Advertisement

Osteoporosis International

, Volume 28, Issue 3, pp 819–832 | Cite as

Clinical fracture risk evaluated by hierarchical agglomerative clustering

  • C. Kruse
  • P. Eiken
  • P. Vestergaard
Original Article

Abstract

Summary

Clustering analysis can identify subgroups of patients based on similarities of traits. From data on 10,775 subjects, we document nine patient clusters of different fracture risks. Differences emerged after age 60 and treatment compliance differed by hip and lumbar spine bone mineral density profiles.

Introduction

The purposes of this study were to establish and quantify patient clusters of high, average and low fracture risk using an unsupervised machine learning algorithm.

Methods

Regional and national Danish patient data on dual-energy X-ray absorptiometry (DXA) scans, medication reimbursement, primary healthcare sector use and comorbidity of female subjects were combined. Standardized variable means, Euclidean distances and Ward’s D2 method of hierarchical agglomerative clustering (HAC), were used to form the clustering object. K number of clusters was selected with the lowest cluster containing less than 250 subjects. Clusters were identified as high, average or low fracture risk based on bone mineral density (BMD) characteristics. Cluster-based descriptive statistics and relative Z-scores for variable means were computed.

Results

Ten thousand seven hundred seventy-five women were included in this study. Nine (k = 9) clusters were identified. Four clusters (n = 2886) were identified based on low to very low BMD with differences in comorbidity, anthropometrics and future bisphosphonate compliance. Two clusters of younger subjects (n = 1058, mean ages 30 and 51 years) were identified as low fracture risk with high to very high BMD. A mean age of 60 years was the earliest that allowed for separation of high-risk clusters. DXA scan results could identify high-risk subjects with different antiresorptive treatment compliance levels based on similarities and differences in lumbar spine and hip region BMD.

Conclusions

Unsupervised HAC presents a novel technology to improve patient characteristics in bone disease beyond traditional T-score-based diagnosis. Technological and validation limitations need to be overcome to improve its use in internal medicine. Current DXA scan indication guidelines could be further improved by clustering algorithms.

Keywords

Clustering Densitometry Machine learning Osteoporosis Risk factors 

Notes

Acknowledgements

The Obel Family Foundation and the Department of Clinical Medicine at Aalborg University are acknowledged for providing grants that enable the PhD fellowship of Dr. Christian Kruse. Statistics Denmark is acknowledged for providing data. The community around the R statistical software is acknowledged for the programming packages and guidance that enable studies such as this.

Compliance with ethical standards

Conflicts of Interest

CK has received travel grants from Eli Lilly, Otsuka Pharmaceutical, and is a speaker for Novartis and Otsuka Pharmaceutical. PE is an advisory board member with Amgen, MSD and Eli Lilly and at the speakers bureau with Amgen and Eli Lilly, stocks from Novo Nordisk A/S. PV has received unrestricted grants from MSD and Servier, and travel grants from Amgen, Eli Lilly, Novartis, Sanofi-Aventis, and Servier.

Supplementary material

198_2016_3828_MOESM1_ESM.pdf (156 kb)
ESM 1 (PDF 156 kb)

