5, 6, 7]. Most studies show that incidence rates in females increase somewhat with age starting at around 55–65 years, but little increase with age is seen for males at these older ages [2, 3, 6]. Across all ages, on average these fractures require a stay in the hospital of more than a week [8]. Seeley et al. [9] have shown that in women of age 65 years and older, ‘‘leg’’ fractures (not including the hip, foot, toe, patella, and ankle) are associated with low bone mineral density and are usually caused by minimal or moderate trauma. Honkanen et al. [10] found that risk for fracture of the tibia (excluding the ankle) was elevated in lactose-intolerant women of ages 38– 57 years, although it was unclear whether this was attributable to the lower calcium consumption of the lactose intolerant women. Otherwise, little is known of risk factors for fracture of the shaft of the tibia/fibula in 144 older individuals. In this paper, we report the results of a case-control study to identify risk factors for fracture of the shaft of the tibia/fibula. Subjects and Methods This case-control study of fracture of the shaft of the tibia/fibula was part of a larger case-control study to identify risk factors for fractures of five sites: distal forearm, foot, proximal humerus, pelvis and shaft of the tibia/fibula. Cases and controls were identified during the period October 1996 to May 2001 from five Northern California Kaiser Permanente Medical Centers: Hayward, Oakland, San Francisco, Santa Clara and South San Francisco. Details of the study design have been described previously [11, 12] and will be presented relatively briefly here. The study was approved by the Institutional Review Boards of the Kaiser Permanente Division of Research and the Stanford University School of Medicine. Cases All fractures of the shafts of the tibia and/or fibula in females and males of 45 years of age and older at the time the fracture was diagnosed were identified each week from computerized radiology reports and medical records by a trained medical record abstractor. The fracture had to be confirmed by X-ray, bone scan or magnetic resonance imaging to be included. Persons with previous fracture of the shaft of the tibia/fibula at age 45 years and older were excluded from these analyses. Pathologic fractures resulting from diseases such as Paget’s disease and cancer were also excluded. If a person had simultaneous fractures of multiple sites in addition to the tibia/fibula shaft, only cases in which the fracture of the tibia/fibula shaft was listed first on the medical record among the five fracture sites considered in this study were included in these analyses. Participation among eligible cases was 77.6%. Controls Controls were randomly selected within gender and age groups from the membership lists of the same five Kaiser Permanente medical centers over the same period of time. Because the same controls were used for each of the five fracture sites included in the overall study, controls were not specifically matched to cases with tibia/fibula shaft fracture, but rather were chosen so that there would be sufficient numbers across the age range of the study. Every 3 months the computerized Kaiser membership lists were stratified by gender and nine 5year age groups. Within each gender and age group the members were randomly ordered and the first 34 females and 7 males selected. All who belonged to a minority group or were of unknown race/ethnicity and 39% of white females and 78% of white males were randomly chosen. Controls with a previous fracture of the shaft of the tibia/fibula since age 45 years of older were excluded from these analyses. The participation among the selected and eligible controls was 65.5%. Collection of data on possible risk factors Most information on possible risk factors was obtained using a standardized questionnaire administered by trained interviewers in English or Spanish. For the first 3 years of the study, most of the interviews were inperson; after 15 November 2000, most interviews were conducted over the telephone in order to enhance participation and to increase numbers. No evidence was found of effect modification by mode of interview; that is, the associations between potential risk factors and tibia/fibula shaft fracture were similar in those interviewed in-person (112 cases and 999 controls) and by telephone (67 cases and 1,400 controls). The mode of interview is included as a covariate in the analyses. Data from proxy respondents (2 cases and 75 controls) are included, but could not be considered as a separate stratum because of the small numbers of cases. The questionnaire covered the period of time before the fracture for cases and before the interview for controls. Areas covered in the questionnaire were demographic characteristics; weight and height (used to compute body mass index: weight in kilograms divided by height in meters squared); handedness; a family history of hip fracture; a history of practitioner-diagnosed