SEARCH

SEARCH BY CITATION

Keywords:

  • BMD;
  • former athletes;
  • osteoporosis;
  • peak bone mass;
  • physical activity;
  • fractures

Abstract

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

Former male young athletes partially lost benefits in BMD (g/cm2) with cessation of exercise, but, despite this, had a higher BMD 4 years after cessation of career than a control group. A higher BMD might contribute to the lower incidence of fragility fractures found in former older athletes ⩾60 years of age compared with a control group.

Introduction: Physical activity increases peak bone mass and may prevent osteoporosis if a residual high BMD is retained into old age.

Materials and Methods: BMD was measured by DXA in 97 male young athletes 21.0 ± 4.5 years of age (SD) and 48 controls 22.4 ± 6.3 years of age, with measurements repeated 5 years later, when 55 of the athletes had retired from sports. In a second, older cohort, fracture incidence was recorded in 400 former older athletes and 800 controls ⩾60 years of age.

Results: At baseline, the young athletes had higher BMD than controls in total body (mean difference, 0.08 g/cm2), spine (mean difference, 0.10 g/cm2), femoral neck (mean difference, 0.13 g/cm2), and arms (mean difference, 0.05 g/cm2; all p < 0.001). During the follow-up period, the young athletes who retired lost more BMD than the still active athletes at the femoral neck (mean difference, 0.07 g/cm2; p = 0.001) and gained less BMD at the total body (mean difference, 0.03 g/cm2; p = 0.004). Nevertheless, BMD was still higher in the retired young athletes (mean difference, 0.06-0.08 g/cm2) than in the controls in the total body, femoral neck, and arms (all p < 0.05). In the older cohort, there were fewer former athletes ⩾60 of age than controls with fragility fractures (2.0% versus 4.2%; p < 0.05) and distal radius fractures (0.75% versus 2.5%; p < 0.05).

Conclusions: Although exercise-induced BMD benefits are reduced after retirement from sports, former male older athletes have fewer fractures than matched controls.


INTRODUCTION

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

PHYSICAL ACTIVITY DURING adolescence increases peak BMD, and cross-sectional studies have shown that high-level impact-loaded exercise during growth is associated with 1.0-2.0 SD higher peak BMD,(1–3) a benefit likely to more than halve the fracture risk.(4, 5) If these exercise-induced benefits are retained into old age, physical activity during the first decades of life may be recommended as a prevention strategy against osteoporosis and fractures later in life. So far, no long-term prospective bone mass data exist that follow athletes from their active career into long-term retirement. The few published cross-sectional studies that do evaluate former athletes after decades of retirement suggest that exercise-induced high BMD is maintained by 0.5-1.0 SD above the age-predicted mean in athletes retired from sport for 10–20 years, a benefit no more than one-half that observed in active athletes. Furthermore, from studies over four to five decades, it seems questionable whether any residual high BMD may persist in elderly former athletes.(6–10)

However, irrespective of the level of BMD found in retired athletes, to evaluate the hypothesis that exercise during growth reduces the clinical problem of fragility fractures, we will need to show that retired athletes have fewer fractures than controls do. Traits independent of BMD, such as skeletal architecture, bone size, balance, muscle strength, and neuromuscular proprioception, may also be affected by exercise, all of which could influence the fracture risk.(11–15)

The aim of this study was to prospectively follow the development of BMD in former young athletes during years of reduced activity and evaluate whether exercise during adolescence is associated with fewer fractures at old age.