References

  1. 1.
    Genant HK, Cooper C, Poor G et al (1999) Interim report and recommendations of the World Health Organization task-force for osteoporosis. Osteoporos Int 10(4):259–264CrossRefPubMedGoogle Scholar
  2. 2.
    Kanis JA, Johnell O, Oden A, Jonsson B, De laet C, Dawson A (2000) Risk of hip fracture according to the World Health Organization criteria for osteopenia and osteoporosis. Bone 27(5):585–590CrossRefPubMedGoogle Scholar
  3. 3.
    Kanis JA (1994) Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: synopsis of a WHO report. WHO study group. Osteoporos Int 4(6):368–381CrossRefPubMedGoogle Scholar
  4. 4.
    Sandor T, Felsenberg D, Brown E (1999) Comments on the hypotheses underlying fracture risk assessment in osteoporosis as proposed by the World Health Organization. Calcif Tissue Int 64(3):267–270CrossRefPubMedGoogle Scholar
  5. 5.
    De laet C, Kanis JA, Odén A et al (2005) Body mass index as a predictor of fracture risk: a meta-analysis. Osteoporos Int 16(11):1330–1338CrossRefPubMedGoogle Scholar
  6. 6.
    Mattson RH, Gidal BE (2004) Fractures, epilepsy, and antiepileptic drugs. Epilepsy Behav 5(Suppl 2):S36–S40CrossRefPubMedGoogle Scholar
  7. 7.
    Richards JB, Papaioannou A, Adachi JD, Joseph L, Whitson HE, Prior JC, Goltzman D (2007) Effect of selective serotonin reuptake inhibitors on the risk of fracture. Arch Intern Med 167(2):188–194CrossRefPubMedGoogle Scholar
  8. 8.
    Vestergaard P (2007) Discrepancies in bone mineral density and fracture risk in patients with type 1 and type 2 diabetes—a meta-analysis. Osteoporos Int 18(4):427–444CrossRefPubMedGoogle Scholar
  9. 9.
    Vestergaard P, Rejnmark L, Mosekilde L (2006) Proton pump inhibitors, histamine H2 receptor antagonists, and other antacid medications and the risk of fracture. Calcif Tissue Int 79(2):76–83CrossRefPubMedGoogle Scholar
  10. 10.
    Vestergaard P, Rejnmark L, Mosekilde L (2004) Fracture risk associated with use of antiepileptic drugs. Epilepsia 45(11):1330–1337CrossRefPubMedGoogle Scholar
  11. 11.
    Schwartz AV, Sellmeyer DE, Ensrud KE et al (2001) Older women with diabetes have an increased risk of fracture: a prospective study. J Clin Endocrinol Metab 86(1):32–38CrossRefPubMedGoogle Scholar
  12. 12.
    Ao SI, Yip K, Ng M et al (2005) CLUSTAG: hierarchical clustering and graph methods for selecting tag SNPs. Bioinformatics 21(8):1735–1736CrossRefPubMedGoogle Scholar
  13. 13.
    Agnelli L, Mosca L, Fabris S, Lionetti M, Andronache A, Kwee I, Todoerti K, Verdelli D, Battaglia C, Bertoni F, Deliliers GL, Neri A (2009) A SNP microarray and FISH-based procedure to detect allelic imbalances in multiple myeloma: an integrated genomics approach reveals a wide gene dosage effect. Genes Chromosomes Cancer 48(7):603–614. doi: 10.1002/gcc.20668 CrossRefPubMedGoogle Scholar
  14. 14.
    Cotsapas C, Voight BF, Rossin E, Lage K, Neale BM, Wallace C, Abecasis GR, Barrett JC, Behrens T, Cho J, De Jager PL, Elder JT, Graham RR, Gregersen P, Klareskog L, Siminovitch KA, van Heel DA, Wijmenga C, Worthington J, Todd JA, Hafler DA, Rich SS, Daly MJ (2011) FOCiS network of consortia. Pervasive sharing of genetic effects in autoimmune disease. PLoS Genet 7(8):e1002254. doi: 10.1371/journal.pgen.1002254 CrossRefPubMedPubMedCentralGoogle Scholar
  15. 15.
    Linnaeus C (1758) Systema naturae per regna tria naturae : secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locisGoogle Scholar
