COLIA1 polymorphism contributes to bone mineral density to assess prevalent wrist fractures

Bone Vol. 26, No. 3 March 2000:287–290

COLIA1 Polymorphism Contributes to Bone Mineral Density to Assess Prevalent Wrist Fractures ´ ,1 J. J. SˇTEˇPA ´ N,1 D. MICHALSKA ´ ,1 T. HAAS,1 H. A. P. POLS,2 and M. WEICHETOVA 2 A. G. UITTERLINDEN 1

Department of Internal Medicine III, Charles University Faculty of Medicine, Prague, Czech Republic Department of Internal Medicine III, Erasmus University Medical School, Rotterdam, The Netherlands

2

26:287–290; 2000) rights reserved.

Wrist fractures associated with postmenopausal women are only partially explained by osteoporosis. Recent studies have shown that polymorphism of an Sp1 binding site in the first intron of the collagen I alpha 1 gene (COLIA1) may determine risk for vertebral and nonvertebral fractures in postmenopausal women independent of bone mass. We investigated the relationship between the COLIA1 polymorphism, lumbar spine and femoral neck bone mineral density (BMD), ultrasound stiffness of the heel, anthropometric variables, and risk for wrist fractures in 126 Czech postmenopausal women with low bone mass who suffered one or more wrist fracture in the last 5 years and in 126 postmenopausal women with low bone mass without any fracture. Genotypes for the Sp1 COLIA1 polymorphism were determined by polymerase chain reaction, digestion with BalI restriction enzyme, and agarose gel electrophoresis. The test discriminates two alleles, S and s, which correspond to the presence of guanine and thymidine, respectively, at the first bases in the Sp1-binding site in the first intron of the gene for COLIA1. No significant differences were found between the fracture and control group with regard to age, weight, and years since menopause. However, BMD of the lumbar spine and femoral neck and ultrasound stiffness of the heel were significantly lower in patients with prevalent wrist fracture. Femoral neck BMD was the strongest determinant of prevalent fracture of the wrist. COLIA1 genotyping significantly strengthened prediction of prevalent fracture of the wrist. After multivariate adjustment, women in the Ss group had 2.0 times the risk of the women in the SS group (95% confidence interval [CI] ⴝ 1.1–3.8), and the women in the ss group had 2.8 times the risk of the women in the SS group (95% CI ⴝ 0.5–14.6). The overall gene-dose effect was an odds ratio of 2.1 per copy of the “s” allele (95% CI ⴝ 1.2–3.8). In the stepwise logistic regression, COLIA1 acted synergistically with femoral neck BMD and weight in increasing prediction of wrist fracture. The results demonstrate that COLIA1 Sp1 polymorphism is associated with an increased risk of wrist fracture in postmenopausal women independent of BMD and may be helpful in clinical practice by identifying patients with an increased fracture risk. (Bone

Key Words: Bone density; Collagen I; Fracture; Genotype; Osteoporosis; Ultrasonometry. Introduction Bone mineral density (BMD) measurements have been used to identify patients at increased risk of osteoporotic fracture and to predict fracture risk. However, in individual patients, additional factors have to be used to improve the predictive value of BMD. Twin and family studies have shown that 50 – 85% of the population variance in BMD is under genetic determination.1,6,9 In epidemiological studies, several genetic factors that play a role in the pathogenesis of osteoporosis have been associated with BMD.15 Type I collagen is the major protein constituent in bone. Both structural mutations of the protein-coding region17 and sequence changes in regulatory regions of the type I collagen gene may predispose to osteoporosis. A guanine-to-thymidine substitution at a binding site for the transcription factor Sp1 in the collagen type I alpha 1 gene (COLIA1) has been associated with decreased BMD and with an increased risk for vertebral and nonvertebral fractures in the United Kingdom,5,10 France,4 The Netherlands,18 the United States,16 and Denmark.13 However, the Sp1 polymorphism at the COLIA1 gene was not found in Korean women,7 and the Sp1 polymorphisms in the COLIA1 gene was found unlikely to be of clinical value in identifying Swedish subjects who are at risk of postmenopausal osteoporosis.14 The aim of this study was to assess the association of the COLIA1 polymorphism with the risk of wrist fractures in Czech women after menopause and to investigate the predictive value of this genetic marker in comparison with BMD measurements and bone ultrasonography. Materials and Methods Study Subjects The association between the COLIA1 genotype and osteoporosis was evaluated in 126 Caucasian postmenopausal women with osteoporosis or osteopenia according to World Health Organization criteria20 who suffered one or more wrist fracture in the last 5 years, and in a control group of 126 postmenopausal women with osteoporosis or osteopenia, but without any fracture (Table

Address for correspondence and reprints: Dr. Jan Sˇteˇpa´n, 3rd Department of Internal Medicine, U nemocnice 1, CZ 128 00 Prague, Czech Republic. E-mail: [email protected] © 2000 by Elsevier Science Inc. All rights reserved.

