A Single Determination of a Urinary Biochemical Marker of Bone Turnover for Detecting Bone Density in the Hip

Alfred K Pfister MD^*

Shelda Martin MD^*

Chris Welch MS^†

Paul D Saville, MD^*

^*Department of Medicine, West Virginia University School of Medicine,
Charleston, West Virginia

^†CAMCARE Research Institute, Charleston, West Virginia

KEY WORDS: urinary deoxypyridinoline, hip, osteoporosis, and bone mineral density.

ABSTRACT

Objective: The determination of a single urinary biochemical marker of bone turnover, N-telopeptide, has been proposed as a screening test to classify bone mineral density (BMD) of the hip as normal, osteopenic, or osteoporotic bone. The purpose of this study was to determine the value of a different urinary biochemical marker of bone turnover, deoxypyridinoline (Dpd), as a cost-effective method to discriminate osteopenia or osteoporosis of the hip.

Methods: A urinary assay of Dpd and a BMD were prospectively performed in estrogen-depleted women (n = 120) who had never received remedial bone treatment.

Results: The urinary Dpd showed a negative correlation to all regions of the hip, but it did not reliably distinguish normal, osteopenic, or osteoporotic BMD. In fact, the body-mass index showed a stronger correlation to the hip BMD. Moreover, the Dpd excretion would have missed 53% of clinically relevant osteopenia or osteoporosis cases.

Conclusions: A single measurement of the urinary Dpd cannot be used as a screening test or as a replacement for a BMD study to determine hip osteopenia or osteoporosis.

INTRODUCTION

Hip fracture is a devastating event that usually occurs after the age of 65; women are affected more than twice as often as men.1,2 After 1 year, the mortality rate is between 12% and 24%, and up to 25% of the survivors remain in nursing home care.3,4 Furthermore, most of the survivors experience a functional decline in both the upper and lower extremities as well as a deterioration in the social quality of life.4,5 Subsequent hospitalizations for these patients are more lengthy, costly, and frequent than those in matched cohorts.5 The annual cost of hip fractures to the United States health care budget is estimated to be over 9 billion dollars.6 Although many factors are involved in the pathogenesis of the hip fracture, the force of the fall and the strength of the hip bone are the ultimate determinants that define whether a break will occur.7–9

The bone mineral density measurement (BMD) at the hip site is a powerful predictor of future fracture rates.10–12 An increase in the BMD by drug therapy significantly reduces the rate of hip fracture.13 Measurement of the BMD at a peripheral area as a screening test for osteoporosis (for example, in the distal forearm) is less expensive, but the prediction rate of relative risk of hip fracture using this area is much lower and frequently discordant with the central hip measurement.10,12,14 Ultrasound of the calcaneus, also an inexpensive technique, predicts the risk of hip fracture as effectively as BMD of the hip.15,16 Additionally, the measurement of a urinary biochemical marker of bone turnover has been proposed as an inexpensive screening test for osteoporosis of the hip and spine.17

The skeletal structure maintains its balance by the continuous remodeling process of osteoclastic resorption of established bone coupled with the osteoblastic build up of new bone. In osteoporosis, the bone metabolism becomes accelerated, with an uncoupling of this process. Osteoclastic resorption now exceeds osteoblastic bone formation, resulting in the net loss of bone.18,19 Thus, a resultant increase in the urinary markers of bone resorption occurs. This has been historically measured by the determination of hydroxyproline excretion. Newer assays that reflect osteoclastic bone resorption are the telopeptides, N-telopeptide (Ntx) and C-telopeptide (Ctx), as well as the pyridium crosslinks, pyridinoline and deoxypyridinoline (Dpd). Their commercial availability has been popularized in recent years because of their higher specificity, low cost, and ease of use.20

Quantifying these urinary markers has been advocated as a means of predicting an earlier response than a BMD determination to antiresorptive therapy in osteoporotic patients.21–24 Additionally, these assays are useful in predicting the amount of bone gain in response therapy,21–24 future bone loss,25 and future fracture rates.26 Moreover, one report indicated that a single determination of a telopeptide urinary marker (Ntx) had the power to classify normal, osteopenic, and osteoporotic BMD using the criteria of the World Health Organization (WHO).17,27 Based on this study and the high incidence of hip fracture noted in our region,7 we evaluated whether a single determination of a pyridium crosslink urinary bone resorption marker (Dpd) would inexpensively discriminate between normal, osteopenic, and osteoporotic BMD of the hip.

