Volume 87, Issue 11, Pages 1459-1462 (November 2006)

10 of 333

ABSTRACT

FULL TEXT

FULL-TEXT PDF (88 KB)

Bone Mineral Density in Hip-Fracture Patients With Parkinson’s Disease: A Case-Control Study

Abstract

Di Monaco M, Vallero F, Di Monaco R, Tappero R, Cavanna A. Bone mineral density in hip-fracture patients with Parkinson’s disease: a case-control study.

Objective

To investigate bone mineral density (BMD) levels in patients with Parkinson’s disease (PD) who sustained a hip fracture.

Design

Case-control study.

Setting

Rehabilitation hospital in Italy.

Participants

We investigated 831 out of 887 white patients consecutively admitted to a rehabilitation hospital because of an original hip fracture resulting from a fall. Twenty-eight (3.37%) of the 831 patients were affected by PD. Twenty-eight controls matched for sex, age, and hip-fracture type (cervical or trochanteric) were found among the 803 non-PD patients.

Interventions

Not applicable.

Main Outcome Measures

BMD was assessed by dual-energy x-ray absorptiometry (DXA) at the unfractured femur. Five sites were investigated in each subject: total proximal femur, femoral neck, trochanter, intertrochanteric area, and Ward’s triangle. DXA scan was performed a mean ± standard deviation of 22.2±7.8 days after fracture occurrence in the 28 patients and 22.0±5.3 days after fracture occurrence in the 28 controls.

Results

BMD expressed as a T score did not differ significantly between the 28 PD patients and the 28 controls, whereas z score in the PD patients was significantly lower than 0±1 in the age- and sex-matched general population at 4 of the 5 sites of BMD assessment.

Conclusions

A sample of PD fallers who sustained a hip fracture had femoral BMD levels similar to those found in matched hip-fracture fallers who did not suffer from PD and significantly lower than those found in the matched reference population.

Key Words: Bone mineral density, Hip fracture, Osteoporosis, Parkinson’s disease, Rehabilitation

Article Outline

• Abstract

• Methods

• Data Analysis

• Results

• Discussion

• Prevalence of PD Among Hip-Fracture Patients

• BMD in PD Patients, Controls, and the Reference Population: Meaning of the Study

• Study Limitations

• Conclusions

• References

• Copyright

FRACTURE RISK IS SIGNIFICANTLY increased among patients with Parkinson’s disease (PD).1, 2, 3, 4, 5 The high fracture rate could be attributable to the increased incidence of falls in PD patients who suffer from postural instability, bradykinesia, rigidity, and postural hypotension.6, 7, 8, 9, 10, 11 Alternatively, the increased risk of fracture may depend on bone fragility, given the significant association between PD and low bone mineral density (BMD).11, 12, 13, 14, 15 Among fracture sites, a relatively high proportion of hip fractures was observed1, 4, 5: PD more than doubled hip-fracture risk over a 10-year follow-up,4 and by 10 years after diagnosis, an estimated 27% of a PD cohort experienced a new hip fracture.1 The absence of adequately directed arm movements in PD fallers, together with their tendency to fall backward or to the side, may explain the high incidence of hip injuries.10 Otherwise, the neuromuscular deficit in PD may differentially reduce hip BMD relative to other sites, thus increasing hip-fracture risk.12 On the whole, prospective data did not establish the independence of bone fragility and falls as determinants of hip-fracture risk.1, 4, 10 To investigate BMD levels in PD patients who sustained a hip fracture, we compared femoral BMD among PD fallers with hip fracture; control fallers matched for age, sex, and hip-fracture type; and a reference population.