© 2000 by Elsevier Science Inc. All

287

8756-3282/00/$20.00 PII S8756-3282(99)00280-X

288

M. Weichetova´ et al. COLIA1 polymorphism and wrist fractures

Bone Vol. 26, No. 3 March 2000:287–290

Table 1. Characteristics of the populations of postmenopausal women with wrist fracture and the control group

No. of patients Age (years) Time since menopause (years) Weight (kg) Height (cm) Lumbar spine BMD (T-score) Lumbar spine BMD (Z-score) Femoral neck BMD (T-score) Femoral neck BMD (Z-score) Stiffness (%) Genotype frequency SS n(%) Ss n(%) ss n(%) Allele frequency S (%) s (%)

Wrist fracture

Control group

p

126 62.2 ⫾ 6.0 13.7 ⫾ 6.7 66.2 ⫾ 11.1 164.1 ⫾ 5.8 ⫺2.25 ⫾ 1.17 ⫺0.73 ⫾ 1.19 ⫺2.44 ⫾ 0.82 ⫺0.84 ⫾ 0.74 73.9 ⫾ 12.7

126 60.8 ⫾ 5.8 12.8 ⫾ 6.1 64.7 ⫾ 8.4 162.5 ⫾ 5.7 ⫺1.97 ⫾ 0.97 ⫺0.64 ⫾ 0.97 ⫺1.69 ⫾ 1.10 ⫺0.46 ⫾ 0.93 78.3 ⫾ 11.9

0.06 0.27 0.22 0.02 0.04 0.27 ⬍0.001 ⬍0.001 0.005

79 (62) 40 (32) 7 (6)

94 (74) 30 (24) 2 (2)

0.068a

79 21

87 13

0.019a

Data are means ⫾ SD. a Chi square test, two-sided.

1). All patients were residents of the city of Prague. Patients with metabolic bone diseases other than osteoporosis and diseases known to affect bone metabolism (corticosteroid use, pituitary disease, hyperparathyroidism, neoplasia, and thyrotoxicosis) were excluded from the study. Postmenopausal women aged 45–70 years who experienced wrist fracture during the last 5 years and were treated at the University Hospital Department of Surgery were invited for screening. All the patients who responded to invitation and fulfilled the inclusion and exclusion criteria were enrolled in the fracture group. Control women were enrolled in sequence from postmenopausal outpatients aged 45–70 years examined for osteoporosis in this outpatient Osteoporosis Clinic. All patients from the control group who fulfilled the inclusion (age and menopausal status) and exclusion criteria were enrolled in the study.

reactions were performed in a DNA thermocycler (Cyclogene Dri-Block Cycler, Eppendorf, Hamburg, Germany). The cycling protocol involved initial denaturation step of 94°C for 5 min, 40 cycles of 94°C, 60°C, and 72°C for 1 min each, and a final extension step of 72°C for 5 min. Polymerase chain reaction products were digested with BalI restriction enzyme (Amersham, Little Chalfont, UK) for 2 h according to the manufacturer’s instructions, and were separated by electrophoresis on a ethidium bromide-stained 3% NuSieve agarose gel (Promega, Madison, WI) in a buffer (44.5 mmol/L Tris, 44.5 mmol/L boric acid, 1 mmol/L disodium EDTA) for 100 volt-hours. The separation patterns were documented by Polaroid photography under ultraviolet illumination (312 nm). The results were confirmed by repeated analysis on all samples, including those from all “ss” homozygotes.