medical conditions (diabetes; angina, a heart attack or heart failure; stroke or a blood clot in the brain; epilepsy, seizures, convulsions, or fits; kidney disease; cataracts; glaucoma; Parkinson’s disease; arthritis; depression; cancer; hyperthyroidism; hypothyroidism); self-reported foot problems; a history of certain neuromuscular symptoms in the past year; a history of fractures; a history of use of selected medications (thiazide diuretics, water pills, Tums, other calcium supplements, multivitamins, melatonin, steroid pills, and seizure medications) at least once per week for at least 1 year, and recent use of medications to help sleep, calm nerves or lift mood. Detailed questions, taken from the Women’s Health Initiative [13], were asked of women about the use of menopausal hormone therapy for at least 3 consecutive months, and about age at last menstrual period and prior hysterectomy and/or oophorectomy [14]. Overall health status compared to others of similar age was queried. Measurement of physical functioning covered ability in the past month to do heavy housework; walk up and down stairs; walk half a mile without help; pull or push objects; stoop, crouch or kneel; lift 10 lbs; extend the arms above shoulder level; and write or handle small objects. Ability to perform activities of daily living in the 145 past month covered using the telephone, getting groceries, getting to places outside of walking distance, preparing meals, doing chores around the house, taking medications, and handling finances. Leisure-time physical activity was assessed by questioning participants about the frequency and duration of walking/hiking, gardening, exercise classes, swimming, bicycling, tennis, calisthenics/weight training, social dancing, jogging, bowling, golfing, stretching exercises or yoga, Tai Chi, and heavy housework in the past year. Questions modified from the Physical Activity History questionnaire [15] were used. The data obtained were converted into metabolic equivalent hours of exercise per month [16]. A calcium-validated food frequency questionnaire [17], to which a few ethnic foods were added, was used to assess dietary calcium intake. Data on cigarette smoking, alcohol consumption, a history of falling, and the immediate cause and circumstances surrounding the fracture were also obtained. Vision was assessed by asking participants whether they could see well enough to recognize a friend across the room and by a short test of visual acuity. Hearing loss was assessed by two questions about use of a hearing aid. Ten percent of participants agreed to a slightly abbreviated interview that did not include questions on dietary calcium intake, physical activity, cigarette smoking, and part of the medication history. Because of the reduction in sample size for these variables, analyses involving them are based on somewhat smaller numbers than other analyses. Statistical analysis Data were analyzed using SAS version 8.2 software. The odds ratio, which approximates the relative risk, was used as a measure of the magnitude of the association between putative risk factors and fracture of the shaft of the tibia/fibula. All odds ratios were adjusted by unconditional logistic regression for the sampling variables 5-year age group, gender, and race/ethnicity as recorded in Kaiser records (white versus nonwhite or unknown), and whether the interview was conducted in person or by telephone. Age in years and self-reported race/ethnicity (White, Native American or other; Asian or Pacific Islander; Black; Hispanic) were also included to control more tightly for these variables. Unconditional logistic regression was also used to control for several other variables associated with fracture of the shaft of the tibia/fibula. When odds ratios are shown according to quartiles, the distribution of the variable in the control group was used to create the quartile boundaries. Effect modification by gender, age, race/ethnicity, and selected other variables was assessed both by visual inspection and by including cross-product terms in logistic regression models. Because of the relatively small number of cases, we had little power to detect effect modification, but little evidence of it was apparent. Consequently, all subgroups are combined in the analyses. Results One hundred seventy-nine cases and 2,399 controls were included (Table 1). The number of female cases (n =103) exceeded that of male cases (n =76). About 90% of the cases were younger than 75 years of age. There were fewer members of minority racial/ethnic groups among the cases than controls, but in this study minority controls were oversampled, whereas all cases of tibia/ fibula fracture were included. Seventy-three percent of tibia/fibula shaft fractures occurred as a result of a fall. Eighteen percent were attributed to motor vehicle accidents, 5% to recreational accidents, 3% to the bone just breaking, and 1% to other causes. Little difference was seen between odds Table 1 Number (and percentage) of cases (n =179) and controls (n =2,399) by gender, age and race/ethnicity Characteristic Females Males Cases Age (years) 45–54 55–64 65-74 75–84 85+ Total Race/ethnicity White, Native American and Other Asian/Pacific Islander Black Hispanic 30 37 24 8 4 103 Controls (29.1) (35.9) (23.3) (7.8) (3.9) (100.0) Cases (females and males) 118 10 24 27 (65.9) (5.6) (13.4) (15.1) 513 456 451 363 102 1885 Cases (27.2) (24.2) (23.9) (19.3) (5.4) (100.0) Controls (females and males) 1,268 480 415 236 (52.9) (20.0) (17.3) (9.8) 34 25 11 6 0 76 Controls (44.7) (32.9) (14.5) (7.9) (0.0) (100.0) 120 130 119 107 38 514 (23.4) (25.3) (23.1) (20.8) (7.4) (100.0) 146 ratios computed with and without the inclusion of fractures in which the immediate cause was a multiple vehicle accident. Therefore, all cases were included in subsequent analyses in order to maximize sample size. Table 2 presents the prevalence in the control group of several characteristics of interest, and Table 3 shows odds ratios for the relation between these characteristics and tibia/fibula shaft fracture, first adjusted for the design variables only (third column) and then adjusted for the design variables and all the other variables in the table (fourth column). Among factors that may be in part markers of low bone mass or to affect fracture risk through their effect on bone mass, the number of previous fractures since age 45 was associated with increased risk. High body mass index, use of thiazide diuretics or water pills, and current use of menopausal hormone therapy, generally thought to be protective against loss of bone mass, were associated with decreased risks, although the use of menopausal hormone Table 2 Percent prevalence of selected characteristics among 2,399 controls Characteristic Percent prevalence At least one fracture since age 45 Body mass index (kg/m2) ‡25.0 Ever used thiazide diuretics/water pills Current use of menopausal hormone therapy (females only) a Current smokerb Leisure-time physical activity <18 MET hours per month in past yearb Dietary calcium <800 mg/day in past yearb Ever used calcium supplements for ‡1 year Fell at least twice in past year Self-reported poor or fair health status Has at least one neuromuscular symptomc Has difficulty performing one or more physical functionsd Needs help or unable to do at least one activity of daily livinge Self-reported history of one or more practitioner-diagnosed diseasesf 5.6 53.6 20.2 21.3 11.6 25.1 60.6 32.1 11.4 16.5 59.5 37.3 41.1 66.0 therapy was based on only the 103 women and did not reach statistical significance. Current cigarette smoking and to some extent lack of leisure-time physical activity, thought to be associated with low bone mass, were associated with increased risks. Higher dietary calcium intake and use of calcium supplements, which would be expected to be associated with higher bone mass, did not have much effect on risk. The number of falls in the past year was not associated with tibia/fibula shaft fracture. Factors thought to increase the risk for falls, such as vision problems, hearing problems, and epilepsy or use of seizure medication, were also not associated with risk (data not shown). No clear picture emerged of the relation between general health status and risk for fracture of the tibia/ fibula shaft. Self-reported fair or poor health, number of neuromuscular symptoms reported, and difficulty with or inability to perform various physical functions were associated with somewhat decreased risks, although the association for physical functioning was not as strong in the multivariate-adjusted analysis as in the analysis adjusted only for design variables. In contrast, needing help or inability to perform activities of daily living was associated with a slightly increased risk, an association that was somewhat stronger in the multivariate-adjusted analysis than in the analysis adjusted only for design variables. As noted above, lack of physical activity, which may be an indicator of poorer health, was also associated with a slightly increased risk. Number of diseases reported was not associated with risk, and none of the individual diseases queried was associated with an elevation or reduction in risk (data not shown). Use of physical aids was also not associated with tibia/fibula shaft fracture (data not shown). Other variables, including height, vision problems, hearing problems, foot problems, other lower extremity problems, maternal history of hip fracture, oophorectomy or hysterectomy, alcohol consumption, and use of any of the other medications queried, were not associated with risk. a Current use refers to use in the month before fracture for cases and before interview for controls and continuously used for at least 3 months. bExcludes 16 cases and 265 controls who were given abbreviated interviews. See Methods for explanation of abbreviated interview. cNeuromuscular symptoms include numbness or weakness in the feet or legs, problems with balance or