Methods

We evaluated 887 white patients consecutively admitted to our physical medicine and rehabilitation division because of their first hip fracture. We focused on white patients because few nonwhite elderly people live in Italy. Fifty of the 887 patients were excluded from the study because their fractures resulted from major trauma or cancer affecting bone or because they occurred spontaneously. The remaining 837 patients sustained fractures that resulted from a fall. Three of these 837 patients were excluded from our study because they had arthroplasties at the unfractured hip, and we could not measure the hip BMD. The remaining 834 patients were asked to undergo dual-energy x-ray absorptiometry (DXA) assessment. Three of these 834 patients refused to undergo DXA assessment and were excluded from the study. The final study sample included 831 patients. Twenty-eight (3.37%) of the 831 patients were affected by PD (the diagnosis either was self-reported or derived from clinical records, and it was confirmed by a neurologist during the hospital stay). Eighteen of the 28 PD patients were women (10 were men), whereas 16 suffered from trochanteric fractures (12 suffered from cervical fractures). The mean age ± standard deviation (SD) in the 28 patients was 77.5±7.8 years (age range, 60–90y). Number of years of PD and levodopa dosage were 7.5±3.6 and 336±246mg, respectively. Hoehn and Yahr stage scale scores ranged from 1 to 4 (mean, 2.96±0.84; median, 3; interquartile range, 2). Nine of the 28 patients had parkinsonism variants. No smokers were found among the 28 patients. Twenty-eight nonsmoking controls matched for sex, age, and hip-fracture type (cervical or trochanteric) were found among the 803 non-PD patients. When multiple matched controls were available, we included in the study the one whose body weight was the most similar to the corresponding patient’s body weight. None of the controls was affected by neurologic diseases.

BMD was assessed by DXAa 22.2±7.8 days after fracture occurrence in the 28 patients (22.0±5.3d after fracture occurrence in the 28 controls). Five sites were assessed in each patient at the unfractured hip: total proximal femur, femoral neck, trochanter, intertrochanteric area, and Ward’s triangle. Body weight and height were measured at the time of DXA assessment. None of the patients included in the study had begun specific treatment with drugs for osteoporosis before DXA assessment.

Data Analysis

BMD values were expressed as a T score. (BMD values were compared with the reference range of the young sex-matched population by using SD units.) Because BMD values were normally distributed, BMD comparison between PD patients and controls was performed by an unpaired t test. The same test was used to compare anthropometric variables (ie, body weight, body height, body mass index [BMI]) between the 2 groups. For the 24 of the 28 patients who were younger than 85 years, a z score was available. (It was routinely calculated by the software supplied by the manufacturer.) BMD values were compared with the reference range of the age- and sex-matched population by using SD units. In the 24 patients, we performed an unpaired t test to assess the hypothesis that the mean z score differed significantly from 0±1 (ie, mean ± SD in the age- and sex-matched general population). The reference population for both T- and z-score calculations was derived from the Third National Health and Nutrition Examination Survey.

In the 28 patients with PD we performed bivariate correlations by using a Spearman rank test between femoral BMD (expressed as a T score) and 5 variables: age, BMI, number of years of PD, Hoehn and Yahr stage, and levodopa dosage. The variables correlated significantly with femoral BMD were included as the independent variables in a linear multiple regression model with femoral BMD as the dependent variable. The residuals were normally distributed in the regression model.

The statistical package used was SPSS.b

Results

BMD values did not differ significantly between the 28 PD patients and the 28 controls (table 1). Also, no significant differences in anthropometric variables were found between the 2 groups. The mean body weight was 54.4±11.2kg in the PD patients and 56.4±9.4kg in the controls (mean difference, 2.0kg; 95% confidence interval [CI], −3.49 to 7.57; P=.46), body height was 160.7±9.5 and 161.9± 6.5cm, respectively, in the 2 groups (mean difference, 1.2cm; 95% CI, −3.17 to 5.53; P=.59), and BMI was 21.2±4.7kg/m2 in the PD patients and 21.4±2.7kg/m2 in the controls (mean difference, 0.2kg/m2; 95% CI, −1.85 to 2.27; P=.84). In the 24 PD patients who were younger than 85 years, the z score was significantly lower than 0±1 (ie, mean in the age- and sex-matched general population) at 4 of the 5 sites of BMD assessment (table 2).

Table 1.