Measurements

Statistical Methods

Height and weight were measured at the initial examination. Age at menopause was assessed by questionnaire. Wrist fractures were diagnosed by the X-ray and classified by a physician. Bone mineral density (in g/cm2) measurements were determined by dual-energy X-ray absorptiometry (QDR 4500; Hologic, Waltham, MA) at the femoral neck and lumbar spine (vertebrae L1– 4). The ultrasound stiffness of the heel was measured using the Achilles Plus Ultrasound Bone Densitometer (Lunar Corp., Madison, WI). The long-term coefficient of variation of lumbar spine BMD was 1.1%, femoral neck BMD was 1.4%, and stiffness was 3.9%. Genomic DNA was extracted from peripheral venous blood using a kit (Nucleon, Amersham, UK). The intronic polymorphism of the COLIA1 gene was detected by the polymerase chain reaction (PCR) with a mismatched primer that introduces a diallelic restriction site.5 The test discriminates two alleles, S and s, which correspond to the presence of guanine and thymidine, respectively, at the first bases in the Sp1-binding site in the first intron of the gene for COLIA1. The reaction mixture of 25 ␮L contained 100 ng of genomic DNA, 1.0 U of Taq polymerase (Promega, Madison, WI), 2.5 ␮L of DNA polymerase 10⫻ buffer (100 mmol/L Tris HCl, 500 mmol/L KCl, 1% Triton X-100), 5 nmol of dideoxynucleotide triphosphatase, 37.5 nmol of magnesium chloride, and 0.004 fmol of each primer. The

Quantitative variables from the different groups were compared using two-sample t test and categorical variables using Chi square test. The COLIA1 genotypes were assigned codes of 0 for SS, 1 for Ss, or 2 for ss genotypes, according to the number of alleles present. The linearity assumption for weight failed; instead, a design variable weight group (⬍59, 60 – 64, 64 –70, ⬎70 kg, coded as 0, 1, 2, 3, respectively) was used. A multivariate logistic-regression analysis was used to test for interaction of BMD measurements, ultrasonometry, and polymorphism of the COLIA1 in predicting wrist fractures and to adjust the variables for confounding factors such as age, time since menopause, and anthropometric measurements. Odds ratios (with 95% confidence intervals [CI]) were calculated by multivariate logisticregression analysis to analyze predictors of wrist fractures. Hosmer-Lemeshow goodness-of-fit statistics was used to assess the fit of the model.8 The likelihood ratio test (G-statistic) was used to check for the significance of the addition of new terms to the model. Results Table 1 gives characteristics of the patients’ population. No significant differences were found between the fracture and control groups with regard to the age, height, and years since

Bone Vol. 26, No. 3 March 2000:287–290

M. Weichetova´ et al. COLIA1 polymorphism and wrist fractures

289

Table 2. Odds ratios for prevalent wrist fracture according to femoral neck and lumbar spine BMD, ultrasound heel stiffness, and COLIA1 genotype Multivariatea

Age-adjusted Variable Femoral neck BMD Lumbar spine BMD Stiffness COLIA1 genotype SS Ss ss per copy of the “s” allele

Odds ratio

95% confidence intervals

Odds ratio

95% confidence intervals

0.5 0.8 1.0

0.3–0.6 0.6–1.1 0.9–1.0

0.4 1.0 1.0

0.3–0.5 0.7–1.3 0.97–1.02

1.0 1.5 3.7 1.7

0.9–2.7 0.7–18.5 1.0–2.9

1.0 2.0 2.6 2.1

1.1–3.9 0.5–13.7 1.1–3.9

a Odds ratios for each variable have been adjusted for each of the remaining variables (age, years since menopause, weight, height, and femoral neck BMD, lumbar spine BMD, stiffness, or COLIA1, respectively).

menopause. Patients with fractures were slightly taller compared with controls. Genotype frequencies observed in the total group were similar to those previously observed and were in HardyWeinberg equilibrium: SS 68.7%, Ss 27.8%, and ss 3.5%. Bone mineral density of the lumbar spine and femoral neck and ultrasound stiffness of the heel were significantly lower in patients with prevalent wrist fracture. Table 1 also shows the distribution of prevalent wrist fractures according to COLIA1 genotype. There was a significant overrepresentation of the “s” allele in wrist fracture cases. Femoral neck BMD was the strongest predictor of prevalent fracture of the wrist. In the logistic stepwise regression model, the odds ratio for prevalent wrist fracture was significantly increased based on femoral neck T-score alone and was further strengthened by COLIA1 genotype and weight. None of the remaining variables (age, height, lumbar spine BMD, ultrasound stiffness, and years since menopause) entered the regression model. The odds ratios for prevalent fracture were of borderline significance when assessed using calcaneal stiffness. However, the model using femoral neck, COLIA1 genotype, and weight design variables was not strengthened by adding stiffness (p ⫽ 0.58), and was significantly weakened by replacing COLIA1 genotype with stiffness. Logistic-regression analysis showed a gene-dose effect for COLIA1 genotypes Ss and ss. For prevalent wrist fractures, the women in the Ss group had 1.5 times the risk of the women in the SS group, and the women in the ss group had 3.7 times the risk of the women in the SS group (Table 2). The odds ratio per copy of the “s” allele was 1.7. Adjustment for femoral and lumbar BMD, calcaneal stiffness, years since menopause, and anthropometric measurements only slightly strengthened the association between COLIA1 genotypes and the risk of wrist fractures. Discussion The incidence of wrist fractures increases rapidly in the first 5 years after menopause, reaching a peak between ages of 60 –70 years. The wrist fractures are only partially explained by osteoporosis.2 From cross-sectional studies, the difference of BMD between patients with fracture of the distal end of the forearm and controls was about 0.5 SD,12 whereas the difference measured at the femoral neck or trochanter was between 0.9 and 1 SD.11 Thus, the wrist fractures seem to be slightly less osteoporotic than the hip fractures. The polymorphism, which affects a binding site for the transcription factor Sp1 in a regulatory region of the COLIA1 gene, was associated with risk of fracture in British,5,10 Dutch,18