MATERIALS AND METHODS

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

BMD (g/cm2) was measured at baseline in 65 male ice hockey players and 73 soccer players 21.0 ± 4.5 years of age, all of whom were competitive white athletes without diseases or medications known to affect the bone metabolism. At baseline, the athletes exercised for a mean of 8.2 ± 2.8 h/week, predominantly playing ice hockey and soccer and doing some impact-loaded and aerobic training. Their starting age of training was 7.7 ± 2.3 years (range, 4–16 years). Controls were 61 age-matched male individuals who exercised for a mean of 2.5 ± 2.6 h/week doing recreational exercise, predominantly playing soccer or floorball, or doing distance running and weight training. After 5 years, 146 of the original participants could be located, 55 ice hockey players, 42 soccer players, and 49 controls (73% response rate), and attended a second measurement session. Fifty-five athletes had by then retired from their active sports career (age, 22.5 ± 5.8 years; range, 16–37 years) and exercised recreationally for a mean of 3.6 ± 2.2 h/week, whereas the still active athletes and the controls exercised at the same level as at baseline (Table 1). One man among the controls who had developed anorexia nervosa during the follow-up was excluded, leaving 48 controls.

BMD (g/cm2) was measured by DXA (Lunar DPX-L software version 1.3y; Lunar, Waukesha, WI, USA) in the total body, spine, and arms by a total body scan and at the femoral neck by a hip scan. Whereas the total body, spine, and femoral neck represent weight-loaded regions, the arms represent a minor loaded region. Bone area (cm2) of the right femoral neck was taken to represent bone size. The precision of the DXA technique has previously been discussed in detail,(16, 17) and the CV for the BMD measurements was 0.4-3.0% depending on the region, and for neck area was 0.9% by our equipment.(8, 18) Furthermore, the equipment was calibrated each day during the follow-up, using a standardized phantom. Body weight was measured with an electric scale and body height, with a height meter. Questionnaires both at baseline and at follow-up recorded smoking habits, previous and present illnesses, medications, starting age of exercising, current and past exercise history, years of exercise career, type and level of activity, and years since retirement.

The second cohort included 400 former older soccer and ice hockey players, all at international or national competing level during their active sports career, now all ⩾60 years of age (mean age, 71.0 ± 6.0 years). They had started to train regularly at a mean age of 13 ± 3 years and closed their career at a mean age of 36 ± 8 years. Fracture incidence was determined through a mailed questionnaire, which was sent to former athletes all over Sweden, documenting present and past exercise level, sporting activities, years since retirement, occupation, alcohol and tobacco use, previous and current diseases, injuries during and after the career, medications, and fractures throughout life. The same questions were asked of 800 age- and gender-matched individuals, two controls matched to each former athlete by date, month, and year of birth, and with their names placed closest to those of the former athletes in the Swedish national computerized data files. The response rate was 74% for the former athletes and 64% for the controls.

Statistical analysis included ANOVA when the three groups were compared. Differences between two groups were evaluated by Student's t-test for means or Mann-Whitney's test for independent samples. Spearman's rho correlation coefficient was used to evaluate any bivariate correlation. A p value of <0.05 was considered statistically significant. Data are presented as means ± SD. The study was approved by the local ethics committees in Umeå and Malmö.

RESULTS

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

At baseline, the young athletes had a higher BMD (0.08 g/cm2 in the total body, 0.10 g/cm2 in the spine, 0.13 g/cm2 at the femoral neck, and 0.05 g/cm2 in the arms; all p < 0.001) compared with the controls, but there were no differences in body height or body weight (Table 1). There was no correlation between starting age of training and BMD at any site at baseline (β 0.04-0.15; p > 0.05). Furthermore, at baseline, there were no significant differences in duration of physical activity, BMD, height, or weight between those athletes who would continue their active sports career throughout the study and those who would retire during the study period (Table 1).

Table Table 1. Age, Antropometric Data, Physical Activity, BMD of Different Sites, and Bone Area of the Femoral Neck in What Was to be the Former Athletes, the Active Athletes, and Controls at Baseline
Thumbnail image of

Retirement from exercise led to a decrease in the discrepancy in BMD between young former athletes and controls during the follow-up of 0.02 g/cm2 in total body (nonsignificant [NS]), 0.05 g/cm2 in the spine (p < 0.05), and 0.03 g/cm2 at the femoral neck (NS), whereas the discrepancy increased by 0.02 g/cm2 in the arms (NS; Fig. 1). In the former young athletes, the changes in BMD at the arms were significantly different from the changes in the total body (p = 0.007), the spine (p < 0.001), and the femoral neck (p < 0.001; Fig. 1). Number of years since retirement correlated with the changes in the spine (r = 0.31; p = 0.03), but not with the changes in the total body (r = 0.13, p = 0.36), femoral neck (r = −0.02; p = 0.86), or arms (r = 0.07 p = 0.60).

thumbnail image

Figure Fig. 1.. Changes in BMD (g/cm2) in the young cohort of athletes and controls during the 5 years of follow-up.