BMD Levels (expressed as a T score) in 28 PD Patients and 28 Controls


Site of Hip BMD Assessment	PD Patients	Controls	Mean Difference (95% CI)	P
Totalproximalfemur	−2.72±0.79	−2.73±0.73	−0.01(−41.3 to 40.0)	.97
Femoralneck	−2.67±0.76	−2.85±0.87	−0.18(−0.61 to 0.26)	.43
Trochanter	−2.64±0.91	−2.50±0.94	0.14(−0.35 to 0.64)	.56
Intertrochantericarea	−2.53±0.82	−2.68±0.88	−0.15(−0.61 to 0.30)	.50
Ward’striangle	−3.16±1.02	−3.44±0.88	−0.28(−0.79 to 0.23)	.28

NOTE. Values are mean ± SD or as indicated.

Abbreviation: CI, confidence interval.

Table 2.

BMD Levels (expressed as a z score) in 24 PD Patients Who Were Younger Than 85 Years


Site of Hip BMD Assessment	z Score	95% CI	P
Totalproximalfemur	−1.17±0.88	−1.7to−0.64	<.001
Femoralneck	−0.70±0.82	−1.22to−0.18	.01
Trochanter	−1.32±0.94	−1.8to−0.77	<.001
Intertrochantericarea	−1.13±0.93	−1.68to−0.58	<.001
Ward’striangle	−0.41±0.98	−0.97to0.15	.16

NOTE. Values are mean ± SD or as indicated. Comparisons were performed between the 24 patients and the reference data from subjects matched for age, sex, and ethnicity derived from the Third National Health and Nutrition Examination Survey.

In the 28 PD patients, we found a significant correlation between total femur BMD and Hoehn and Yahr stage (ρ=−.51, P=.006), number of years of PD (ρ =−.56, P=.002), and BMI (ρ=.50, P=.007). Multiple regression showed that illness duration and Hoehn and Yahr stage were independently associated with femoral BMD. Table 3 shows the results of multiple regression with BMD assessed at the total proximal femur. The results were similar with BMD assessed at the other 4 sites (data not shown).

Table 3.

Multiple Regression Model


Independent Variables	Partial Correlation	P
Hoehn and Yahr stage	−.55	.003
Years of PD	−.54	.004
BMI	.13	.53

NOTE. Hoehn and Yahr stage, years of PD, and BMI were included in the model as the independent variables. BMD assessed at total proximal femur was the dependent variable. R2=.48, F=11.56, P<.001.

Discussion

Prevalence of PD Among Hip-Fracture Patients

In 1 sample of elderly patients with hip fracture resulting from a fall, 28 (3.37%) of 831 suffered from PD. This prevalence is identical to that reported in a recent study by Idjadi et al,16 who found 31 PD patients in a sample of 920 hip-fracture subjects, and it is slightly lower than the 4.3% and 5.4% prevalence rates reported in 2 previous studies examining 57717 and 136118 hip-fracture patients, respectively. Overall, the prevalence of PD among hip-fracture patients appears to be higher than that observed in the general population,19 which is consistent with the reported increase in hip-fracture risk resulting from PD.1, 2, 3, 4, 5

BMD in PD Patients, Controls, and the Reference Population: Meaning of the Study

Femoral BMD was low in PD patients who sustained a hip fracture: the mean T score ranged from −3.16 to −2.53, depending on the femoral site of BMD assessment. We did not observe significant differences in T score between PD fallers with hip fracture and control fallers matched for age, sex, and hip-fracture type. To our knowledge, no previous studies have compared BMD between patients with PD with fractures and controls with fractures. Current data indicate that each of our PD and non-PD patients sustained a hip fracture at similar levels of femoral BMD. Several researchers20, 21 showed that femoral BMD levels were strongly associated with the risk of hip fracture in the general population. Likewise, in PD patients, 1 prospective study2 showed that low BMD predicted a high hip-fracture risk over a 1-year follow-up. In our cross-sectional study, femoral BMD was lower in patients who sustained a hip fracture than in the reference population (comparison performed by using z scores), as expected on the basis of the prospective studies.