MATERIALS AND METHODS

Patients

Over an 18-month period, we prospectively studied ambulatory women who underwent intake evaluations at the Bone and Mineral Clinic of the Charleston Area Medical Center. Patients were either self-referred or sent by other physicians. All subjects were estrogen depleted. Patients with medical illnesses (such as chronic renal failure, advanced pulmonary, cardiac, or malignant conditions), fractures within the past year, hyperthyroidism, hyperparathyroidism, alcoholism, or previous therapy with glucocorticoids, calcitonin, bisphosphonates, or estrogens were excluded.

Testing

A certified technician performed the hip dual-energy x-ray absorptiometry (Hologic 1000, Waltham, MA). The daily standardization by a phantom produced a coefficient of variation of less than 1.0%. Urine for Dpd was collected at the intake examination between 8 am and 10 am using the second voided specimen and analyzed using an immunoassay (Pyrilinks D, Metro Biosystems, Mountain View, CA). Results were expressed as the nmols of Dpd divided by nmols of creatinine. The normal reference value for healthy postmenopausal women was defined by the manufacturer’s package insert as up to 7.4.

Statistics

Data analysis was performed using SAS 8.01 (SAS Institute, Cary, NC) software. Age and body-mass index (BMI)—adjusted partial correlations were performed between the urinary Dpd and BMD of total hip as well as the femoral neck, trochanter, intertrochanter, and Ward’s triangle regions of the hip. Similarly, both age-adjusted and BMI-adjusted partial correlations were performed for the hip and its various regions. A 95% confidence interval of the urinary Dpd was used to classify normal, osteopenic, and osteoporotic bone. One-way analyses of variance (ANOVAs) were performed to detect differences in urinary Dpd excretion between bone status groups. Significance level was set at P < 0.05.

RESULTS

The mean age (plus or minus) of the sample was 69 ± 9.7 years (range, 41–90). Forty-three percent of the participants had urinary Dpd values higher than premenopausal upper limits of normal. After adjustment for age and BMI, all regions of the hip correlated negatively with the urinary Dpd excretion (Table 1). The comparison of the Dpd excretion to the hip BMD accounted for about 6% of the variance (Figure 1), but the age and BMI adjusted value extended this variance to 7.1%. Furthermore, the age and BMI variables failed to show any linear trend with the Dpd excretion and accounted for 4% and 12% of the hip BMD, respectively. Additionally, the urinary Dpd did not predict normal, osteopenic, or osteoporotic bone with any degree of confidence. Although the Dpd excretion values did show a significance for the trochanteric region in the osteoporotic group, no significant correlation to the total hip BMD was noted in any of the 3 WHO categories (Table 2).

A single determination of a urinary Dpd yielded a sensitivity of 47% and a specificity of 57%. The diagnostic accuracy of a single elevated urinary Dpd in predicting osteopenia or osteoporosis was poor and yielded a likelihood ratio of only 1.1. The osteoporotic group did excrete a significantly higher Dpd amount than the osteopenic group (P = 0.05, 8.9 ± 4.5 vs 7.01 ± 2.93); however, the normal group did not differ significantly from the osteopenic or osteoporotic group in the amount of Dpd excreted (7.43 ± 2.53). When the urinary Dpd was evaluated within 10-year age-group increments, using partial correlation adjusting for BMI, only the subset of women aged 80 and older showed a significant correlation with BMD (Table 3). No significant correlation was found with the lowest, middle, and highest tertiles of urinary Dpd excretion to the hip BMD.

Of further interest, the age-adjusted partial correlation between the BMI and the total hip BMD and hip region BMDs showed higher correlations than the urinary Dpd (Table 4). This BMI to BMD correlation trend continued through normal, osteopenic, and osteoporotic groups, with the highest correlation to the total hip.

