Skip to main content
NCBI home page
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2019 Feb 1.
Published in final edited form as: J Bone Miner Res. 2018 May 3;33(8):1417–1421. doi: 10.1002/jbmr.3435

Differential mortality and the excess rates of hip fracture associated with type 2 diabetes: accounting for competing risks in fracture prediction matters

C Tebé 1,2,3, D Martinez-Laguna 4,5, V Moreno 2,3,4,5,6, C Cooper 7,8, A Diez-Perez 9, GS Collins 7,8,9,10, D Prieto-Alhambra 4,7,8,9,10,✉
PMCID: PMC6108421  EMSID: EMS78015  PMID: 29624726

Abstract

Type 2 diabetes (T2DM) is associated with a reduced life expectancy. Latest published evidence suggests an increased risk of fractures among T2DM patients. We conducted a population-based cohort study to determine the impact of mortality as a competing risk in the study of the association between T2DM and hip fracture rates. Participants were all diagnosed T2DM patients registered in the SIDIAP database aged 65 years and older; up to two non-T2DM were matched by age, sex, and primary care facility. We used Cox regression models to estimate cause-specific Hazard Ratio (HR) of death or hip fracture according to T2DM status. Fine and Gray models were then fitted to estimate the Subhazard Ratio (SHR) of hip fracture while accounting for competing risk with death and to estimate the probability of hip fracture within 5 years. 55,891 T2DM and 103,093 matched non-T2DM patients were observed for a median of 8 years. Mortality was 48.8 per 1000 person years (py) in T2DM, and 33.8per 1,000 py in non-T2DM; hip fracture rates were 6.0 per 1,000 py and 4.9per 1,000 py respectively. Cox models confirmed a significant association for death and hip fracture: HR=1.51 [95% CI 1.48 to 1.55], and HR=1.32 [95% CI 1.24 to 1.40] respectively. Accounting for death as a competing event (Fine-Gray models), the association between T2DM and hip fracture risk remained statistically significant (sHR=1.15 [95% CI 1.09 to 1.21]) and the probability of a hip fracture within 5 years was 2.3% for TD2M and 1.9% for non-TD2M patients compared to 2.6% and 2.1% respectively using KM estimates. T2DM patients have a 50% increased mortality and, after adjusting for differential survival at 5 years, a 21% increased incidence of hip fracture when compared to matched non-T2DM. Failing to account for differential mortality leads to an overestimation of fracture risk.

Keywords: Fracture Risk Assessment, Type 2 Diabetes Mellitus, Competing Risk, Epidemilogy and Osteoporosis

Introduction

Diabetes is a leading underlying or contributing cause of death in high-income countries1 and is associated with a known reduction in life expectancy2. Moreover, diabetes is a long-term increasingly prevalent condition in both developed and developing countries. People with diabetes are at a greater risk of developing cardiovascular diseases, sight loss, foot and leg amputation and renal failure requiring dialysis or transplantation3,4. If diabetes is poorly controlled, the risk of developing these diseases and suffering severe consequences is higher5,6. A meta-analysis of studies of mortality in type 2 diabetes (T2DM) patients showed an 85% increase risk for all-cause mortality, a 76% excess cardiovascular mortality and over a 2-fold increase risk for stroke7. In the US alone, diabetes is reported to account for an average of 3.3 to 18.7 years lifespan loss8.

Published evidence suggests an association between diabetes and bone fragility9. Despite diabetes being associated with higher bone mineral density (BMD)10, a systematic review and meta-analysis concluded that type 2 diabetics have a 40% excess risk of hip fracture compared to non-diabetic peers11. Recent evidence, mostly published in the last 10-15 years, reinforces these findings12–16. Furthermore, the time from diabetes onset (metabolic control) appears to contribute to this association and T2DM complications such as neuropathy, nephropathy, and visual impairment or cataracts, are associated with an increased number of falls and related fractures17.