Download figure to PowerPoint

In contrast, continued exercise during the follow-up, in the young athletes, increased the difference in BMD compared with the controls by 0.01 g/cm2 in total body (NS), 0.04 g/cm2 at the femoral neck (p < 0.05), and 0.04 g/cm2 in the arms (p < 0.05), whereas the discrepancy decreased by 0.03 g/cm2 in the spine (NS). Finally, the young athletes who retired from their active sports career during the study period lost more BMD at the total body (mean difference, 0.03 g/cm2; p = 0.004) and femoral neck (mean difference, 0.07 g/cm2; p = 0.001) than those athletes who continued their active sports career (Fig. 1).

Despite a higher bone loss associated with retirement, BMD at follow-up remained higher in former young athletes than in controls by 0.06 g/cm2 in the total body (p < 0.01), 0.08 g/cm2 in the femoral neck (p < 0.05), and 0.07 g/cm2 in the arms (p < 0.001). However, retired young athletes at follow-up had a lower BMD than active athletes in the femoral neck (mean difference, 0.11 g/cm2; p = 0.002; Table 2).

Table Table 2. Age, Antropometric Data, Physical Activity, BMD of Different Sites, and Neck Area in Former Athletes, Active Athletes, and Controls at Follow-up
Thumbnail image of

In the second cohort, the proportion of subjects with one or more fractures before the age of 35 years was higher among old former soccer and ice hockey players than it was among the controls (17.5% versus 12.9%; p < 0.05; Fig. 2), the implication being that the former older athletes had sustained these fractures during their active sports career. In contrast, after the age of 35 years (i.e., after retirement from sports), the prevalence of fractures was lower in the old former athletes than in the controls (8.5% versus 12.9%; p < 0.05). Similarly, at ⩾50 years of age, the proportion of fragility fractures (proximal humerus, distal radius, spine, pelvic, hip, and tibial condyles) was lower in former older athletes than in controls (2.0% versus 4.2%; p < 0.05). The only site-specific fragility fracture reaching statistically significant difference was in individuals with a distal radial fracture (0.75% for former athletes versus 2.5% for controls; p < 0.05), whereas the corresponding figures for hip fractures were 0.75% versus 1.25% (NS; Fig. 2).

thumbnail image

Figure Fig. 2.. Proportion of former older athletes and controls ⩾60 years of age with fractures.

Download figure to PowerPoint

To validate the questionnaire used in the old cohort of former athletes and controls, we identified those individuals who were born, and still lived, in the city of Malmö in southern Sweden. Because Malmö has only one emergency hospital and <3% of fracture cases attend medical facilities outside the hospital and because all radiographs taken this past century have been saved in the hospital archives, we could objectively validate the questionnaire with regard to fracture incidence in these individuals.(19) Among the 278 individuals who were identified as Malmö citizens, 35 (12.6%) stated in the questionnaire that they had sustained no fractures before age 35, although such fractures had been registered in the archives. Another 32 (11.5%) stated that they had had no fractures after age 35, despite such fractures having been registered in the archives, whereas 4 individuals (1.4%) stated that they had had no fractures after the age of 50, despite such fractures having been registered in the archives. On the other hand, 16 individuals (5.8%) stated that they had had fractures before the age of 35, when no record of such fractures could be found in the archives, and 3 individuals (0.7%) stated that they had had fractures after 35 years, despite no such fractures having been registered in the archives. All stated fractures after the age of 50 were later identified in the archives.