DISCUSSION

A single determination of the urinary Dpd, a biochemical marker of bone turnover, was not useful in classifying women with normal, osteopenic, or osteoporotic hip bone. Fifty-three percent of hip osteopenia or osteoporosis cases would have gone undetected had we relied on this assay as a screening test. A similar study that used the NTX urinary assay was able to make this distinction in the hip and lumbar spine sites; however, their larger, community-based sample had a higher percentage of women with normal hip BMD (43.6 vs 17.5) but a smaller percentage of women with osteoporosis (23.9 vs 35.8) than ours and included women who had previously taken hormone replacement therapy.17 Although we noted stronger or similar correlations compared with other studies using urinary Dpd or Ntx assays to the various regions of the hip BMD, the BMI accounted for a greater variance than the urinary Dpd.24,26,28,29 Likewise, another trial (PEPI) in younger postmenopausal women showed the BMI and age to account for a greater variance to the hip and spine BMD than any biochemical markers of bone turnover.30

The inability of the urinary bone resorption marker to predict the hip BMD is not surprising. The level of bone mass depends not only on the rate of postmenopausal bone loss but also on the peak bone gain during adolescence. During early postmenopause, women undergo bone loss from trabecular bone much faster than cortical bone. This is particularly noticeable in the clinically relevant areas, such as the spine, hip, and distal radius. After the age of 60, however, bone loss is slower and about equal from both trabecular and cortical regions.18 In spite of this more gradual decline of the skeletal mass in the elderly years, urinary bone resorption markers continue to be elevated in a substantial percent of women.31 This could reflect accelerated osteoclastic resorption not only from trabecular bone regions but also from the proportionally more numerous cortical sites in the skeletal structure. Estrogen deficiency, secondary hyperparathyroidism, and age-related bone loss may be totally or partially responsible.18,31

An elevated urinary Dpd with a low BMD is an independent risk factor for the prediction of a future hip fracture.26 Our failure to classify Dpd excretion with bone mass may be related to this assay’s ability to reflect microdamage in the trabecular architecture. Furthermore, the urinary Dpd correlates strongly with the calcaneal ultrasound.32 Although calcaneal ultrasound and DXA predict future hip fractures independently by determination of the BMD, the calcaneal ultrasound may yield some additional clues about bone architecture. These techniques, however, do not provide the additional information about other factors felt to be important in bone strength, specifically microdamage.33 The elevated urinary Dpd in the face of a reduced bone mass may signify microdamage and thus becomes an added risk factor for future fracture.

The chief limitation of the study was that all patients came from a specialized clinic rather than from the community. Consequently, we had a lower proportion of women with a normal BMD. Conversely, had we included women who had received current or remote estrogen replacement therapy, this group with a normal BMD would have constituted 23% of our population. This would have changed our mean Dpd excretion in this group to 6.85 (95% confidence intervals, 5.80 to 7.90). However, these values still do not change our outcomes in classifying BMD. Furthermore, the use of a single urine specimen for a biochemical marker of bone turnover imposes limitations due to the well-recognized analytical and biological variability.34 However, obtaining 2 or more urine specimens to overcome this variability would reduce the advantage of a inexpensive screening test.

CONCLUSIONS

Urinary markers of bone turnover are a dynamic test of bone status and have considerable variability. The BMD, however, is not only a cross-sectional measurement of past and current bone loss, but also a representation of peak bone mass attained during adolescence. These biochemical assays serve, on the other hand, as a useful guide in osteoporotic patients to identify an early response to antiresorptive treatment. They may identify individuals with higher future bone loss and probably reflect trabecular microdamage. They cannot, however, accurately and inexpensively serve as a screening test to identify patients with osteopenia or osteoporosis. A single measurement of the BMD remains the most reliable method. Further studies are needed to validate the high correlation noted between calcaneal ultrasound and urinary Dpd and to determine whether this assay combined with ultrasound can independently predict the risk of future hip fracture in a prospective study.

