 |
|
High-Sensitivity C-Reactive Protein and Fracture Risk in Elderly Women
To the Editor: Reduced bone mass and increased risk of fracture have been associated with inflammatory diseases such as rheumatoid arthritis,1 inflammatory bowel disease,2 and ankylosing spondylitis.3 Inflammatory cytokines stimulate osteoclastic bone resorption4 and inhibit osteoblast function.5 We hypothesized that serum high-sensitivity C-reactive protein (hsCRP), a marker of systemic inflammation, is associated with fracture risk.
Methods
An age-stratified sample of 1494 women (99% white), representing 77.1% of eligible participants, was randomly recruited from electoral rolls for the Geelong Osteoporosis Study.6 The study region is characteristic of Australia in age distribution and socioeconomic indicators. The inclusion criterion of age 65 years or older was met by 522 women. Of these, 33 were excluded because serum was unavailable for analysis and 45 were excluded for baseline use of hormone therapy or oral glucocorticoids for at least 6 months, leaving a study population of 444 women. Participants were comparable with those excluded except that a smaller proportion used calcium and vitamin D supplements (13.3% vs 24.4%, respectively; P = .01).
Baseline hsCRP levels were measured by the Roche/Hitachi Tina-quant (Latex) assay (Roche Diagnostics, Sydney, Australia).7 Serum C-telopeptide ( -Crosslaps, Roche Diagnostics) and bone-specific alkaline phosphatase (Bone-ALP, Roche Diagnostics) were measured as markers of bone resorption and formation. Bone mineral density (BMD) was measured by Lunar DPX-L and analyzed using Lunar DPX-L software, version 1.31 (Lunar Corp, Madison, Wis).
Incident fractures were prospectively identified from radiological reports using a validated procedure and blinded ascertainment.8 Pathologic fractures and minor fractures of finger, face, and patella were excluded from analysis.
Baseline assessments were performed from 1994 to 1997, and participants were followed up until fracture, death, migration from the study region, or the end of 2002. Hazard ratios (HRs) for fracture were determined from Cox proportional hazards models predicting time to first fracture, controlling for BMD, prevalent fracture, age, anthropometry, lifestyle, medication use, and comorbidity, as defined in Table 1. Because of skewed distributions, values for hsCRP were log-transformed (ln-hsCRP); ln-hsCRP and BMD values are expressed as standard deviation units. Linear regression was used to determine associations between ln-hsCRP and log-transformed C-telopeptide, log-transformed bone-specific alkaline phosphatase, and BMD. The study was approved by the Barwon Health Research and Ethics Advisory Committee, and participants provided written consent.
|
|
|
|
Table 1. Baseline Participant Characteristics (N = 444)
|
|
|
Results
The median follow-up period was 5.5 years (interquartile range, 3.8-6.2 years), resulting in 2208 person-years of follow-up. The median hsCRP level was 2.44 mg/mL (interquartile range, 1.32-4.38 mg/mL) (Table 1). Only 26 of the 444 women had hsCRP values greater than 10 mg/L. During the study period, 56 women died, 13 left the region, and 102 had a fracture. Ninety-six incident fracture cases were included in the analysis: 32 spine, 21 hip, 9 humerus, 8 rib, 7 wrist, 7 radius/ulna, 3 pelvis, and 11 other; 2 women had 2 coincident fractures.
The unadjusted HR for fracture increased by 23% for each SD increase in ln-hsCRP (HR, 1.23; 95% confidence interval, 1.01-1.51). The age-standardized absolute risk of fracture during the study period increased from 16.3% (95% CI, 6.8%-25.8%) for ln-hsCRP less than –1 SD (0.96 mg/L) to 28.9% (95% confidence interval, 17.7%-40.1%) for ln-hsCRP greater than +1 SD (6.35 mg/L). Multivariate models consistently included significant contributions from ln-hsCRP, prevalent fracture, and BMD (Table 2). For each SD increase in ln-hsCRP, there was an independent 24% to 32% increase in fracture risk, depending on site-specific BMD used in the model. Fracture risk was independently increased 52% to 79% for each SD decrease in BMD and 52% to 73% by previous fracture. Further adjustment for bone turnover markers, lifestyle factors, diet, medication use, and comorbidity had only minimal effects on the point estimate and did not change the significance of hsCRP. Excluding participants with baseline inflammatory diseases, coronary artery disease, or stroke within 1 year of baseline did not significantly change the results. There were no significant correlations between ln-hsCRP and BMD, log-transformed C-telopeptide, or log-transformed bone-specific alkaline phosphatase.
|
|
|
|
Table 2. Hazard Ratios (HRs) for Fracture Associated With a 1-SD Increase in ln-hsCRP, 1-SD Decrease in BMD, and Prevalent Fracture*
|
|
|
Comment
To our knowledge, this is the first report that increased serum hsCRP concentration is associated with increased fracture risk. The 23% increased risk associated with each SD increase in ln-hsCRP was not explained by differences in BMD, prevalent fracture, markers of bone turnover, diet, lifestyle, medications, or comorbidity. This is consistent with hsCRP-associated alterations in remodeling balance at individual bone remodeling units, possibly mediated by inflammatory processes, that might compromise bone strength.9
Limitations of this study include self-reporting of medication use, history of prevalent fracture, and comorbidities. Vertebral fractures not presenting clinically were not ascertained. Results may not be generalizable beyond a white population. Complex interactions between inflammation and other factors modulating bone metabolism are likely, and circulating hsCRP levels may be a surrogate for unrecognized confounders that affect fracture risk.