In our sample of hip-fracture fallers with PD, we showed that illness duration, BMI, and Hoehn and Yahr stage correlated significantly with femoral BMD. These data are in agreement with a previous prospective investigation2 and 2 cross-sectional studies.22, 23 However, BMI was not significantly associated with femoral BMD after adjustment for clinical stage and illness duration. We think that the inverse correlation between BMI and both duration and severity of the disease may explain this result. The association between a high Hoehn and Yahr stage and a low BMD level suggests that heightened awareness by physicians may prevent bone loss in PD, because reduced mobility may be preventable. Earlier studies showed that low BMD in patients with PD may be related to other preventable factors we did not study: depletion of vitamin D14 and vitamin K1.22 Furthermore, prevention of both osteoporosis and vitamin D deficiency may improve functional outcome after hip fracture, apart from decreasing hip fracture risk, because low BMD and vitamin D depletion were associated with a worse ability to function after acute inpatient rehabilitation.24, 25 Although this hypothesis needs confirmation by prospective studies, the chance of ameliorating functional outcome after hip fracture in PD deserves special consideration, because functional prognosis after hip fracture is negatively affected by PD. PD patients need longer hospitalization, have lower levels of independence in basic activities of daily living, and are more likely to be discharged to nursing facilities than non-PD patients after hip fracture.16, 17, 18

Study Limitations

This study has several limitations. BMD values are influenced by race,26 and our study included white patients only. As a consequence, our results are not generalizable to the overall population of patients who sustain hip fractures. Although hip BMD is a strong predictor of hip fracture risk,20, 21 bone strength and fracture risk are also influenced by BMD-independent factors.27 In PD patients, a recent report showed that high homocysteine concentrations enhanced the risk for hip fracture independently of age-adjusted BMD levels, likely because of homocysteine interference with collagen cross-linking.28 Levodopa treatment induces hyperomocystinemia and may contribute to a high incidence of hip fracture by a BMD-independent mechanism.28 We showed no significant associations between levodopa dosage and BMD in our patients with hip fracture, in agreement with the lack of association between homocysteine concentration and BMD,28 but we did not evaluate homocysteinemia that could affect fracture risk independently of BMD.

We performed DXA assessment after hip fracture. Time between fracture occurrence and DXA assessment may be a confounding variable in our study, because relevant changes in body composition, including a decrease in BMD, have been shown after hip fracture.29 We performed DXA assessment about 3 weeks after fracture occurrence. A shorter time (ie, a few days) may be better to minimize the changes in BMD, but many patients cannot undergo DXA assessment a few days after fracture occurrence.30 The time between fracture occurrence and DXA assessment was quite similar in patients and controls. The reference population for both T- and z-score calculations was not from Italy, because no population-based studies were performed in Italy to establish normative data for femoral BMD. This limitation does not affect the main result of the study—that is, the finding of similar BMD levels between hip-fracture fallers with and without PD—because both groups were derived from the same sample of hip-fracture fallers admitted to our hospital. Conversely, the comparison between hip-fracture patients and reference data we performed by using z scores should be considered with caution, given the potential differences between the Italian and reference populations. Adequate matching of patients and controls for confounding variables is a crucial issue. Controls were matched for sex and age, that is, 2 factors that strongly influence fracture risk.27 Furthermore, we matched controls for hip-fracture type, because BMD’s role in affecting hip-fracture risk varies with the site of fracture: trochanteric fractures are associated with a more severe bone loss than cervical fractures, at least in some age groups.30, 31 However, other confounders may be relevant. The role of smoking habits in influencing fracture risk was ruled out, because all patients and controls were nonsmokers. Anthropometric variables strongly affect the risk of hip fracture.27 We did not observe significant differences in body weight, body height, and BMI between PD patients and controls, but we cannot exclude a confounding role exerted either by minor differences in the variables we investigated or by differences in soft-tissue body composition32 we did not investigate. Finally, we did not evaluate other risk factors for hip fracture, including familiarity and previous fragility fractures. All patients and controls sustained hip fractures that resulted from a fall; however, we did not investigate fall characteristics in detail, and we cannot be sure that there were no differences in fall type between patients and controls.

Conclusions

PD patients who sustained a hip fracture after a fall had hip BMD levels similar to those found in matched non-PD fallers who sustained a hip fracture and significantly lower than those found in the reference population matched for age, sex, and ethnicity.