Acknowledgments

The authors wish to thank Dr. William H. Carter for his review of this manuscript and Ms. Vickie Starcher for her technical assistance.

REFERENCES

1. Kannus P, Pakkari J, Sievanen H, et al: Epidemiology of hip fractures. Bone 18(suppl): 57–63, 1998.

2. Lips P: Epidemiology and predictors of fractures associated with osteoporosis. Am J Med 103(suppl):3–11, 1997.

3. Browner WS, Pressman AR, Nevitt MC, Cummings SR: Mortality following fractures in older women. Arch Intern Med 156:1521–1525, 1996.

4. Zimmerman SI, Fox SM, Magaziner J: Pyschological aspects of osteoporosis. Phys Med Rehab Clin North Am 5:441–453, 1995.

5. Wolinsky FD, Fitzgerald JF, Stump TE: The effect of hip fracture on mortality, hospitalization, and functional status: A prospective study. Am J Public Health 87:398–403, 1997.

6. Ray NF, Chan JF, Thamer M, Melton LJ: Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: Report from the National Osteoporosis Foundation. J Bone Miner Res 12:24-35, 1997.

7. Pfister AK, McJunkin J, Santrock DA: Hip fracture outcomes and their prevention in Kanawha County, West Virginia. WV Med J 95:170–174, 1999.

8. Slemenda C: Prevention of hip fractures: A risk factor modification. Am J Med 103(suppl):65–73, 1997.

9. Cummings SR, Nevitt MC, Browner WS, et al: Risk factors for hip fractures in white women. N Engl J Med 332:767–773, 1995.

10. Cummings SR, Black DM, Nevitt MC: Bone density at various sites for prediction of hip fracture. Lancet 341:72–75, 1993.

11. Marshall D, Johnell O, Wedel H: Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fracture. Br Med J 312:1254–1259, 1996.

12. Cummings SR, Black DM, Nevitt MC, et al: Appendicular bone density and age predict hip fracture in women. JAMA 263:665–668, 1990.

13. Black DM, Cummings SR, Karpf DB, et al: Randomized trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Lancet 348:1535–1541, 1996.

14. Varney LF, Parker RA, Vincelette A, Greenspan SL: Classification of osteoporosis and osteopenia in postmenopausal women is dependent on site-specific analysis. J Clin Densitometry 2:275–283, 1999.

15. Hans D, Dargent-Molina P, Schott AM, et al: Ultrasound heel measurements to predict hip fracture in elderly women: The EPIDOS study. Lancet 348:511–514, 1996.

16. Bauer DC, Gluer CC, Cauley JA, et al: Broadband ultrasound attenuation predicts fractures strongly and independently of densitometry in older women. Arch Intern Med 157:629–624, 1997.

17. Schneider DL, Barret-Conner EL: Urinary N-telopeptide levels discriminate normal, osteopenic, and osteoporotic bone mineral density. Arch Intern Med 157:1241–1245, 1997.

18. Riggs BL, Khosla S, Melton LJ: A unitary model for involutional osteoporosis: estrogen deficiency causes both type I and type II osteoporosis in postmenopausal women and contributes to bone loss in aging men. J Bone Miner Res 13:763–773, 1998.

19. Eastell R: Treatment of postmenopausal osteoporosis. N Engl J Med 338:736–746, 1998.

20. Seibel MJ, Woitge HW: Basic principles and clinical applications of biochemical markers of bone metabolism. J Clin Densitometry 2:200–321, 1999.

21. Miller PD, Baran DT, Bilezikian NJS, et al: Practical clinical application of biochemical markers of bone turnover. J Clin Densitometry 3:323–332, 1999.

22. Rosen CJ, Chesnut CH, Mallinak NJS, et al: The predictive value of biochemical markers of bone turnover for bone mineral density in early postmenopausal women treated with hormone replacement therapy or calcium supplementation. J Clin Endocrinol Metab 82:1904–1910, 1997.

23. Garnero P, Shih WJ, Ginetys E, et al: Comparison of new biochemical markers of bone turnover in late postmenopausal osteoporotic women in response to alendronate treatment. J Clin Endocrinol Metab 79:1693–1700, 1994.

