Abstract
Summary
We used a large population-based health care database to determine the impact of common co-morbidities on hip fracture risk amongst elderly men. We demonstrated that diabetes, chronic obstructive pulmonary disease, renal failure, HIV infection, dementia, and cerebrovascular disease are independent predictors of hip fracture, as is a Charlson score of ≥3.
Introduction
Risk factors for hip fractures in men are still unclear. We aimed to identify common co-morbidities (amongst those in the Charlson index) that confer an increased risk of hip fracture amongst elderly men.
Methods
We conducted a population-based cohort study using data from the SIDIAP Q database. SIDIAPQ contains primary care and hospital inpatient records of a representative 30 % of the population of Catalonia, Spain (>2 million people). All men aged ≥65 years registered on 1 January 2007 were followed up until 31 December 2009. Both exposure (co-morbidities in the Charlson index) and outcome (incident hip fractures) were ascertained using ICD codes. Poisson regression models were fitted to estimate the effect of (1) each individual co-morbidity and (2) the composite Charlson index score, on hip fracture risk, after adjustment for age, body mass index, smoking, alcohol drinking, and use of oral glucocorticoids.
Results
We observed 186,171 men for a median (inter-quartile range) of 2.99 (2.37–2.99) years. In this time, 1,718 (0.92 %) participants had a hip fracture. The following co-morbidities were independently associated with hip fractures: diabetes mellitus, chronic obstructive pulmonary disease (COPD), renal failure, HIV infection, dementia, and cerebrovascular disease. A Charlson score of ≥3 conferred an increased hip fracture risk.
Conclusion
Common co-morbidities including diabetes, COPD, cerebrovascular disease, renal failure, and HIV infection are independently associated with an increased risk of hip fracture in elderly men. A Charlson score of 3 or more is associated with a 50 % higher risk of hip fracture in this population.
We’re sorry, something doesn't seem to be working properly.
Please try refreshing the page. If that doesn't work, please contact support so we can address the problem.
Introduction
Hip fracture rates are half in men compared to women [1], with a high heterogeneity worldwide: recently reported incidence rates range from 12.3 per 10,000 in Catalonia, Spain [2], to 41.4 per 10,000 in the USA [3]. There is also an important gender difference in the management of hip fracture that needs to be taken into consideration: men are under-diagnosed and under-treated [4] in spite of having an excess fracture-related mortality compared to women [5].
Risk factors for hip fracture have been exhaustively studied in women, but there are fewer studies in men. Some of the most relevant risk factors reported for hip fracture in men have been body mass index [6], age [7], and smoking [8]. In 2012, a meta-analysis added to these associations is the following predictors: alcohol consumption, diabetes, stroke, and dementia [9]. However, there is still a scarcity of data on risk factors for hip fracture in men.
The Charlson co-morbidity index (CCI) was designed on the basis of information obtained from a cohort of hospitalized patients [10]. It is a well-validated tool, widely used to predict inpatient mortality [11], as well as functional outcomes [12]. We hypothesized that common co-morbidities in the CCI, as well as the index itself contribute to elderly male patients’ risk of hip fracture.
For this reason, we carried out a study using routinely collected data in SIDIAPQ [13] in order to determine which of the different co-morbidities registered in the CCI could be identified as risk factors for hip fracture in men. Second, we looked at the association between CCI score and hip fracture risk amongst elderly men.
Methods
Source of data
We carried out a population-based cohort study using data from the SIDIAP Q (Sistema d‘Informació per al Desenvolupament de l‘Investigació en Atenció Primària) database. SIDIAPQ contains the primary care computerized medical records of >1,300 GPs in Catalonia (northeast Spain), with information on a representative 30 % of the population (>2 million people). It comprises the clinical and referral events registered by primary care health professionals (GPs and nurses) in electronic medical records, socio-demographic information, pharmacy invoicing data, and referrals to imaging and secondary care. Only GPs who achieve quality control standards can contribute to the SIDIAPQ database [13].