Hip fractures are the most devastating consequence of osteoporosis and bone fragility, and their incidence increases markedly with age18–20. Death and fracture are clearly competing events that cluster in the elderly, and T2DM is reportedly a risk factor for both. Standard methods, such as Kaplan-Meier or Cox regression fail to account for competing events, and alternative analytical approaches are required21. However, most of the literature on the association between type 2 diabetes and hip fractures has failed to account for a competing risk with death. We therefore used a large population-based computerized records database to study the impact of differential mortality on the association between type 2 diabetes and hip fracture risk.

Materials and Methods

We conducted a population-based cohort study using data from the SIDIAP Database (www.sidiap.org). SIDIAP contains clinical information from primary care records, hospital admissions, and pharmacy invoice data for >5 million patients (80% of the population) in Catalonia, Spain22.

Eligible participants were all those diagnosed with T2DM registered in SIDIAP and ≥65 years old on 1 January 2006 (“the index date”). Up to two non-T2DM were matched to each of the participants in the T2DM cohort by age (+/- 2 years), sex, and primary care practice. Non-T2DM patients that were users of anti-T2DM drugs or with two measures of glycated haemoglobin above 6.5% were excluded. Non-T2DM and T2DM patients with a recorded hip fracture or death on index date were excluded (as they did not contribute to observation time). Patients (either T2DM or non-T2DM) that ended unmatched (i.e. with no matched peer remaining) because of the above exclusion criteria were also excluded.

Both T2DM and matched non-T2DM cohorts were followed from index date until death, fractured a hip, or transferred out of catchment area or end of study on 12/31/2013, whichever came first (treated as censored). The main outcome was incident hip fracture and all-cause mortality was a secondary (competing) event. Previously validated lists of ICD10 codes22,23 were used to identify both the study exposure (T2DM status) and outcome/s (fracture and death). SIDIAP is a validated source of information for epidemiological research22. Potential confounders considered for adjustment in the multivariable models were those with a raw statistical significant association and a Hazard Ratio (HR) above 1.1 or below 0.9 (see directed acyclic graphs for hip fracture and death at appendix 1).

Baseline characteristics of T2DM and non-T2DM patients were described using mean and standard deviation for continuous variables and frequencies for categorical variables. We used Cox regression24 models stratified by match sets to estimate cause-specific hazards (Hazard Ratio –HR– and 95% Confidence Intervals –95%CI–) in T2DM compared to non-T2DM patients. Two cause-specific hazards were studied: hip fracture and death. Similarly, Cox regression models were then used to estimate (matched) multivariable adjusted cause-specific hazards according to T2DM status. Finally, Fine and Gray modelling25 was used to estimate risk prediction whilst illustrating the effect of competing risk with death. Cause-specific HR are reported from the Cox model; as are Subhazard ratio (SHR) and cumulative incidence function (CIF) from the Fine and Gray regressions26. We tested the PH assumption graphically with the log-log plot of survival and analytically with the test of proportional-hazards assumption. All statistical analyses were conducted using STATA v13.

Results

In total, 55,891 T2DM patients and 103,093 matched non-T2DM met inclusion criteria (Figure 1). T2DM patients were observed for a median of 6.4 years comprising 358,428 person-years of observation and non-T2DM patients were observed for a median of 8.0 years making up to a total of 695,340 person-years of observation.

Figure 1.

Figure 1

Open in a new tab

Population Flow-chart

Baseline characteristics for study participants stratified by T2DM status are presented in Table 1. T2DM patients had a higher prevalence of cardiovascular disease, neuropathy, and nephropathy. No discernible differences were observed for history of previous osteoporotic fracture, but non-T2DM subjects were more often users of calcium, vitamin D, and anti-osteoporotic treatments in the year before inclusion.

Table 1.