  16. 16.
    Rousseeuw PJ (1987) Silhouettes: a graphical aid to the interpretation and validation of cluster analysis. J Comput Appl Math 20:53–65CrossRefGoogle Scholar
  17. 17.
    Tibshirani R, Walther G, Hastie T (2001) Estimating the number of clusters in a data set via the gap statistic. J R Stat Soc Ser B Stat Methodol 63(2):411–423CrossRefGoogle Scholar
  18. 18.
    Kruse C, Eiken P, Vestergaard P (2015) Hyponatremia and osteoporosis: insights from the Danish National Patient Registry. Osteoporos Int 26(3):1005–1016CrossRefPubMedGoogle Scholar
  19. 19.
    Kruse C, Eiken P, Verbalis J, Vestergaard P (2016) The effect of chronic mild hyponatremia on bone mineral loss evaluated by retrospective national Danish patient data. Bone 84:9–14CrossRefPubMedGoogle Scholar
  20. 20.
  21. 21.
    Charlson M, Szatrowski TP, Peterson J, Gold J (1994) Validation of a combined comorbidity index. J Clin Epidemiol 47(11):1245–1251CrossRefPubMedGoogle Scholar
  22. 22.
    Sundararajan V, Henderson T, Perry C, Muggivan A, Quan H, Ghali WA (2004) New ICD-10 version of the Charlson comorbidity index predicted in-hospital mortality. J Clin Epidemiol 57(12):1288–1294CrossRefPubMedGoogle Scholar
  23. 23.
    Ward JH (1963) Hierarchical grouping to optimize an objective function. J Am Stat Assoc 58(301):236–244CrossRefGoogle Scholar
  24. 24.
    Murtagh F (2014) Ward’s hierarchical agglomerative clustering method:which algorithms implement Ward’s criterion? J Classif 31:274–295CrossRefGoogle Scholar
  25. 25.
    Warming L, Hassager C, Christiansen C (2002) Changes in bone mineral density with age in men and women: a longitudinal study. Osteoporos Int 13(2):105–112CrossRefPubMedGoogle Scholar
  26. 26.
    Rannevik G, Jeppsson S, Johnell O, Bjerre B, Laurell-Borulf Y, Svanberg L (2008) A longitudinal study of the perimenopausal transition: altered profiles of steroid and pituitary hormones, SHBG and bone mineral density. Maturitas 61(1–2):67–77CrossRefPubMedGoogle Scholar
  27. 27.
    Liberman UA, Weiss SR, Bröll J, Minne HW, Quan H, Bell NH, Rodriguez-Portales J, Downs RW Jr, Dequeker J, Favus M (1995) Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. The alendronate phase III osteoporosis treatment study group. N Engl J Med 333(22):1437–1443CrossRefPubMedGoogle Scholar
  28. 28.
    Burger H, van Daele PL, Odding E, Valkenburg HA, Hofman A, Grobbee DE, Schütte HE, Birkenhäger JC, Pols HA (1996) Association of radiographically evident osteoarthritis with higher bone mineral density and increased bone loss with age. The Rotterdam study. Arthritis Rheum 39(1):81–86CrossRefPubMedGoogle Scholar
  29. 29.
    Liu G, Peacock M, Eilam O, Dorulla G, Braunstein E, Johnston CC (1997) Effect of osteoarthritis in the lumbar spine and hip on bone mineral density and diagnosis of osteoporosis in elderly men and women. Osteoporos Int 7(6):564–569CrossRefPubMedGoogle Scholar
  30. 30.
    Lane NE, Oehlert JW, Bloch DA, Fries JF (1998) The relationship of running to osteoarthritis of the knee and hip and bone mineral density of the lumbar spine: a 9-year longitudinal study. J Rheumatol 25(2):334–341PubMedGoogle Scholar
  31. 31.
    Mäkinen TJ, Alm JJ, Laine H, Svedström E, Aro HT (2007) The incidence of osteopenia and osteoporosis in women with hip osteoarthritis scheduled for cementless total joint replacement. Bone 40(4):1041–1047CrossRefPubMedGoogle Scholar
  32. 32.