unsteadiness, limping, tremors or shakes, dizziness or lightheadedness, difficulty walking in dim light and pain, numbness, burning or tingling in the legs or feet when not walking. dHas some or a lot of difficulty or unable to perform one or more physical functions including heavy housework, walking up and down stairs, walking half a mile without help, pulling or pushing objects, stooping, crouching or kneeling, lifting 10 lbs, extending arm above shoulder level, writing or handling small objects. eActivities of daily living include using the telephone, getting groceries, getting to places beyond walking distance, preparing meals, doing chores around the house, taking medications and handling finances. fDiseases queried were diabetes, angina, a heart attack or heart failure, stroke or a blood clot in the brain, epilepsy, seizures, convulsions or fits, kidney disease, cataracts, glaucoma, Parkinson’s disease, arthritis, depression, cancer, hyperthyroidism and hypothyroidism Discussion These results suggest that among older individuals, fracture of the shaft of the tibia/fibula may tend to occur among those who have low bone mass, but who do not show any particular propensity to fall, and whose health status, on balance, is not much different from people who do not fracture. Our finding that indicators of low bone mass may be associated with an increased risk is consistent with the report of Seeley et al. [9] of a strong correlation between measured bone mineral density and risk for fractures of the lower leg exclusive of the hip, foot, toe, patella, and ankle. A history of fractures since age 45 years, which, it should be noted, has been found to be associated with a risk for fractures above and beyond its association with Yes/no, one or more physical functionsh Per activity Per disease Yes/no Yes/no Yes/no Lowest quartile vs. all othersf Per 500 mg/day Yes/no Per fall Fair/poor vs. good/excellent Per symptom 1.49 (1.09–2.03) 1.53 (0.80–2.92) 1.94 (0.72–5.23) 0.82 (0.69–0.97) 0.80 (0.56–1.16) 0.67 (0.43–1.05) 0.62 (0.38–1.02) 0.84 (0.53–1.32) 1.55 (1.01–2.39) e 1.31 (0.87–1.96) e 1.04 (0.84–1.28) e 0.86 (0.59–1.26) 1.01 (0.98–1.03) 0.70 (0.43–1.13) 0.86 (0.77–0.96) 0.90 (0.61–1.34) 0.59 (0.40–0.88) 0.71 (0.49–1.04) 1.17 (1.00–1.37) 1.02 (0.89–1.17) Adjusted (179 cases, 2,399 controls) (0.38–1.14) (0.49–1.37) (0.99–2.45) (0.91–2.23) (0.85–1.31) (0.60–1.39) (0.99–1.04) (0.36–1.20) (0.75–1.01) 0.83 (0.52–1.34) 1.29 (1.04–1.61) 1.06 (0.89–1.26) 0.66 0.82 1.56 1.42 1.05 0.91 1.01 0.66 0.87 0.81 (0.66–0.99) 1.53 (1.09–2.15) Multivariate-adjusted (154 cases, 1,997 controls) Odds ratio (95% confidence interval) c Adjusted for gender, 5-year age group, Kaiser-reported race/ethnicity, in-person vs. telephone interview, age in years and self-reported race-ethnicity. bAdjusted for gender, 5-year age group, Kaiser-reported race/ethnicity, respondent status (self vs. proxy), in-person vs. telephone interview, age in years, self-reported race/ethnicity and all the other variables in the table. cExcludes cases and controls with missing values for the variables included in the tables. See Methods for explanation of abbreviated interviews, which account for most of the missing values. dCurrent use refers to use in the month before fracture for cases and before interview for controls and continuously used for at least 3 months. eExcludes 16 cases and 265 controls who were given abbreviated interviews. See Methods for explanation of abbreviated interview. fLowest quartile: <18 MET hours per month of leisure-time physical activity. gNeuromuscular symptoms include numbness or weakness in the feet or legs, problems with balance or unsteadiness, limping, tremors or shakes, dizziness or lightheadedness, difficulty walking in dim light and pain, numbness, burning or tingling in the legs or feet when not walking. hHas some or a lot of difficulty or unable to perform one or more physical functions including heavy housework, walking up and down stairs, walking half a mile without help, pulling or pushing objects, stooping, crouching, or kneeling, lifting 10 lbs, extending arm above shoulder level and writing or handling small objects. iActivities of daily living include using the telephone, getting groceries, getting to places beyond walking distance, preparing meals, doing chores around the house, taking medications and handling finances. jDiseases queried were diabetes mellitus, angina, a heart attack or heart failure, stroke or a blood clot in the brain, epilepsy, seizures, convulsions or fits, kidney disease, cataracts, glaucoma, Parkinson’s disease, arthritis, depression, cancer, hyperthyroidism and hypothyroidism a Per fracture Number fractures since age 45 1 fracture versus none 2+ fractures versus none Body mass index (BMI) (kg/m2) BMI 25.0–29.9 versus <25.0 BMI ‡30.0 versus <25.0 Ever thiazide diuretic/water pill use for ‡1 year Current use of menopausal hormone therapyd Current smoker Leisure-time physical activity (MET hours per month in past