Baseline characteristics of the SIDIAP cohort, according to T2DM status.

non-T2DM T2DM patients
N 103,093 55,891
mean sd mean sd
Age (years) 74.68 6.52 74.63 6.46
BMI (kg/m2) 28.70 4.51 29.91 4.84
n % n %
Sex: Men 45094 43.74 24413 43.68
Smoking status: Never Smoker 29442 28.56 20708 37.05
Former Smoker 7822 7.59 5503 9.85
Current Smoker 4914 4.77 2834 5.07
Missing 60915 59.09 26846 48.03
Previous hip fracture 220 0.21 142 0.25
Previous major fracture 282 0.27 160 0.29
Previous osteoarthritis 23483 22.78 13367 23.92
Previous CVA 5466 5.30 4752 8.50
Previous IHD 6681 6.48 7405 13.25
Previous nephropathy 12137 11.77 11084 19.83
Previous nuropathy 14 0.01 931 1.67
Previous Falls 883 0.86 668 1.20
Corticoids prescriptions 6749 6.55 3816 6.83
Calcium + vitamin D 13686 13.28 6108 10.93
Anti-Osteporosis drugs 12271 11.90 5162 9.24
Insulin prescriptions 0 0.00 11619 20.79
Oral antidiabetic drug prescriptions 0 0.00 37163 66.49
Open in a new tab

BMI: Body Mass Index; IHD: Ischemic heart disease; CVA: Cerebrovascular accident; SD: Standard Deviation.

A total of 17,774/55,891 (31.8%) T2DM and 23,794/103,093 (23.1%) non-T2DM patients died during follow-up, equivalent to mortality (95% CI) rate of 48.8 (48.1, 49.5) per 1,000 person years and 33.8 (33.3, 34.2) per 1,000 person years respectively (Figure 2). In an age-sex-practice matched Cox model a higher mortality hazard in T2DM respect non-T2DM was found: HR 1.51 (95% CI 1.48 to 1.55), which remained significant in a multivariable-adjusted model (adjusted HR 1.46; 95%CI 1.42 to 1.49) (Table 2).

Figure 2.

Figure 2

Open in a new tab

Kaplan-Meier mortality risk function in T2DM and matched non-T2DM patients

Table 2.

Hazard and Subhazard ratio of risk fracture between groups.

Raw* CI 95% Adjusted CI 95%
HRdeath 1.51 1.48 to 1.55 1.46** 1.42 to 1.49
HRhip fracture 1.32 1.24 to 1.40 1.31*** 1.23 to 1.40
SHR 1.17 1.11 to 1.23 1.15*** 1.09 to 1.21
Open in a new tab

HRdeath: Death Cause-Specific Hazard Ratio ; HRhip fracture: Hip fracture Cause-Specific Hazard Ratio; SHR: SubHazard Ratio; CI95%; 95% confidence interval.

*

Match by age and sex

**

Adjusted by age, previous cerebrovascular accident, previous Ischemic heart disease, previous osteoarthritis, previous nephropathy, previous falls, corticoids prescriptions and anti-osteoporosis prescriptions.

***

Adjusted by previous major osteoporotic fracture, previous cerebrovascular accident, previous Ischemic heart disease, previous osteoarthritis, previous nephropathy, previous falls, corticoids prescriptions and anti-osteoporosis prescriptions.

Similarly, 2,154/55,891 (3.9%) T2DM and 3,388/103,093 (3.3%) non-T2DM patients sustained a hip fracture in the study period, with estimated incidence rates of 6.0 (5.76 to 6.27) per 1,000 person-years and 4.9 (4.71 to 5.04) per 1,000 person-years respectively. Cox models showed a significant association between T2DM and hip risk fracture in a matched model (HR 1.32; 95%CI 1.24 to 1.40), which remained significant in a multivariable-adjusted model (adjusted HR 1.31; 95%CI 1.23 to 1.40). Accounting for death as a competing event, the association between T2DM and hip hazard fracture remained statistically significant (matched –unadjusted- sHR 1.17; 95% CI 1.11 to 1.23). Adjustment for significant confounders did not modify substantially these estimates (adjusted sHR 1.15; 95%CI 1.09 to 1.21) (Table 2). In Figure 3, Kaplan-Meier functions of hip fracture probability according to T2DM status (failing to account for differential mortality) are compared to Cumulative Incidence Functions, which do account for the effect of death. The 5-year cumulative incidence of hip fracture using Kaplan–Meier was 2.6% in T2DM patients and 2.1% in non-T2DM patients. Using competing risk methodology, 5-year cumulative incidence of hip fracture differed, with an incidence of 2.3% among T2DM, slightly worse than the estimate of 1.9% for non-T2DM.