Suppliers

References

1. 1Johnell O, Melton LJ, Atkinson EJ, O’Fallon WM, Kurland LT. Fracture risk in patients with parkinsonism: a population-based study in Olmsted County, Minnesota. Age Ageing. 1992;21:32–38. MEDLINE

2. 2Sato Y, Kaji M, Tsuru T, Oizumi K. Risk factors for hip fracture among elderly patients with Parkinson’s disease. J Neurol Sci. 2001;82:89–93. CrossRef

3. 3van Staa TP, Leufkens HG, Cooper C. Utility of medical and drug history in fracture risk prediction among men and women. Bone. 2002;31:508–514. Abstract | Full Text | Full-Text PDF (125 KB) | CrossRef

4. 4Taylor BC, Schreiner PJ, Stone KL, et al. Long-term prediction of incident hip fracture risk in elderly white women: study of osteoporotic fractures. J Am Geriatr Soc. 2004;52:1479–1486. MEDLINE | CrossRef

5. 5Genever RW, Downes TW, Medcalf P. Fracture rates in Parkinson’s disease compared with age- and gender-matched controls: a retrospective cohort study. Age Ageing. 2005;34:21–24. MEDLINE | CrossRef

6. 6Gray P, Hildebrand K. Fall risk factors in Parkinson’s disease. J Neurosci Nurs. 2000;32:222–228. MEDLINE

7. 7Bloem BR, Grimbergen YA, Cramer M, Willemsen M, Zwinderman AH. Prospective assessment of falls in Parkinson’s disease. J Neurol. 2001;248:950–958. MEDLINE | CrossRef

8. 8Grimbergen YA, Munneke M, Bloem BR. Falls in Parkinson’s disease. Curr Opin Neurol. 2004;17:405–415. MEDLINE | CrossRef

9. 9Wielinski CL, Erickson-Davis C, Wichmann R, Walde-Douglas M, Parashos SA. Falls and injuries resulting from falls among patients with Parkinson’s disease and other parkinsonian syndromes. Mov Disord. 2005;20:410–415. MEDLINE | CrossRef

10. 10Fink HA, Kuskowski MA, Orwoll ES, Cauley JA, Ensrud KE. Association between Parkinson’s disease and low bone density and falls in older men: the osteoporotic fractures in men study. J Am Geriatr Soc. 2005;53:1559–1564. MEDLINE | CrossRef

11. 11Winogrodzka A, Wagenaar RC, Booij J, Wolters EC. Rigidity and bradykinesia reduce interlimb coordination in Parkinsonian gait. Arch Phys Med Rehabil. 2005;86:183–189. Abstract | Full Text | Full-Text PDF (151 KB) | CrossRef

12. 12Taggart H, Crawford V. Reduced bone density of the hip in elderly patients with Parkinson’s disease. Age Ageing. 1995;24:326–328. MEDLINE

13. 13Revilla M, de la Sierra G, Aguado F, Varela L, Jimenez-Jimenez FJ, Rico H. Bone mass in Parkinson’s disease: a study with three methods. Calcif Tissue Int. 1996;58:311–315. CrossRef

14. 14Sato Y, Kikuyama M, Oizumi K. High prevalence of vitamin D deficiency and reduced bone mass in Parkinson’s disease. Neurology. 1997;49:1273–1278. MEDLINE

15. 15Petroni ML, Albani G, Bicchiega V, et al. Body composition in advanced-stage Parkinson’s disease. Acta Diabetol. 2003;40:S187–S190. CrossRef

16. 16Idjadi JA, Aharonoff GB, Su H, et al. Hip fracture outcomes in patients with Parkinson’s disease. Am J Orthop. 2005;34:341–346. MEDLINE

17. 17Di Monaco M, Vallero F, Di Monaco R, Mautino F, Cavanna A. Functional recovery and length of stay after hip fracture in patients with neurologic impairment. Am J Phys Med Rehabil. 2003;8:143–148.

18. 18Jonsson B, Sernbo I, Johnell O. Rehabilitation of hip fracture in patients with Parkinson’s disease. Scand J Rehabil Med. 1995;27:227–230. MEDLINE

19. 19de Rijk MC, Launer LJ, Berger K, et al.Neurologic Diseases in the Elderly Study Group Prevalence of Parkinson’s disease in Europe: a collaborative study of population-based cohorts. Neurology. 2000;54(11 Suppl 5):S21–S23. MEDLINE

20. 20Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ. 1996;312:1254–1259.