Within these constraints, circulating hsCRP appears to be an independent predictor of fracture risk in elderly women. Serum hsCRP should be further studied as a potential marker of fracture risk.
Author Contributions: Dr Pasco had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Pasco, Kotowicz, Nicholson, Schneider.
Acquisition of data: Pasco, Kotowicz, Henry, Spilsbury, Box, Schneider.
Analysis and interpretation of data: Pasco, Kotowicz, Schneider.
Drafting of the manuscript: Pasco, Kotowicz.
Critical revision of the manuscript for important intellectual content: Pasco, Kotowicz, Henry, Nicholson, Spilsbury, Box, Schneider.
Statistical analysis: Pasco, Henry.
Obtained funding: Pasco, Kotowicz, Nicholson, Schneider.
Administrative, technical, or material support: Pasco, Nicholson, Spilsbury, Box, Schneider.
Study supervision: Pasco, Kotowicz, Nicholson, Schneider.
Financial Disclosures: Dr Schneider reports having applied for an Australian patent, provisional application 2005902311. Dr Schneider and Dr Nicholson have applied for an international patent, application PCT/AU2006/000592. Both patents are for the use of hsCRP to predict fractures. The beneficiaries of the patent income would be Bayside Health and the University of Melbourne. No other financial disclosures were reported.
Funding/Support: The study was funded by the Victorian Health Promotion Foundation and the Geelong Region Medical Research Foundation.
Role of the Sponsors: The sponsors played no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript.
Julie A. Pasco, PhD
juliep{at}barwonhealth.org.au
Mark A. Kotowicz, MBBS, FRACP;
Margaret J. Henry, PhD;
Geoffrey C. Nicholson, PhD, FRCP, FRACP
Department of Clinical and Biomedical Sciences: Barwon Health
University of Melbourne
Geelong, Australia
Heather J. Spilsbury, MSc;
Jeffrey D. Box, PhD
Alfred Pathology Service
Melbourne, Australia
Hans G. Schneider, MD, FRACP, FRCPA
Department of Medicine
Monash University
Melbourne, Australia
1. Orstavik RE, Haugeberg G, Mowinckel P, et al. Vertebral deformities in rheumatoid arthritis: a comparison with population-based controls. Arch Intern Med. 2004;164:420-425.
FREE FULL TEXT
2. van Staa TP, Cooper C, Brusse LS, Leufkens H, Javaid MK, Arden NK. Inflammatory bowel disease and the risk of fracture. Gastroenterology. 2003;125:1591-1597.
FULL TEXT
|
ISI
| PUBMED
3. Mitra D, Elvins DM, Speden DJ, Collins AJ. The prevalence of vertebral fractures in mild ankylosing spondylitis and their relationship to bone mineral density. Rheumatology (Oxford). 2000;39:85-89.
FREE FULL TEXT
4. Quinn JM, Gillespie MT. Modulation of osteoclast formation. Biochem Biophys Res Commun. 2005;328:739-745.
FULL TEXT
|
ISI
| PUBMED
5. Tanabe N, Ito-Kato E, Suzuki N, et al. IL-1 affects mineralized nodule formation by rat osteoblasts. Life Sci. 2004;75:2317-2327.
FULL TEXT
|
ISI
| PUBMED
6. Henry MJ, Pasco JA, Nicholson GC, Seeman E, Kotowicz MA. Prevalence of osteoporosis in Australian women: Geelong Osteoporosis Study. J Clin Densitom. 2000;3:261-268.
FULL TEXT
|
ISI
| PUBMED
7. Roberts WL, Moulton L, Law TC, et al. Evaluation of nine automated high-sensitivity C-reactive protein methods: implications for clinical and epidemiological applications: part 2. Clin Chem. 2001;47:418-425.
FREE FULL TEXT
8. Pasco JA, Henry MJ, Gaudry TM, Nicholson GC, Kotowicz MA. Identification of incident fractures: the Geelong Osteoporosis Study. Aust N Z J Med. 1999;29:203-206.
ISI
| PUBMED
9. Seeman E, Delmas PD. Bone quality—the material and structural basis of bone strength and fragility. N Engl J Med. 2006;354:2250-2261.
FREE FULL TEXT
Letters Section Editor: Robert M. Golub, MD, Senior Editor.
JAMA. 2006;296:1353-1355.
 CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati  Twitter
What's this?
RELATED LETTER
High-Sensitivity C-Reactive Protein and Prediction of Nontraumatic Fractures Reply
and
Arch Intern Med. ;167():2007-2008.
FULL TEXT
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
Circulating Levels of Inflammatory Markers Predict Change in Bone Mineral Density and Resorption in Older Adults: A Longitudinal Study
Ding et al.
J. Clin. Endocrinol. Metab. 2008;93:1952-1958.
ABSTRACT
| FULL TEXT
High-Sensitivity C-Reactive Protein and Prediction of Nontraumatic Fractures Reply
Schett and Kiechl
Arch Intern Med 2007;167:2007-2008.
FULL TEXT
What's new in the other general journals
Tonks
BMJ 2006;333:699-700.
FULL TEXT
|