Figure 3.

Figure 3

Open in a new tab

Estimates probabilities of hip fracture in non-T2DM and T2DM using Kaplan-Meyer and Cumulative Incidence Function.

Discussion

T2DM patients have a 50% higher mortality than matched non-T2DM. Using survival analysis methods such as Cox regression, T2DM patients have a 35% higher risk of hip fracture; we have shown, however, that models accounting for the differential mortality as a competing event reduce this estimate of excess risk of hip fracture amongst T2DM to 15% and show a 21% increased incidence of hip fracture at 5 years in T2DM participants when compared to matched non-T2DM. Failure to account for competing risk with death overestimates the effect size of the association between T2DM and hip fracture risk.

T2DM is a well-known risk factor for numerous health problems. Individuals with high blood glucose levels during a long period are at higher risk to develop cardiovascular disease, blindness, kidney failure, and lower limb amputation. Therefore, T2DM is a key predictor, such as smoking or hypertension, in cardiovascular risk functions27,28. This does therefore explain the known increased mortality associated with this long-term condition7. In addition, previous studies have also shown that T2DM patients have higher hip fracture risk with a 38% increase compared to non-T2DM11.

The existence of these two parallel associations is a crucial methodological conundrum, as death precludes the occurrence of a hip fracture. Therefore, the competing risks with death should be taken into account when analysing the association between T2DM and fractures. However, most of the existing literature fails by using classic methods such as Cox PH with data death censored. For illustration, in the eight studies used in the metanalysis by Vestergaard11 all use Cox PH taking with death censored to estimate the effects of diabetes on the hazard of hip fracture. In a competing event scenario, there is no correspondence between cause-specific hazard and cumulative incidence thus failure to account for death as a competing risk results in an unrealistic estimate of this association26,29. According to our results, failing to account for differential mortality is an overestimation in the excess risk of hip fracture associated with T2DM.

Other studies on T2DM or in fracture risk have described this same issue in other settings. A study focused on assessing the impact of diabetes on relapse-free period and overall mortality in elderly breast cancer patients30. They found that when taking competing mortality into account, relapse-free period was better in elderly breast cancer patients with diabetes compared with patients without diabetes. Another study assessed whether assisted peritoneal dialysis was associated with a lower risk for technique failure31. In contrast with the results of a traditional cox model, they found using Fine and Gray that assisted peritoneal dialysis was associated with a lower risk for transfer to hemodialysis. A study to determine the association of weight loss with risk of clinical fractures at the hip, spine, and pelvis in older men, found that ignoring the competing mortality risk among men with weight loss substantially overestimates their long-term fracture probability and relative fracture risk32. Finally another study focused on analyse the peritonitis-free survival in the peritoneal dialysis. They found that using a competing risk approach the cumulative incidence of at least one peritonitis episode was lower than reported by the Kaplan–Meier method33.

However, there are some limitations in our study. First, our data is based on clinical records and we lack validation of each individual fracture. However, coding of fractures in SIDIAP has been compared to classical cohort data and hospital databases and shown to be highly specific (>95% for all fracture sites tested) and moderately sensitive (almost 70% for hip fractures)19. Also, ICD-10 does not distinguish between traumatic fractures and fragility fractures. A recent study including a random sample of 300 SIDIAP participants aged > 50 years old who suffered a fracture during 2012 has showed that >90% of hip fractures and >80% of major fractures were fragility (not related to high impact trauma)23. Residual confounding is possible in our study, as some key potential risk factors for fracture such as parental hip fracture history were not available in the dataset. Finally, we are uncertain whether these findings are generalizable to other populations, but previous research by Giangregorio et al34 have found a similar association in Manitoba cohort, Canada.