year) Dietary calcium intake (mg per day, past year) Ever used calcium supplements for ‡1 year Number of falls in past year Self-reported health status Number neuromuscular symptomsg 1 symptom versus none 2+ symptoms versus none Has difficulty performing Number of activities of daily living with which needs help or unable to doi Number of diseases reportedj Per 5 units Unit Variable Table 3 Adjusteda and multivariate-adjustedb odds ratios (and 95% confidence intervals) for the associations between selected variables and fracture of the shaft of the tibia/fibula 147 148 low bone mineral density [18], was associated with an increased risk in this study. A protective effect of greater body mass index, as suggested by these data, is probably attributable to the higher endogenous estrogen and testosterone concentrations in obese individuals and to the greater mechanical stress on weight-bearing bones [19]. The mechanism for a possible protective effect of thiazide diuretics, if such an effect exists for fractures of the shaft of the tibia/fibula, is not known with certainty, but is believed to be related to lower calcium excretion [19]. Current use of menopausal hormone therapy is known to protect against loss of bone mass and fracture of most sites [19, 20], but the number of women in this study was not large enough to show a statistically significant effect. An increased risk from cigarette smoking is consistent with the findings that smoking is associated with lower endogenous estrogen concentrations and lower bone mass [21]. Lack of physical activity may be a risk factor because of its association with lower bone mineral density [22]. It can also be considered a marker of frailty, but other indicators of frailty were not strongly associated with tibia/fibula shaft fracture in this study. The lack of association in our study with previous falls and with fall-related risk factors finds some support from the European Vertebral Osteoporosis Study. Among the 30 centers included in that study, a moderate correlation was noted between the incidence of falls and the incidence of upper limb fractures, but little correlation was found between the incidence of falls and the incidence of lower limb fractures [23]. Although a propensity to fall may not be a risk factor for tibia/fibula shaft fracture, how a person falls when she/he does fall appears to be of significance. In a previous publication from this study, Keegan et al. [24] reported that among controls who had fallen at least once in the previous year and cases who attributed their fracture to a fall, tibia/fibula shaft fractures were associated with non-vigorous outside activity, falling from more than a standing height, falling sideways, straight down and particularly falling in multiple directions (such as backward and sideways) and tumbling or rolling. Wearing high-heeled shoes was associated with an increased risk, while putting out a hand or grabbing or hitting something with the body to break a fall was protective. Also, as noted above, 18% of fractures of the shaft of the tibia/fibula in this study were attributed to motor vehicle accidents. Thus, common sense measures such as avoiding situations in which one falls from a substantial height, wears high-heeled shoes, or is involved in a motor vehicle accident, as well as breaking a fall with an outstretched hand or other means, may reduce the incidence of these fractures. Breaking a fall with an outstretched hand would, however, increase the risk for a distal forearm fracture [24]. To our knowledge no previous studies have considered health status in relation to risk for fracture of the shaft of the tibia/fibula. In the present study, the associations were weak and varied according to the indicator of health status used. Although more study is needed before any conclusions are reached, health status does not appear to be a major risk factor for fracture of the shaft of the tibia/fibula. The relatively modest increase in incidence rates with age reported in other studies [2, 3, 6] also suggests that fracture of the shaft of the tibia/fibula is not mainly a fracture of frail, elderly people. There are several limitations to this study. It was a case-control study, and was therefore dependent upon participants to remember and report past events. Undoubtedly some inaccuracies in remembering and reporting occurred. However, a prospective study of a relatively uncommon occurrence such as fracture of the tibia/fibula shaft would require a very large study population. Although the 179 cases included in this study make it by far the largest study to date of risk factors for fracture of the shaft of the tibia/fibula, we had limited power to detect small associations or to examine subgroups. We did, however, have adequate power to detect moderate or strong associations. For instance, power was 92% to detect an odds ratio of 2.0 if only 10% of the controls were exposed to a factor, and 81% to detect and odds ratio as low as 1.7 for factors