Health professionals gather this information using International Classification of Diseases (ICD-10) codes and structured forms designed for the standardized collection of variables relevant for primary health care, including lifestyle risk factors (smoking and alcohol drinking) and anthropometric measurements (height, weight, and body mass index), amongst others.
SIDIAPQ is further linked to the official hospital inpatient records database (CMBD-AH for its acronym in Catalan language) to improve data completeness.
Study participants
We included all men aged ≥65 years old [14] permanently registered in the SIDIAPQ database in 2007.
Exposure: co-morbid conditions and CCI
Common co-morbid conditions in the CCI index were identified, up to 1 January 2007, using a validated algorithm for the calculation of CCI scores in administrative datasets [15]. Coding of common co-morbid conditions in SIDIAPQ has been shown to be comparable to that gathered in national health surveys for Catalonia, particularly in the elderly population [16, 17].
Outcome: hip fractures
Incident fractures in the study period (1 January 2007 to 31 December 2009) were ascertained within both primary care and hospital episodes data using ICD-10 codes. Coding of hip fractures has been validated in SIDIAPQ compared to prospective cohort data as a reference and shown to be highly specific (99 %) but less sensitive (68 %) [2]. Therefore, hospital inpatient data were also used to identify hip fractures to minimize misclassification.
Statistical analyses
Hip fracture incidence rates and 95 % confidence intervals (95 % CI) were calculated assuming a Poisson distribution. We used manual backwards stepwise methods to identify key co-morbidities within the CCI that appeared significantly associated with hip fractures. All variables were then adjusted for age, body mass index, smoking, alcohol intake, and use of oral glucocorticoids. Similarly, adjusted Poisson models were used to calculate the relative risk (RR, 95 % CI) according to CCI score compared to CCI = 0, which was defined as the reference group. Kaplan-Meier estimates of age-adjusted cumulative hip fracture risk stratified by key independent predictors as well as by CCI scores were plotted for visual purposes. Study participants were followed from enrolment (01 January 2007) to study end (31 December 2009), death or date of transfer out of the catchment area whichever came first. All these analyses were carried out using Stata/SE for Mac version 12.0.
Ethics
The study was conducted in accordance with Good Clinical Practice and the Declaration of Helsinki and was approved by the ethical committee of reference (CEIC IDIAP Jordi Gol i Gurina).
Results
We included 186,171 men, observed for a median (inter-quartile range) of 2.99 (2.37–2.99) years. In this time, 1,718 (0.92 %) participants sustained a hip fracture. Men with a hip fracture were significantly older, thinner, more likely to be oral glucocorticoid users, and had a higher number of long-term co-morbidities. Details on baseline characteristics of men with and without an incident hip fracture in the study period are shown in Table 1.
Hip fracture incidence rates increased with age particularly from the age of 80 (54 per 10,000 persons/year), doubling its incidence every 5 years and reaching values of 184 hip fractures per 10,000 persons/year in men over 90 years old.
A number of co-morbidities amongst those in the CCI appeared independently associated with an increased risk of hip fracture in elderly men, even after adjusting for potential confounders. Diabetes mellitus, with and without complications, appeared associated with an increased risk: adjusted RR of 1.45 (95 % CI, 1.25 to 1.69) and 1.89 (95 % CI, 1.15 to 3.12), respectively. Chronic obstructive pulmonary disease (COPD) was also related to an increased hip fracture risk: adjusted RR of 1.20 (95 % CI, 1.03 to 1.40). Other conditions associated with an excess risk of hip fractures were the following: chronic kidney disease, adjusted RR of 1.32 (95 % CI, 1.07–1.65); HIV infection, adjusted RR of 5.03 (95 % CI, 1.25–20.21); dementia, adjusted RR of 1.65 (95 % CI, 1.30–2.09); cerebrovascular disease, adjusted RR of 1.51 (95 % CI of 1.27–1.80); and mild liver disease (MLD), adjusted RR of 1.53 (95 % CI, 1.10–2.13). Detailed results for these analyses are reported in Table 2, and age-adjusted Kaplan-Meier estimates of cumulative risk of hip fracture for participants with and without each these co-morbidities are shown in Figs. 1, 2 and 3.