In summary, we conclude that estimation of hip fracture risk without accounting for death as a competing event results in an overestimation of fracture risk. All fracture risk estimation/s on T2DM patients should therefore be addressed accounting for death as a competing risk. Our data do however confirm that T2DM patients are at higher risk of hip fracture even after accounting for death as a competing risk. Conventional fracture risk assessment models like FRAX do not include T2DM as a risk factor so a limited predictive accuracy in T2DM patients is expected. Studies on the anti-fracture effectiveness of anti-osteoporosis medications (or non-pharmacological treatments) should account for a competing risk with death. This is particularly relevant if such interventions can affect mortality, similar to the effects observed in our T2DM vs non-T2DM example.

Acknowledgements

Cristian Tebe, Daniel Martinez-Laguna, and Daniel Prieto-Alhambra designed the study, performed data management, statistical analysis, and drafted the paper. Cyrus Cooper, Victor Moreno, Adolfo Diez-Perez and Gary S. Collins supervised epidemiological/analytical aspects, participated in the discussion and interpretation of results, and reviewed and critically edited the paper. All authors read the final version and approved submission.

This work was supported by the NIHR Biomedical Research Centre, Oxford. DPA is funded by a National Institute for Health Research Clinician Scientist award (CS-2013-13-012). This article presents independent research funded by the National Institute for Health Research (NIHR). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.

