Hip fracture in chronic kidney disease: incidence and mortality

Angharad Marks, Lynn Robertson, Gordon Prescott, Corri Black, Nick Fluck, Rosemary Hollick, Sharon Gordon

Research output: Contribution to conferenceAbstractpeer-review


Introduction: Chronic kidney disease (CKD) has many complications. For those on renal replacement therapy (RRT) there is an increase in the risk of fracture. The evidence regarding those with less advanced CKD is not so clear.

Aim: To understand the association between CKD and (i) hip-fracture related admissions; and (ii) mortality after hip-fracture.

Methods: The Grampian Laboratory Outcomes Mortality and Morbidity Study (GLOMMS-II) cohort was used, in particular, ~20,000 Grampian residents in 2003 who had evidence of CKD and ~20,000 with normal eGFR. Data-linkage to hospital episode data and National Records of Scotland allowed ascertainment of comorbidities, hip fracture and mortality events up until 2009. The incidence of hip fracture in those with CKD and those with normal eGFR in 2003 was calculated and the incidence rate ratios (IRR), unadjusted and adjusted for potential confounders. To estimate the risk of hip fracture attributable to CKD, the population attributable risk (PAR) and fraction (PAF) were calculated. To investigate whether mortality after a hip fracture differed between those with CKD and normal eGFR, the mortality rates in those who had suffered a hip fracture was calculated and the mortality rate ratio (MRR) calculated (unadjusted and adjusted).

Results: There were 19,537 persons with CKD in 2003 and a 19,748 sample with normal eGFR, the incidence of hip fracture was 10.01 and 1.45 per 1000 patient-years (py) respectively. Unadjusted IRRs for stage 3a, 3b, 4 and 5 were 5.61, 9.74, 10.96 and 2.27 [6.90 (CI 5.83, 8.15) overall for stage 3-5 CKD]. When adjusted for age and sex the overall IRR for stage 3-5 CKD was 1.56 (CI 1.31, 1.86). In the population of Grampian as a whole in 2003 49% of fractures (18% after adjusting for confounders) could be attributed to CKD. From the time of fracture the mortality rates in those with CKD and normal eGFR were 488.04 and 308.86 per 1000py respectively. The unadjusted MRRs for those with stage 3a, 3b, 4 and 5 CKD were 1.36, 1.75, 2.87 and 2.47 [1.58 (CI 1.26, 1.98) overall for stage 3-5 CKD]. When adjusted for age and sex, the overall MRR for stage 3-5 CKD was 1.11 (CI 0.88, 1.41).

Conclusion: Amongst those with CKD there is a higher rate of hip fracture than in those without, this represents a significant proportion of those who suffer hip fractures within the region, even adjusting for confounders. Although the mortality rates after a fracture are not significantly different between those with CKD and those with a normal eGFR, reducing the incidence of fractures would reduce the number of fracture related deaths in CKD. Prospective management of future fracture risk in those with CKD is warranted.

Source of funding: The analysis for this work was done with funding from NHS Grampian Endowments fund (14/30), however the set-up the cohort was funded by a CSO grant (CZH/4/656).
Conflict of Interests statement: None of the authors have any conflicting interests.
Original languageEnglish
Number of pages1
Publication statusUnpublished - Oct 2015
EventScottish Renal Association: 2015 - Dundee, United Kingdom
Duration: 1 Oct 20152 Oct 2015


ConferenceScottish Renal Association
Country/TerritoryUnited Kingdom


Dive into the research topics of 'Hip fracture in chronic kidney disease: incidence and mortality'. Together they form a unique fingerprint.

Cite this