to which 15% of the controls were exposed. In summary, the study of Seeley et al. [9] and the present study indicate that people with low bone mass may be at increased risk for tibia/fibula shaft fractures, but otherwise these fractures may be hard to predict. Thus, on the basis of the limited available information, preventive measures include preservation of bone mass and employing general common sense to avoid hazardous situations created by falling from more than a standing height, wearing high-heeled shoes and being involved in motor vehicle accidents. Further study is needed, however, before firm conclusions can be reached. Acknowledgements This study was supported by grants from the National Institute of Arthritis and Musculoskeletal Diseases (R01 AR42421 and T32 AR07588). We thank Beverly Peters and Luisa Hamilton for project management, Michael Sorel for computing and database management and Carolyn Salazar for medical record abstraction. References 1. Barrett JA, Baron JA, Karagas MR, Beach ML (1999) Fracture risk in the US Medicare population. J Clin Epidemiol 52:243–249 2. Griffin MR, Ray WA, Fought RL, Melton LJ III (1992) Blackwhite differences in fracture rates. Am J Epidemiol 136:1378– 1385 3. Baron JA, Karagas M, Barrett J Kniffin W, Malenka D, Mayor M et al (1996) Basic epidemiology of fractures of the upper and lower limb among Americans over 65 years of age. Epidemiology 7:612–618 4. Baron JA, Barrett JA, Karagas MR (1996) The epidemiology of peripheral fractures. Bone 18:209S–213S 5. Baron YM, Brincat M, Galea R, Baron AM (1994) The epidemiology of osteoporotic fractures in a Mediterranean country. Calcif Tissue Int 54:365–369 6. Van Staa TP, Dennison EM, Leufkens HGM, Cooper C (2001) Epidemiology of fractures in England and Wales. Bone 29:517– 522 149 7. Bengner U, Ekbom T, Johnell O, Nilsson BE (1990) Incidence of femoral and tibial shaft fractures. Epidemiology 1950–1983. Acta Orthop Scand 61:251–254 8. Praemer A, Furner S, Rice DP (1992) Musculoskeletal Conditions in the United States. American Academy of Orthopedic Surgeons, Park Ridge, Ill 9. Seeley DG, Browner WS, Nevitt MC, Genant HK, Scott JC, Cumming SR (1991) Which fractures are associated with low appendicular bone mass in elderly women? Ann Intern Med 115:837–842 10. Honkanen R, Kröger H, Alhava E, Turpeinen P, Tuppurainen M, Saarikoski S (1997) Lactose intolerance associated with fractures of weight-bearing bones in Finnish women aged 38– 57 years. Bone 21:473–477 11. Keegan THM, Kelsey JL, Sidney S, Quesenberry CP Jr (2002) Foot problems as risk factors of fractures. Am J Epidemiol 155:926–931 12. Luetters CM, Kelsey JL, Keegan THM, Quesenberry CP, Sidney S (2003) Left-handedness as a risk factor for fractures. Osteoporosis Int 14:918–922 13. Writing Group for the Women’s Health Initiative Investigators (2002) Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288:321– 333 14. Keegan THM, Gopalakrishnan G, Sidney S, Quesenberry CP Jr, Kelsey JL (2003) Hormone replacement therapy and risk for foot, distal forearm, proximal humerus, and pelvis fractures. Osteoporosis Int 14:469–475 15. Sidney S, Jacobs DR Jr, Haskell W, Armstrong MA, Dimicco A, Oberman A et al (1991) Comparison of two methods of assessing physical activity in the Coronary Artery Risk Development of Young Adults (CARDIA) Study. Am J Epidemiol 133:1231–1245 16. Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ et al (2000) Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc 32:S498–504 17. Cummings SR, Block G, McHenry K, Baron RB (1987) Evaluation of two food frequency methods of measuring dietary calcium intake. Am J Epidemiol 126:796–802 18. Kanis JA, Johnell O, De Laet C, Johansson H, Oden A, Delmas P et al (2004) A meta-analysis of previous fracture and subsequent fracture risk. Bone 35:375–382 19. Cauley JA, Salamone LM (2001) Postmenopausal endogenous and exogenous hormones, degree of obesity, thiazide diuretics, and risk of osteoporosis. In: Marcus R, Feldman D, Kelsey J (eds) Osteoporosis, vol 1. Academic Press, San Diego, pp 741– 769 20. Togerson DJ, Dell-Syer SE (2001) Hormone replacement therapy and prevention of nonvertebral fractures: a metaanalysis of randomized trials. JAMA 285:2891–2897 21. Seeman E (2001) The effects of tobacco and alcohol use on bone. In: Marcus R, Feldman D, Kelsey J (eds) Osteoporosis, vol 1. Academic Press, San Diego, pp 771–794 22. Beck BR, Shaw J, Snow CM (2001) Physical activity and osteoporosis. In: Marcus R, Feldman D, Kelsey J (eds) Osteoporosis, vol 1. Academic Press, San Diego, pp 701–720 23. Roy DK, Pye SR, Lunt M, O’Neill TW, Todd C, Raspe H, Reeve J et al (2002) Falls explain between-center differences in the incidence of limb fracture across Europe. Bone 31:712–717 24. Keegan THM, Kelsey JL, King AC, Quesenberry CP JR, Sidney S (2004) Characteristics of fallers who fracture at the distal forearm, foot, proximal humerus, pelvis, and shaft of the tibia/fibula compared to fallers who do not fracture. Am J Epidemiol 159:192–203