References

  • 1.Ogurtsova K, da Rocha Fernandes JD, Huang Y, et al. IDF Diabetes Atlas: Global estimates for the prevalence of diabetes for 2015 and 2040. Diabetes Res Clin Pract. 2017;128:40–50. doi: 10.1016/j.diabres.2017.03.024. [DOI] [PubMed] [Google Scholar]
  • 2.Fuller JH, Elford J, Goldblatt P, Adelstein AM. Diabetes mortality: new light on an underestimated public health problem. [Accessed February 19, 2018];Diabetologia. 1983 24(5):336–341. doi: 10.1007/BF00251820. [DOI] [PubMed] [Google Scholar]
  • 3.Sandoz MS, Ess SM, Keusch GW, Schwenkglenks M, Szucs TD. Prevalence and direct medical costs of end-stage renal disease in patients with type 2 diabetes mellitus in Switzerland for 2001. Swiss Med Wkly. 2004;134(31–32):448–458. doi: 10.4414/smw.2004.10682. doi:2004/31/smw-10682. [DOI] [PubMed] [Google Scholar]
  • 4.Narres M, Claessen H, Droste S, et al. The Incidence of End-Stage Renal Disease in the Diabetic (Compared to the Non-Diabetic) Population: A Systematic Review. Jandeleit-Dahm K, editor. PLoS One. 2016;11(1):e0147329. doi: 10.1371/journal.pone.0147329. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Holman N, Hillson R, Young RJ. Excess mortality during hospital stays among patients with recorded diabetes compared with those without diabetes. Diabet Med. 2013;30(12):1393–1402. doi: 10.1111/dme.12282. [DOI] [PubMed] [Google Scholar]
  • 6.Lutgers HL, Gerrits EG, Sluiter WJ, et al. Life Expectancy in a Large Cohort of Type 2 Diabetes Patients Treated in Primary Care (ZODIAC-10) Sorensen TIA, editor. PLoS One. 2009;4(8):e6817. doi: 10.1371/journal.pone.0006817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Nwaneri C, Cooper H, Bowen-Jones D. Mortality in type 2 diabetes mellitus: magnitude of the evidence from a systematic review and meta-analysis. Br J Diabetes Vasc Dis. 2013;13(4):192–207. doi: 10.1177/1474651413495703. [DOI] [Google Scholar]
  • 8.Leung M-YM, Pollack LM, Colditz GA, Chang S-H. Life Years Lost and Lifetime Health Care Expenditures Associated With Diabetes in the U.S., National Health Interview Survey, 1997–2000. Diabetes Care. 2015;38(3):460–468. doi: 10.2337/dc14-1453. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.British Orthopaedic Association. THE CARE OF PATIENTS WITH FRAGILITY FRACTURE. [Accessed February 19, 2018];2007 www.boa.ac.uk.
  • 10.Tuominen JT, Impivaara O, Puukka P, Rönnemaa T. Bone mineral density in patients with type 1 and type 2 diabetes. [Accessed February 19, 2018];Diabetes Care. 1999 22(7):1196–1200. doi: 10.2337/diacare.22.7.1196. [DOI] [PubMed] [Google Scholar]
  • 11.Vestergaard P. Discrepancies in bone mineral density and fracture risk in patients with type 1 and type 2 diabetes--a meta-analysis. Osteoporos Int. 2007;18(4):427–444. doi: 10.1007/s00198-006-0253-4. [DOI] [PubMed] [Google Scholar]
  • 12.Hothersall EJ, Livingstone SJ, Looker HC, et al. Contemporary Risk of Hip Fracture in Type 1 and Type 2 Diabetes: A National Registry Study From Scotland. J Bone Miner Res. 2014;29(5):1054–1060. doi: 10.1002/jbmr.2118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Schwartz AV, Sellmeyer DE, Ensrud KE, et al. Older Women with Diabetes Have an Increased Risk of Fracture: A Prospective Study. J Clin Endocrinol Metab. 2001;86(1):32–38. doi: 10.1210/jcem.86.1.7139. [DOI] [PubMed] [Google Scholar]
  • 14.Bonds DE, Larson JC, Schwartz AV, et al. Risk of Fracture in Women with Type 2 Diabetes: the Women’s Health Initiative Observational Study. J Clin Endocrinol Metab. 2006;91(9):3404–3410. doi: 10.1210/jc.2006-0614. [DOI] [PubMed] [Google Scholar]
  • 15.Melton LJ, Leibson CL, Achenbach SJ, Therneau TM, Khosla S. Fracture risk in type 2 diabetes: update of a population-based study. J Bone Miner Res. 2008;23(8):1334–1342. doi: 10.1359/jbmr.080323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Martinez-Laguna D, Tebe C, Javaid MK, et al. Incident type 2 diabetes and hip fracture risk: a population-based matched cohort study. Osteoporos Int. 2015;26(2):827–833. doi: 10.1007/s00198-014-2986-9. [DOI] [PubMed] [Google Scholar]