Anthropometrics did not differ between the old former athletes and the controls now all ⩾60 years of age; for former athletes versus controls, height was 176.4 ± 5.9 versus 176.3 ± 7.5 cm (NS), whereas weight was 81.3 ± 10.3 versus 81.5 ± 12.0 kg (NS), respectively. Lifestyle factors differed minimally; for former athletes versus controls, respectively, the figures were as follows: any foodstuff excluded from the diet, 2.0% versus 3.5% (NS); milk drinkers, 93.2% versus 92.3% (NS); cups of coffee per day, 3.5 ± 2.0 versus 3.7 ± 2.3 (NS); smokers, 12.3% versus 14.2% (NS); alcohol users, 84.0% versus 73.2% (p < 0.001); drug intake, 62.0% versus 65.9% (NS); blue-collar workers, 21.8% versus 26.6% (NS); early retirement, 51.8% versus 47.0% (NS); current level of physical activity, 4.5 ± 4.0 versus 4.7 ± 3.9 h/week (NS).

DISCUSSION

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

This study suggests that there is an association between retirement from exercise and loss of exercise-induced BMD benefits. Thus far, few prospective studies have investigated the effect of reduced training on BMD. The existing studies indicate that, after reduction in activity, there is a decrease in BMD to pretraining levels; however, published studies predominantly include elderly subjects with low exercise-induced BMD benefits and shorter follow-up periods.(20–23) In this study, the young retired athletes still had a 4–8% higher BMD than controls, despite the reduction in residual high BMD at most sites. Possible explanations for this finding may be that the athletes in this study had reached a much higher BMD during the period of high physical activity, that they had been active for a longer duration, that they had started exercising at an earlier age, and that they had exercised at a high level in comparison with the exercisers in previously cited studies.

This study could not establish whether retired athletes continue to lose the exercise-induced benefits during longer periods of retirement, so that there would exist no residual higher BMD after several decades of retirement. The correlation between years of retirement and magnitude of BMD loss would seem to suggest this. However, because, in agreement with previously published data,(20, 21) number of years in retirement was not strongly associated with changes in BMD and only significantly related to the changes of the spine, we cannot exclude the possibility that a residual high BMD will persist also in elderly ex-athletes. The fact that we found fewer former athletes than controls in the old cohort with fractures does not contradict this hypothesis.

Although BMD is the main determinant of the bone strength, other factors also affect this property, such as the architecture of the bone and the bone size. The slightly larger bone size at the femoral neck among the young athletes at baseline would, if anything, increase their bone strength compared with that of the controls. However, at follow-up, there were no significant differences in femoral neck size between the groups, thus possibly not explaining the lower fracture incidence in the cohort of old former athletes.

Furthermore, retirement from exercise may induce different BMD development in weight-loaded and unloaded skeletal regions. Kontulainen et al.,(12, 24) in a 4-year prospective study, assessed the side-to-side differences in the upper extremities of male and female racket players. The differences in BMC between the dominant and nondominant arm remained with decreased activity, suggesting that the increased BMD from training was maintained despite reduced activity. Some prospective studies even imply that, with reduced activity, unloaded skeletal regions increase in BMD, possibly because of a redistribution of BMD with reduced activity level, from weight-loaded to unloaded skeletal regions.(25) Furthermore, Vico et al.(26) measured BMD in cosmonauts before and after microgravity during spaceflights. Interestingly, BMD was lost at the weight-bearing tibia, whereas no changes were seen at the non-weight-bearing radius. In this study, no BMD of the arms was lost with reduced activity in the young former athletes we studied. Furthermore, the fact that we also found fewer old former athletes than controls with distal radius fractures does not contradict the hypothesis that the skeleton of former athletes may be regionally more resistant to fractures.