21. 21Kanis JA, Johnell O, Oden A, Dawson A, de Laet C, Jonsson B. Ten year probabilities of osteoporotic fractures according to BMD and diagnostic thresholds. Osteoporos Int. 2001;12:989–995. CrossRef

22. 22Sato Y, Kaji M, Tsuru T, Satoh K, Kondo I. Vitamin K deficiency and osteopenia in vitamin D–deficient elderly women with Parkinson’s disease. Arch Phys Med Rehabil. 2002;83:86–91. Abstract | Full Text | Full-Text PDF (54 KB) | CrossRef

23. 23Kao CH, Chen CC, Wang SJ, Chia LG, Yeh SH. Bone mineral density in patients with Parkinson’s disease measured by dual photon absorptiometry. Nucl Med Commun. 1994;15:173–177. MEDLINE | CrossRef

24. 24Di Monaco M, Di Monaco R, Mautino F, Cavanna A. Femur bone mineral density is independently associated with functional recovery after hip fracture in elderly women. Arch Phys Med Rehabil. 2002;83:1715–1720. Abstract | Full Text | Full-Text PDF (95 KB) | CrossRef

25. 25Di Monaco M, Vallero F, Di Monaco R, Mautino F, Cavanna A. Serum levels of 25-hydroxyvitamin D and functional recovery after hip fracture. Arch Phys Med Rehabil. 2005;86:64–68. Abstract | Full Text | Full-Text PDF (111 KB) | CrossRef

26. 26Kanis JA, Gluer C. An update on the diagnosis and assessment of osteoporosis with densitometry. Osteoporos Int. 2000;11:192–202. CrossRef

27. 27Hodgson SF, Watts NB, Bilezikian JP, et al. American Association of Clinical Endocrinologists medical guidelines for clinical practice for the prevention and treatment of postmenopausal osteoporosis: 2001 edition, with selected updates for 2003. [published erratum in: Endocr Pract 2004;10:90] Endocr Pract. 2003;9:544–564.

28. 28Sato Y, Iwamoto J, Kanoko T, Satoh K. Homocysteine as a predictive factor for hip fracture in elderly women with Parkinson’s disease. Am J Med. 2005;118:1250–1255. Abstract | Full Text | Full-Text PDF (140 KB) | CrossRef

29. 29Fox KM, Magaziner J, Hawkes WG, et al. Loss of bone density and lean body mass after hip fracture. Osteoporos Int. 2000;11:31–35. CrossRef

30. 30Greenspan S, Myers E, Maitland L, Kido TH, Krasnow MB, Hayes WC. Trochanteric bone mineral density is associated with type of hip fracture in the elderly. J Bone Miner Res. 1994;9:1889–1894. MEDLINE | CrossRef

31. 31Di Monaco M, Di Monaco R, Mautino F, Cavanna A. Femur bone mineral density, age and fracture type in 300 hip-fractured women. Aging Clin Exp Res. 2002;14:47–51. MEDLINE

32. 32Di Monaco M, Vallero F, Di Monaco R, Mautino F, Cavanna A. Body composition and hip-fracture type in elderly women. Clin Rheumatol. 2004;23:6–10. MEDLINE | CrossRef

a Osteoporosis Research Center, Presidio Sanitario San Camillo, Torino, Italy

b Division of Physical Medicine and Rehabilitation, Presidio Sanitario San Camillo, Torino, Italy

c Institute of Social and Economical Research, Società Ricerca e Formazione, Torino, Italy.

Reprint requests to Marco Di Monaco, MD, Osteoporosis Research Center, Presidio Sanitario San Camillo, Strada Santa Margherita 136, 10131, Torino, Italy

No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated.

a QDR 4500W; Hologic Inc, 35 Crosby Dr, Bedford, MA 01730.

b Version 10.0; SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

PII: S0003-9993(06)00870-7

doi:10.1016/j.apmr.2006.07.265

10 of 333