The CCI was calculated for the study population, and all men were classified depending on the score obtained; Table 3 reflects the number of subjects with each CCI score. In order to determine the excess risk of hip fracture generated by the co-morbidities registered in CCI score, score results above 0 (1, 2, 3, 4 and >4) were compared to subject with no co-morbidities (CCI score of 0). When comparing to patients without co-morbidities, men with a CCI score of 3, 4 and >4 had an increased risk of hip fracture (Table 3).
Discussion
Association between the CCI and hip fracture in men
In this large cohort study, a number of long-term co-morbidities (diabetes, COPD, renal failure, HIV infection, dementia, and cerebrovascular disease) were found to be independent predictors of hip fracture in elderly men with different effect sizes. In addition, we have shown that a score of 3 or above in the widely used CCI index is associated with a 50 % increase in hip fracture risk in this same population, even after adjustment for potential confounders (age, body mass index, smoking, alcohol intake, use or corticoids and other co-morbid conditions).
Other co-morbidity scores, such as the Elixhauser score [18], have been used recently to predict different health outcomes. As the CCI index, the Elixhauser score was aimed to predict mortality in an inpatient population [18], extending its scope to include other outpatient outcomes [19]. Both scores have been validated for the ICD-10 codes and have shown similar performance predicting mortality outcomes [20]. Nevertheless, the CCI showed a greater adaptability to the SIDIAP database proving to be representative of the common diagnosis in primary care.
To our knowledge, this is the first study that uses the CCI score in order to identify potential risk factors of hip fracture in men and to grade this increased risk depending on the score obtained. In 2012, a Swedish study [19] based on 2,841 men found that the use of a co-morbidity score, such as the Elixhauser score [18], added to the WHO Fracture Risk Assessment Tool (FRAX) algorithm improved the fracture prediction, reaching a prognostic value of hip fracture of up to 80 % in this population. One of the main limitations of FRAX is the lack of representation of co-morbidities, which are able to increase the risk of hip fracture. This was remarked by Byberg et al. [19] and confirmed in our study, where men with high score of CCI had a higher risk of hip fracture compared with those with a lower score.
Previous studies have associated higher CCI scores to an increased risk of death after hip fracture [5, 11]. In 2010, a study published in Denmark [5] found an increased risk of mortality in men with a hip fracture who had a previous diagnosis of COPD, dementia, renal failure and/or mild liver disease. Our results extend these previous findings, showing that these same conditions are, in fact, risk factors for the hip fracture in men.
Concerning the severe liver disease, we found no association with hip fracture in men. Despite that this association would be expected, since a milder liver disease has proven to increase the risk [5], we accounted for a low number of subjects identified with this diagnosis, which could have underpowered our study and explain these results.
Association between risk factors included in the CCI and hip fracture in men
In accordance with previous studies, some of the risk factors identified were increasing age [9, 21], corticosteroid use, diabetes (DM) [9] and cerebrovascular disease [9, 21, 22]. The association between type 2 DM and hip fracture had given, in the past, uneven results, probably due to the complexity of this illness [23]: More recent publications that aimed to establish this relationship found an increased risk of hip fracture in patients with DM and specifically in those with a bad glycemic control [24]. We found that elderly men with a diagnosis of DM had a 45 % increased risk of hip fracture, and this risk was higher if there were any complications added.
Regarding cerebrovascular disease and dementia, our results support the current literature [22, 25], which points towards an increased fracture risk in affected patients. Mechanisms proposed for the appearance of hip fracture in this population involve a decrease in vitamin D levels [25], increasing falls, and a reduction in bone mineral density in hemiparetic/hemiplegic limbs [26].