Although this study was not a blinded, randomized, controlled trail, the fact that we found fewer fractures in former athletes in old age supports the hypothesis of an association between physical activity during childhood and adolescence and fewer fragility fractures at old age. In this study, we used the clinically relevant endpoint, fractures, rather than a predefined surrogate endpoint, such as BMD. The fracture incidence, reported in the questionnaire, was also validated against radiological verified fractures in a subcohort of individuals. Thus, we found that about 10% of the fractures sustained after retirement had not been reported in the questionnaire. Furthermore, only 1.4% of fractures sustained after the age of 50, and defined, in this study, as “osteoporosis-related,” were not reported in the same questionnaire. This is far better than in previous reports, where fracture recall is described as unreliable, with an over- or under-reporting of fractures by 20–40%.(19, 27)

Up to now, no prospective randomized trials have been performed using either surrogate endpoints, such as osteoporosis, or fractures in trying to evaluate the long-term effects of exercise. Such studies will probably never be performed because of the extremely large cohorts needed when using fractures as endpoint, and a >50-year study duration needed before the research question can be answered when using such a study design. However, an alternative explanation for the lower fracture incidence in the old former athletes could never be excluded in a case control study. A genetic, inherited stronger bone or a better functioning neuromuscular system, present already before the start of the exercise and making these individuals prone to exercise, as well as differences in lifestyle factors after retirement, may also have led to fewer fractures in the old former athletes. However, because we did not find any significant differences in lifestyle factors (except for the differences in alcohol use) or diseases and because there were no anthropometric differences between old former athletes and controls, the probability of a causal relationship between exercise during adolescence and reduced fracture incidence at old age is increased.

In summary, BMD was still higher in young former athletes a mean of 4 years after cessation of active career than in controls. Furthermore, former male elderly athletes had a lower fracture risk compared with controls. Therefore, it seems that exercise during childhood and adolescence may be associated with lower risk of sustaining fragility fractures at old age.