24. Ravn B, Clemmesen B, Christiansen C: Biochemical markers can predict the response in bone mass during alendronate treatment in early postmenopausal women. Bone 24:237–244, 1999.

25. Ross PD, Knowlton W: Rapid bone loss is associated with increased levels of biochemical markers. J Bone Miner Res 13:297–302, 1998.

26. Garnero P, Hausherr E, Chapuy MC, et al: Markers of bone resorption predict hip fracture in elderly women. J Bone Miner Res 1l:531–1538, 1996.

27. World Health Organization:1994 Assessment of Fracture Risk and its Application to Screening for Postmenopausal Osteoporosis. Technical Report Series. Geneva, Switzerland: World Health Organization; 1994.

28. The Writing Group for the PEPI trial: Effects of hormone therapy on bone mineral density: results from the Postmenopausal Estrogen/ Progesterone Interventions (PEPI) trial. JAMA 276:1389–1296, 1996.

29. Krall E, Dawson-Hughes B, Hirst K, et al: Bone mineral density and biochemical markers of bone turnover in healthy elderly men and women. J Gerontol 52A:M61–M67, 1997.

30. Marcus R, Holloway L, Wells B, et al: The relationship of biochemical markers of bone turnover to bone density changes in postmenopausal women: Results from the Postmenopausal Extrogen/ Progestin Interventions (PEPI) Trial. J Bone Miner Res 14:1583–1595, 1999.

31. Melton LJ, Khosla S, Atkinson EJ, et al: Relationship of bone turnover to bone density and fractures. J Bone Miner Res 12:1083–1091, 1997.

32. Fiore CE, Penisi P, Giblilaro M, et al: Correlation of quantitative ultrasound of bone with biochemical markers of resorption in women with fractures. J Clin Densitometry 2:231–239, 1999.

33. Frost HM: Does bone mass equate with bone health? J Clin Densitometry 4:179–184, 2001.

34. Hsin-Shan J, Leung S, Brown S: Comparison of analytical performance and biological variability of three bone resorption assays. Clin Chem 43:1570–1576, 1997.

Table 1. The correlation of urinary Dpd excretion with the BMD to various regions of the hip in 120 women

Anatomical region r P value r2

Total hip –0.27 <0.01 0.07

Intertrochanteric –0.27 <0.01 0.07

Femoral neck –0.25 <0.01 0.06

Trochanter –0.27 <0.01 0.07

Ward’s triangle –0.28 <0.01 0.08

Figure 1. The linear regression of age, body-mass index, and urinary Dpd to the bone mineral density (gms/cm2) of the hip.

Table 2. Adjusted age and body mass index relationship of the urinary excretion of Dpd to the hip bone mineral density according to the World Health Organization

Correlation
N Dpd (95% CI)* coefficient P value r2

Normal 21 7.43 (6.28–8.58) –0.15 0.55 0.02

Osteopenia 56 7.01 (6.22–7.79) –0.21 0.12 0.04

Osteoporosis 43 8.91 (7.54–10.28) –0.26 0.10 0.07

* = mean value with 95% confidence intervals (CI).

Table 3. The partial correlation of urinary Dpd excretion (adjusted for body mass index) to the BMD to the various regions in the hip of women aged 80 and older

Table 4. Partial correlation of body mass index to the bone density to the various regions in the hip and World Health Organization bone health categories*

Correlation

Anatomic region coefficient P value r2

Total hip –0.76 <0.01 0.57

Intertrochanter –0.74 <0.01 0.54

Femoral neck –0.57 <0.05 0.45

Trochanter –0.79 <0.01 0.62

Ward’s triangle –0.69 <0.01 0.42

Hip regions r P value r2

Total hip –0.34 0.001 0.12

Intertrochanter –0.35 <0.001 0.12

Neck –0.30 0.001 0.09

Trochanter –0.21 0.02 0.05

WHO categories

Normal –0.53 <0.05 0.28

Osteopenia –0.24 0.08 0.06

Osteoporosis –0.31 <0.05 0.10

*Correlation is with the effect of age removed.