In the osteoporotic fractures in men (MrOs) study [27], carried out in the USA, a 12 % increased risk of hip fracture was found in patients with COPD or asthma, although this association did not reach statistical significance probably due to the low number of men with this condition amongst the study participants. Conversely, we accounted for 32,000 patients with COPD, and our results were statistically significant even after adjustment for the use of oral corticosteroids. In our data, men with mild liver disease had a 20 % excess risk of hip fracture. Similar results were found in a cohort study published in 2012, detecting an increased risk of hip fracture in men with hepatitis C virus (HCV) [28] compared to uninfected peers. Furthermore, we also identified HIV infection as a risk factor for hip fracture, and this is consistent with data reported elsewhere [29, 30], although further studies should account for anti-retroviral medication, since it has been found to increase the risk of hip fracture by itself, mostly at the beginning of the medication use [30].
Strengths and limitations
The main strengths of this study are the large sample size available for each of the co-morbidities studied and the high representativeness of the data, which were routinely collected in actual practice conditions.
The main limitation of this study is the lack of individual validation of each hip fracture as well as of each risk factor in addition to a lack of detail on disease severity for each of the co-morbidities. However, hip fracture coding has been shown to be highly valid in the SIDIAP database [2], and the information gathered in the database on co-morbid conditions is comparable to that collected in local health surveys, especially for the elderly populations [15]. Moreover, despite the fact that only GPs that meet quality standards are accepted to contribute to the SIDIAP database, we cannot discard potential misclassification of the co-morbidities.
Other limitations are the short follow-up of the cohort and the lack of information on bone mineral density, which could explain part of the observed associations. However, our aim was not to describe the mechanisms by which each of the co-morbidities studied confer a higher risk of hip fractures but to quantify the excess risk related to each one of them after multivariate adjustment.
There are some potential residual confoundings that should be mentioned and that could have influenced our results such as other co-morbidities that could act as risk factors of hip fracture and that were not included in our study, such as repeated falls or hypogonadism [31].
In conclusion, common co-morbidities including diabetes, COPD, cerebrovascular disease, renal failure, and HIV infection are independent risk factors of hip fracture in elderly men. Similarly, men with a high CCI score are at increased risk of hip fractures compared to men without co-morbidities. These data add to the scarce but existing literature on the predictors of hip fracture in the elderly male population.
References
Kanis JA, Odén A, McCloskey EV, Johansson H, Wahl DA, Cooper C, IOF Working Group on Epidemiology and Quality of Life (2012) A systematic review of hip fracture incidence and probability of fracture worldwide. Osteoporos Int 23:2239–2256
Pagès-Castellà A, Carbonell-Abella C, Avilés FF, Alzamora M, Baena-Díez JM, Laguna DM, Nogués X, Díez-Pérez A, Prieto-Alhambra D (2012) Burden of osteoporotic fractures in primary health care in Catalonia (Spain): a population-based study. BMC Musculoskelet Disord. doi:10.1186/1471-2474-13-79
Brauer CA, Coca-Perraillon M, Cutler DM, Rosen AB (2009) Incidence and mortality of hip fractures in the United States. JAMA 302:1573–1579