Acknowledgements

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

This study was supported of by grants from the Länsförsäkringar Insurance Company (P4/01) and the Swedish National Center for Research in Sports (112/01).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Nordstrom P, Pettersson U, Lorentzon R 1998 Type of physical activity, muscle strength, and pubertal stage as determinants of bone mineral density and bone area in adolescent boys. J Bone Miner Res 13: 11411148.
  • 2
    Bass S, Pearce G, Young N, Seeman E 1994 Bone mass during growth: The effects of exercise. Exercise and mineral accrual. Acta Univ Carol [Med] 40: 36.
  • 3
    Sundberg M, Gardsell P, Johnell O, Karlsson MK, Ornstein E, Sandstedt B, Sernbo I 2002 Physical activity increases bone size in prepubertal boys and bone mass in prepubertal girls: A combined cross-sectional and 3-year longitudinal study. Calcif Tissue Int 71: 406415.
  • 4
    Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J, Ensrud K, Genant HK, Palermo L, Scott J, Vogt TM 1993 Bone density at various sites for prediction of hip fractures. The Study of Osteoporotic Fractures Research Group. Lancet 341: 7275.
  • 5
    Kanis JA, Johnell O, Oden A, Dawson A, De Laet C, Jonsson B 2001 Ten year probabilities of osteoporotic fractures according to BMD and diagnostic thresholds. Osteoporos Int 12: 989995.
  • 6
    Karlsson MK, Johnell O, Obrant KJ 1995 Is bone mineral density advantage maintained long-term in previous weight lifters. Calcif Tissue Int 57: 325328.
  • 7
    Karlsson MK, Johnell O, Obrant KJ 1993 Bone mineral density in weight lifters. Calcif Tissue Int 52: 212215.
  • 8
    Karlsson MK, Hasserius R, Obrant KJ 1996 Bone mineral density in athletes during and after career: A comparison between loaded and unloaded skeletal regions. Calcif Tissue Int 59: 245248.
  • 9
    Karlsson MK, Linden C, Karlsson C, Johnell O, Obrant K, Seeman E 2000 Exercise during growth and bone mineral density and fractures in old age. Lancet 355: 469470.
  • 10
    Khan KM, Green RM, Saul A, Bonnell KL, Crichton KJ, Hopper JL, Wark JD 1996 Retired elite female ballet dancers and nonathletic controls have similar bone mineral density at weightbearing sites. J Bone Miner Res 11: 15661574.
  • 11
    Jones HH, Priest JD, Hayes WC, Tichenor CC, Nagel DA 1977 Humeral hypertrophy in response to exercise. J Bone Joint Surg Am 59: 204208.
  • 12
    Kontulainen S, Kannus P, Haapasalo H, Hëinonen A, Sievanen H, Oja P, Vuori I 1999 Changes in bone mineral content with decreased training in competitive young adult tennis players and controls: A prospective 4-yr follow-up. Med Sci Sports Exerc 31: 646652.
  • 13
    Vuori I, Heinonen A, Sievanen H, Kannus P, Pasanen M, Oja P 1994 Effects of unilateral strength training and detraining on bone mineral density and content in young women: A study of mechanical loading and deloading on human bones. Calcif Tissue Int 55: 5967.
  • 14
    Haapasalo H, Kontulainen S, Sievanen H, Kannus P, Jarvinen M, Vuori I 2000 Exercise-induced bone gain is due to enlargement in bone size without a change in volumetric bone density: A peripheral quantitative computed tomography study of the upper arms of male tennis players. Bone 27: 351357.
  • 15
    Province MA, Hadley EC, Hornbrook MC, Lipsitz LA, Miller JP, Mulrow CD, Ory MG, Sattin RW, Tinetti ME, Wolf SL 1995 The effects of exercise on falls in elderly patients. A preplanned meta- analysis of the FICSIT Trials. Frailty and Injuries: Cooperative Studies of Intervention Techniques. JAMA 273: 13411347.
  • 16
    Orwoll ES, Oviatt SK, Biddle JA 1993 Precision of dual-energy x-ray absorptiometry: Development of quality controls and their application in longitudinal studies. J Bone Miner Res 8: 693699.
  • 17
    Sievänen H, Oja P, Vouri I 1992 Precision of dual-energy x-ray absorptiometry in determining bone mineral content of various skeletal sites. J Nucl Med 33: 11371142.
  • 18
    Nordstrom P, Lorentzon R 1996 Site-specific bone mass differences of the lower extremities in 17-year-old ice hockey players. Calcif Tissue Int 59: 443448.
  • 19
    Jonsson B, Gardsell P, Johnell O, Redlund-Johnell I, Sernbo I 1994 Remembering fractures: Fracture registration and proband recall in southern Sweden. J Epidemiol Community Health 48: 489490.
  • 20
    Dalsky GP, Stocke KS, Ehsani AA, Slatopolsky E, Lee WC, Birge SJ Jr 1988 Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Ann Intern Med 108: 824828.
  • 21
    Winters KM, Snow CM 2000 Detraining reverses positive effects of exercise on the musculoskeletal system in premenopausal women. J Bone Miner Res 15: 24952503.
  • 22
    Snow CM, Shaw JM, Winters KM, Witzke KA 2000 Long-term exercise using weighted vests prevents hip bone loss in postmenopausal women. J Gerontol A Biol Sci Med Sci 55: M489M491.
  • 23
    Iwamoto J, Takeda T, Ichimura S 2001 Effect of exercise training and detraining on bone mineral density in postmenopausal women with osteoporosis. J Orthop Sci 6: 128132.
  • 24
    Kontulainen S, Kannus P, Haapasalo H, Sievanen H, Pasanen M, Heinonen A, Oja P, Vuori I 2001 Good maintenance of exercise-induced bone gain with decreased training of female tennis and squash players: A prospective 5-year follow-up study of young and old starters and controls. J Bone Miner Res 16: 195201.
  • 25
    Leblanc AD, Schneider VS, Evans HJ, Engelbretson DA, Krebs JM 1990 Bone mineral loss and recovery after 17 weeks of bed rest. J Bone Miner Res 5: 843850.
  • 26
    Vico L, Collet P, Guignandon A, Lafage-Proust MH, Thomas T, Rehaillia M, Alexandre C 2000 Effects of long-term microgravity exposure on cancellous and cortical weight-bearing bones of cosmonauts. Lancet 355: 16071611.
  • 27
    Akesson K, Gardsell P, Sernbo I, Johnell O, Obrant KJ 1992 Earlier wrist fracture: A confounding factor in distal forearm bone screening. Osteoporos Int 2: 201204.