  • 17.Roman de Mettelinge T, Cambier D, Calders P, Van Den Noortgate N, Delbaere K. Understanding the relationship between type 2 diabetes mellitus and falls in older adults: a prospective cohort study. Bayer A, editor. PLoS One. 2013;8(6):e67055. doi: 10.1371/journal.pone.0067055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Harvey N, Dennison E, Cooper C. Osteoporosis: impact on health and economics. Nat Rev Rheumatol. 2010;6(2):99–105. doi: 10.1038/nrrheum.2009.260. [DOI] [PubMed] [Google Scholar]
  • 19.Pagès-Castellà A, Carbonell-Abella C, Avilés FF, et al. Burden of osteoporotic fractures in primary health care in Catalonia (Spain): a population-based study. BMC Musculoskelet Disord. 2012;13(1):79. doi: 10.1186/1471-2474-13-79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Hailer NP, Garland A, Rogmark C, Garellick G, Kärrholm J. Early mortality and morbidity after total hip arthroplasty in patients with femoral neck fracture. Acta Orthop. 2016;87(6):560–566. doi: 10.1080/17453674.2016.1234869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Lau B, Cole SR, Gange SJ. Competing Risk Regression Models for Epidemiologic Data. Am J Epidemiol. 2009;170(2):244–256. doi: 10.1093/aje/kwp107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Bolíbar B, Fina Avilés F, Morros R, et al. [SIDIAP database: electronic clinical records in primary care as a source of information for epidemiologic research] Med Clin (Barc) 2012;138(14):617–621. doi: 10.1016/j.medcli.2012.01.020. [DOI] [PubMed] [Google Scholar]
  • 23.Martinez-Laguna D, Soria-Castro A, Carbonell-Abella C, et al. Validation of fragility fractures in primary care electronic medical records: A population-based study. Reumatol Clin. 2017 Nov; doi: 10.1016/j.reuma.2017.10.013. [DOI] [PubMed] [Google Scholar]
  • 24.Kleinbaum DG, Klein M. Survival Analysis : A Self-Learning Text. Springer; 2012. [Google Scholar]
  • 25.Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509. doi: 10.1080/01621459.1999.10474144. [DOI] [Google Scholar]
  • 26.Austin PC, Lee DS, Fine JP. Introduction to the Analysis of Survival Data in the Presence of Competing Risks. Circulation. 2016;133(6):601–609. doi: 10.1161/CIRCULATIONAHA.115.017719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.D’Agostino RB, Grundy S, Sullivan LM, Wilson P, CHD Risk Prediction Group Validation of the Framingham coronary heart disease prediction scores: results of a multiple ethnic groups investigation. [Accessed February 19, 2018];JAMA. 2001 286(2):180–187. doi: 10.1001/jama.286.2.180. [DOI] [PubMed] [Google Scholar]
  • 28.Marrugat J, Subirana I, Ramos R, et al. Derivation and validation of a set of 10-year cardiovascular risk predictive functions in Spain: The FRESCO Study. Prev Med (Baltim) 2014;61:66–74. doi: 10.1016/j.ypmed.2013.12.031. [DOI] [PubMed] [Google Scholar]
  • 29.Andersen PK, Geskus RB, de Witte T, Putter H. Competing risks in epidemiology: possibilities and pitfalls. Int J Epidemiol. 2012;41(3):861–870. doi: 10.1093/ije/dyr213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Kiderlen M, de Glas NA, Bastiaannet E, et al. Diabetes in relation to breast cancer relapse and all-cause mortality in elderly breast cancer patients: a FOCUS study analysis. Ann Oncol Off J Eur Soc Med Oncol. 2013;24(12):3011–3016. doi: 10.1093/annonc/mdt367. [DOI] [PubMed] [Google Scholar]
  • 31.Lobbedez T, Verger C, Ryckelynck J-P, Fabre E, Evans D. Is assisted peritoneal dialysis associated with technique survival when competing events are considered? Clin J Am Soc Nephrol. 2012;7(4):612–618. doi: 10.2215/CJN.10161011. [DOI] [PubMed] [Google Scholar]
  • 32.Ensrud KE, Harrison SL, Cauley JA, et al. Impact of Competing Risk of Mortality on Association of Weight Loss With Risk of Central Body Fractures in Older Men: A Prospective Cohort Study. J Bone Miner Res. 2017;32(3):624–632. doi: 10.1002/jbmr.3020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Evans DW, Ryckelynck J-P, Fabre E, Verger C. Peritonitis-free survival in peritoneal dialysis: an update taking competing risks into account. Nephrol Dial Transplant. 2010;25(7):2315–2322. doi: 10.1093/ndt/gfq003. [DOI] [PubMed] [Google Scholar]
  • 34.Giangregorio LM, Leslie WD, Lix LM, et al. FRAX underestimates fracture risk in patients with diabetes. J Bone Miner Res. 2012;27(2):301–308. doi: 10.1002/jbmr.556. [DOI] [PubMed] [Google Scholar]

RESOURCES