Cawthon PM (2011) Gender differences in osteoporosis and fractures. Clin Orthop Relat Res 469:1900–1905
Kannegaard PN, van der Mark S, Eiken P, Abrahamsen B (2010) Excess mortality in men compared with women following a hip fracture. National analysis of comedications, comorbidity and survival. Age Ageing 39:203–209
Premaor MO, Compston JE, Fina Avilés F, Pagès-Castellà A, Nogués X, Díez-Pérez A, Prieto-Alhambra D (2013) The association between fracture site and obesity in men: A population-based cohort study. J Bone Miner Res. doi:10.1002/jbmr.1878
Nguyen TV, Eisman JA, Kelly PJ, Sambrook PN (1996) Risk factors for osteoporotic fractures in elderly men. Am J Epidemiol 144:255–263
Hannan MT, Felson DT, Dawson-Hughes B, Tucker KL, Cupples LA, Wilson PW, Kiel DP (2000) Risk factors for longitudinal bone loss in elderly men and women: the Framingham Osteoporosis Study. J Bone Miner Res 15:710–720
Drake MT, Murad MH, Mauck KF, Lane MA, Undavalli C, Elraiyah T, Stuart LM, Prasad C, Shahrour A, Mullan RJ, Hazem A, Erwin PJ, Montori VM (2012) Clinical review. Risk factors for low bone mass-related fractures in men: a systematic review and meta-analysis. J Clin Endocrinol Metab 97:1861–1870
Charlson ME, Pompei P, Ales KL, MacKenzie CR (1987) A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 40:373–383
Souza RC, Pinheiro RS, Coeli CM, Camargo KR (2008) The Charlson comorbidity index (CCI) for adjustment of hip fracture mortality in the elderly: analysis of the importance of recording secondary diagnoses. Cad Saúde Pública 24:315–322, Rio de Janeiro
Soler PA, Paterna Mellinas G, Martınez Sanchez E, Lopez Jimenez E (2010) Evaluación de la comorbilidad en la poblacion anciana: utilidad y validez de los instrumentos de medida. Rev Esp Geriatr Gerontol 45:219–228
García-Gil Mdel M, Hermosilla E, Prieto-Alhambra D, Fina F, Rosell M, Ramos R, Rodriguez J, Williams T, Van Staa T, Bolíbar B (2011) Construction and validation of a scoring system for the selection of high-quality data in a Spanish population primary care database (SIDIAP). Inform Prim Care 19:135–145
Andrews G, Faulkner D, Andrews M (2004) A Glossary of Terms for Community Health Care and Services for Older Persons. WHO. http://www.who.int/kobe_centre/ageing/ahp_vol5_glossary.pdf
Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, Saunders LD, Beck CA, Feasby TE, Ghali WA (2005) Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care 43:1130–1139
Violán C, Foguet-Boreu Q, Hermosilla-Pérez E, Valderas JM, Bolíbar B, Fàbregas-Escurriola M, Brugulat-Guiteras P, Muñoz-Pérez MA (2013) Comparison of the information provided by electronic health records data and a population health survey to estimate prevalence of selected health conditions and multimorbidity. BMC Public Health 13:251
Ramos R, Balló E, Marrugat J, Elosua R, Sala J, Grau M, Vila J, Bolíbar B, García-Gil M, Martí R, Fina F, Hermosilla E, Rosell M, Muñoz MA, Prieto-Alhambra D, Quesada M (2012) Validity for use in research on vascular diseases of the SIDIAP (Information System for the Development of Research in Primary Care): the EMMA study. Rev Esp Cardiol (Engl Ed) 65:29–37
Elixhauser A, Steiner C, Harris DR, Coffey RM (1998) Comorbidity measures for use with administrative data. Med Care 36:8–27
Byberg L, Gedeborg R, Cars T, Sundström J, Berglund L, Kilander L, Melhus H, Michaëlsson K (2012) Prediction of fracture risk in men: a cohort study. J Bone Miner Res 27:797–807
Li B, Evans D, Faris P, Dean S, Quan H (2008) Risk adjustment performance of Charlson and Elixhauser comorbidities in ICD-9 and ICD-10 administrative databases. BMC Health Serv Res 8:12