and Danish13 postmenopausal women. The studies showed that COLIA1 genotyping predicted fracture independent of bone mass. In a cross-sectional study of Dutch women, the overall gene-dose effect was demonstrated to be higher for hip fractures (odds ratio of 3.1, 10 cases) than for wrist fracture (odds ratio of 1.4, 36 cases).18 However, in a nested case-control study, the “s” allele was overrepresented among hip fracture cases (age-adjusted odds ratio of 1.7) but not among vertebral and wrist fracture cases.19 A nested case-control study in the women of the United Kingdom indicated a consistent association between the COLIA1 genotype and increased risk for any fracture.10 The increased fracture risk associated with the “s” allele appeared largely independent of BMD, suggesting that the COLIA1 locus may also have effects on bone structure or quality.10 The present data indicate that fractures of the distal end of the forearm can be considered an early manifestation of an unfavorable genotype in women with a slight postmenopausal bone loss rather than manifestation of osteoporosis. It is supposed that the reduced mechanical strength of bone may be affected by increased production of alpha 1 procollagen relative to alpha 2.3 Our data confirm the overrepresentation of the “s” allele among women with wrist fractures.18 In this study, the odds ratio for prevalent wrist fracture per copy of the “s” allele was 2.1 after multivariate adjustment, whereas the odds ratio for prevalent wrist fracture per 1 SD decrease in femoral neck BMD was 2.6. The statistical power that this study had to detect a difference in the “s” allele distribution between cases (38%) and controls (26%) was low (one-sided test, 66%), though statistically significant (Chi square, p ⫽ 0.042). When compared with other populations, the associations of genotype with a selection criterium phenotype (i.e., fractures) seem very similar. However, in our analysis, low number of subjects and the design of the study did not allow conclusions regarding relationships between the COLIA1 genotype and BMD demonstrated by other studies.4,5,10,18 This study had some important limitations. First, we studied only older, Caucasian females. The ability of genotype to predict fractures may vary across gender, race, and age groups. Second, despite our attempts to avoid preselection of the individuals, this study was just a case-control one. Third, we studied only Colles’ fractures. Finally, prospective studies will be necessary to confirm the clinical value of the genetic marker. In conclusion, we have demonstrated that COLIA1 genotyping significantly strengthened determination of prevalent fracture of the wrist by decrease in femoral neck BMD, indicating that simultaneous measurements of COLIA1 Sp1 polymorphism and femoral neck BMD may help to identify postmenopausal women

290

M. Weichetova´ et al. COLIA1 polymorphism and wrist fractures

with an increased risk for wrist fractures. Our data support the view that COLIA1 may be a candidate gene for regulation of bone quality,10,18 but our results must be treated with caution, in view of the small number of ss individuals, and will require confirmation in larger prospective studies.

Acknowledgments: The authors thank Ing. Marek Michalsky, Jana Krenkova, Anna Masatova, and Vlasta Kubova for their expert technical assistance.