    Antoniades L, MacGregor AJ, Matson M, Spector TD (2000) A cotwin control study of the relationship between hip osteoarthritis and bone mineral density. Arthritis Rheum 43(7):1450–1455CrossRefPubMedGoogle Scholar
  33. 33.
    Chilibeck PD, Sale DG, Webber CE (1995) Exercise and bone mineral density. Sports Med 19(2):103–122CrossRefPubMedGoogle Scholar
  34. 34.
    Suominen H (1993) Bone mineral density and long term exercise. An overview of cross-sectional athlete studies. Sports Med 16(5):316–330CrossRefPubMedGoogle Scholar
  35. 35.
    Lane NE, Bloch DA, Jones HH, Marshall WH Jr, Wood PD, Fries JF (1986) Long-distance running, bone density, and osteoarthritis. JAMA 255(9):1147–1151CrossRefPubMedGoogle Scholar
  36. 36.
    ISCD 2015 Official position. http://www.iscd.org/official-positions/2015-iscd-official-positions-adult/. Accessed 15 June 2016
  37. 37.
    Kanis JA, Johnell O, Oden A, Johansson H, McCloskey E (2008) FRAX and the assessment of fracture probability in men and women from the UK. Osteoporos Int 19(4):385–397. doi: 10.1007/s00198-007-0543-5 CrossRefPubMedPubMedCentralGoogle Scholar
  38. 38.
    Leslie WD, Kovacs CS, Olszynski WP, Towheed T, Kaiser SM, Prior JC, Josse RG, Jamal SA, Kreiger N, Goltzman D (2011) CaMos research group. Spine-hip T-score difference predicts major osteoporotic fracture risk independent of FRAX(®): a population-based report from CAMOS. J Clin Densitom 14(3):286–293. doi: 10.1016/j.jocd.2011.04.011 CrossRefPubMedPubMedCentralGoogle Scholar
  39. 39.
    van den Bergh JP, van Geel TA, Lems WF, Geusens PP (2010) Assessment of individual fracture risk: FRAX and beyond. Curr Osteoporos Rep 8(3):131–137. doi: 10.1007/s11914-010-0022-3 CrossRefPubMedPubMedCentralGoogle Scholar
  40. 40.
    Hamdy RC, Kiebzak GM (2009) Variance in 10-year fracture risk calculated with and without T-scores in select subgroups of normal and osteoporotic patients. J Clin Densitom 12(2):158–161. doi: 10.1016/j.jocd.2008.12.003 CrossRefPubMedGoogle Scholar
  41. 41.
    Reynolds A, Richards G, de la Iglesia B, Rayward-Smith V (1992) Clustering rules: a comparison of partitioning and hierarchical clustering algorithms. Journal of Mathematical Modelling and Algorithms 5:475–504CrossRefGoogle Scholar
  42. 42.
    Müllner D (2013) Fastcluster: fast hierarchical, agglomerative clustering routines for R and python. J Stat Softw 53(9):1–18CrossRefGoogle Scholar
  43. 43.
    Kaufman L, Rousseeuw PJ (1990) (=: “K&R(1990)”) finding groups in data: an introduction to cluster analysis. Wiley, New YorkGoogle Scholar
  44. 44.
    Houle ME, Kriegel H, Kröger P, Schubert E, Zimek A (2010) Can shared-neighbor distances defeat the curse of dimensionality? Proceedings of the 22nd International Conference on Scientific and Statistical Database Management. Heidelberg, GermanyGoogle Scholar
  45. 45.
    Suzuki R, Shimodaira H (2006) Pvclust: an R package for assessing the uncertainty in hierarchical clustering. Bioinformatics, Oxford Univ PressGoogle Scholar

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2016

Authors and Affiliations

  1. 1.Department of EndocrinologyAalborg University HospitalAalborgDenmark
  2. 2.Department of Clinical MedicineAalborg University HospitalAalborgDenmark
  3. 3.Department of Cardiology, Nephrology and EndocrinologyNordsjaellands HospitalHilleroedDenmark
  4. 4.Faculty of Health and Medical SciencesUniversity of CopenhagenCopenhagenDenmark

Personalised recommendations