Lewis CE, Ewing SK, Taylor BC, Shikany JM, Fink HA, Ensrud KE, Barrett-Connor E, Cummings SR, Orwoll E (2007) Osteoporotic Fractures in Men (MrOS) Study Research Group. Predictors of non-spine fracture in elderly men: the MrOS study. J Bone Miner Res 22:211–219
Pouwels S, Lalmohamed A, Leufkens B, de Boer A, Cooper C, van Staa T, de Vries F (2009) Risk of hip/femur fracture after stroke: a population-based case-control study. Stroke 40:3281–3285
Raska I Jr, Broulík P (2005) The impact of diabetes mellitus on skeletal health: an established phenomenon with inestablished causes? Prague Med Rep 106:137–148
Oei L, Zillikens MC, Dehghan A, Buitendijk GH, Castaño-Betancourt MC, Estrada K et al (2013) High bone mineral density and fracture risk in type 2 diabetes as skeletal complications of inadequate glucose control: the rotterdam study. Diabetes Care 36:1619–1628
Zhao Y, Shen L, Ji HF (2012) Alzheimer's disease and risk of hip fracture: a meta-analysis study. ScientificWorldJournal 2012:872173. doi:10.1100/2012/872173
Demirbag D, Ozdemir F, Kokino S, Berkarda S (2005) The relationship between bone mineral density and immobilization duration in hemiplegic limbs. Ann Nucl Med 19:695–700
Dam TT, Harrison S, Fink HA, Ramsdell J, Barrett-Connor E (2010) Osteoporotic Fractures in Men (MrOS) Research Group. Bone mineral density and fractures in older men with chronic obstructive pulmonary disease or asthma. Osteoporos Int 21:1341–1349
Lo Re V 3rd, Volk J, Newcomb CW, Yang YX, Freeman CP, Hennessy S, Kostman JR, Tebas P, Leonard MB, Localio AR (2012) Risk of hip fracture associated with hepatitis C virus infection and hepatitis C/human immunodeficiency virus coinfection. Hepatology 56:1688–1698
Triant VA, Brown TT, Lee H, Grinspoon SG (2008) Fracture Prevalence among Human ImmunodeficiencyVirus (HIV)-Infected Versus Non-HIV-Infected Patients in a Large U.S. Healthcare System. J Clin Endocrinol Metab 93:3499–3504
Güerri-Fernandez R, Vestergaard P, Carbonell C, Knobel H, Avilés FF, Castro AS, Nogués X, Prieto-Alhambra D, Diez-Perez A (2013) HIV infection is strongly associated with hip fracture risk, independently of age, gender, and comorbidities: a population-based cohort study. J Bone Miner Res 28:1259–1263
Yin MT, Kendall MA, Wu X, Tassiopoulos K, Hochberg M, Huang JS, Glesby MJ, Bolivar H, McComsey GA (2012) Fractures after antiretroviral initiation. AIDS 26:2175–2184
Acknowledgments
We would like to acknowledge the health professionals (general practitioners and nurses) responsible for the collection of these data, as well as to the patients involved.
Conflicts of interest
CR: none; DPA: unrestricted research grants from BIOIBERICA SA and AMGEN; PO: none; PE: none; FF: none; XN: advisory board Amgen, educational speaker for Amgen, Lilly and MSD; CC: none; ADP: Advisor or speaker for Lilly, Amgen, Novartis, Pfizer and Active Life Scientific; JGM: none.
Author information
Authors and Affiliations
Corresponding author
Additional information
Conflict of interest
CR: none; DPA: unrestricted research grants from BIOIBERICA SA and AMGEN; PO: none; PE: none; FF: none; XN: advisory board Amgen, educational speaker for Amgen, Lilly and MSD; CC: none; ADP: Advisor or speaker for Lilly, Amgen, Novartis, Pfizer and Active Life Scientific; JGM: none.
About this article
Cite this article
Reyes, C., Estrada, P., Nogués, X. et al. The impact of common co-morbidities (as measured using the Charlson index) on hip fracture risk in elderly men: a population-based cohort study. Osteoporos Int 25, 1751–1758 (2014). https://doi.org/10.1007/s00198-014-2682-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00198-014-2682-9