References 1. Arden, N. K., Baker, K., Hogg, B., Baan, K., and Spector, T. D. The heritability of bone mineral density, ultrasound of the calcaneus and hip axis length: a study of postmenopausal twins. J Bone Miner Res 11:530 –534; 1996. 2. Black, D. M., Arden, N. K., Palermo, L., Pearson, J., and Cummings, S., for the Study of osteoporotic fractures research group. Prevalent vertebral deformities predict hip fractures and new vertebral deformities but not wrist fractures. J Bone Miner Res 14:821– 828; 1999. 3. Dean, V., Hobson, E. E., Aspden, R. M., Robins, S. P., and Ralston, S. H. Relationship between COLIA1 Sp1 alleles, gene transcription, collagen production and bone strength. Bone 23(Suppl.):161; 1998. 4. Garnero, P., Borel, O., Grant, S. F., Ralston, S. H., and Delmas, P. D. Collegen I alpha1 Sp1 polymorphism, bone mass, and bone turnover in healthy French premenopausal women: the OFELY study. J Bone Miner Res 13:813– 817; 1998. 5. Grant, S. F. A., Reid, D. M., Blake, G., Herd, R., Fogelman, I., and Ralston, S. H. Reduced bone density and osteoporosis associated with a polymorphic Sp1 site in the collagen type I alpha 1 gene. Nature Genet 14:203–205; 1996. 6. Gue´guen, R., Jouanny, P., Guillemin, F., Kuntz, C., Pourel, J., and Siest, G. Segregation analysis and variance components analysis of bone mineral density in healthy families. J Bone Miner Res 10:2017–2022; 1995. 7. Han, K. O., Moon, I. G., Hwang, C. S., Choi, J. T., Yoon, H. K., Min, H. K., Han, I. K. Lack of an intronic Sp1 binding-site polymorphism at the collagen type I alpha1 gene in healthy Korean women. Bone 24:135–137; 1999. 8. Hosmer, D. W. and Lemeshow, S. Applied Logistic Regression. New York: John Wiley & Sons; 1989. 9. Jouanny, P., Guillemin, F., Kuntz, C., Jeandel, C., and Pourel, J. Environmental and genetic factors affecting bone mass: similarity of bone density among members of healthy families. Arthritis Rheum 38:61– 67; 1995.

Bone Vol. 26, No. 3 March 2000:287–290 10. Keen, R. W., Woodford-Richens, K. L., Grant, S. F., Ralston, S. H., Lanchbury, J. S., and Spector, T. D. Association of polymorphism at the type I collagen (COL1A1) locus with reduced bone mineral density, increased fracture risk, and increased collagen turnover. Arthritis Rheum 42:285–290; 1999. 11. Kelsey, J. L., Browner, W. S., Seeley, D. G., Nevitt, M. C., and Cummings, S. R. Risk factors for fractures of the distal forearm and proximal humerus. Am J Epidemiol 135:477– 479; 1992. 12. Krølner, B., Tondevol, E., Toft, B., Berthelsen, B., and Pors Nielsen, S. Bone mass of the axial and the appendicular skeleton in women with Colles’ fracture: its relation to physical activity. Clin Physiol 2:147–157; 1982. 13. Langdahl, B. L., Ralston, S. H., Grant, S. F., Eriksen, E. F. An Sp1 binding site polymorphism in the COLIA1 gene predicts osteoporotic fractures in both men and women. J Bone Miner Res 13:1384 –1389; 1998. 14. Liden, M., Wilen, B., Ljunghall, S., and Melhus, H. Polymorphism at the Sp 1 binding site in the collagen type I alpha 1 gene does not predict bone mineral density in postmenopausal women in Sweden. Calcif Tissue Int 63:293–295; 1998. 15. Ralston, S. H. The genetics of osteoporosis. Quart J Med 90:247–251; 1997. 16. Sainz, J., Van Tornout, J. M., Sayre, J., Kaufman, F., and Gilsanz, V. Association of collagen type 1 alpha 1 gene polymorphism with bone density in early childhood. J Clin Endocrinol Metab 84:853– 855; 1999. 17. Spotila, L. D., Colige, A., Sereda, L., Constantinou-Deltas, C. D., Whyte, M. P., Riggs, B. L., Shaker, J. L., Spector, T. D., Hume, E., Olsen, N., Attie, M., Tenenhouse, A., Shane, E., Briney, W., and Prockop, D. J. Mutation analysis of coding sequences for type I procollagen in individuals with low bone density. J Bone Miner Res 9:923–932; 1994. 18. Uitterlinden, A. G., Burger, H., Huang, Q., Yue, F., McGuigan, F. E. A., Grant, S. F. A., Hofman, A., van Leeuwen, J. P. T. M., Pols, H. A. P., and Ralston, S. H. Relation of alleles of the collagen type I␣1 gene to bone density and the risk of osteoporotic fractures in postmenopausal women. New Engl J Med 338:1016 –1021; 1998. 19. Uitterlinden, A. G., Yue, F., Ralston, S. H., van Leeuwen, J. P., and Pols, H. A. The collagen type I polymorphism predicts hip fracture in women. Bone 23(Suppl.):161; 1998. 20. WHO Study Group. Assessment of Fracture Risk and Its Application to Screening for Postmenopausal Osteoporosis. WHO Technical Reports Series 843, Geneva: World Health Organization; 1994.

Date Received: January 8, 1999 Date Revised: October 25, 1999 Date Accepted: October 25, 1999