- Email: [email protected]
Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients
G Model
AGG-2967; No. of Pages 6 Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Archives of Gerontology and Geriatrics journal homepage: www.elsevier.com/locate/archger
Review
Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients Yanbin Zhu a,b, Wei Chen a,b, Tao Sun a,b, Qi Zhang a,b, Jiaxiang Cheng a,b, Yingze Zhang a,b,* a b
Department of Orthopaedic Surgery, The Third Hospital of Hebei Medical University, Shijiazhuang, Hebei 050051, PR China Key Laboratory of Biomechanics of Hebei Province, Shijiazhuang, Hebei 050051, PR China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 2 December 2013 Received in revised form 24 February 2014 Accepted 27 February 2014 Available online xxx
This study aims to quantitatively summarize the risk factors for the incidence of SHF. A meta-analysis was performed with the data obtained from 22 relevant papers published in Pubmed, Embase and Cochrane central database (all through January 2014) following strict selection. The pooled odds ratios (ORs) or standardized mean difference (SMD) with 95% confidence intervals (CIs) were calculated for potential risk factors associated with SHF. Our meta-analysis indicated the significant risk factors for SHF were female (OR, 1.46; 95%CI, 1.29–1.66), living in institutions (OR, 2.23; 95%CI, 1.29–3.83), osteoporosis (Singh index (SI) 1–3) (OR, 10.02; 95%CI, 5.41–18.57), low vision (OR, 2.09; 95%CI, 1.06–4.12), dementia (OR, 1.89; 95%CI, 1.47–2.43), Parkinson (OR, 2.90; 95%CI, 1.41–5.95), cardiac diseases (OR, 1.32; 95%CI, 1.02–1.70) and respiratory disease (OR, 1.97; 95%CI, 1.16–3.32). Related strategies must be implemented on those involved with above-mentioned medical conditions to effectively prevent a SHF. ß 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Second hip fracture Elderly patients Risk factor Meta-analysis
Contents 1. 2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Literature search . . . . . . . . . . . . . . . . Data extraction . . . . . . . . . . . . . . . . . 2.2. Quality of included studies. . . . . . . . 2.3. Meta-analyses . . . . . . . . . . . . . . . . . . 2.4. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Literature search results . . . . . . . . . . 3.1. Methodological quality assessment . 3.2. Pooled analysis of risk factors . . . . . 3.3. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
1. Introduction Hip fracture is the most common fracture in elderly patients and accounts for 20% of all the fractures in this population (Zhang,
* Corresponding author at: Department of Orthopaedics, The Third Hospital, Hebei Medical University, No. 139 Ziqiang Road, Shijiazhuang 050051, PR China. Tel.: +86 311 88603682; fax: +86 311 87023626. E-mail addresses: [email protected], [email protected] (Y. Zhang).
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
000 000 000 000 000 000 000 000 000 000 000 000 000
2012). Consequently, patients involved with hip fractures suffered functional deterioration, limited mobility and increased morbidity, especially at the first year post operation (Fredman et al., 2005; Penrod et al., 2007; Vestergaard, Rejnmark, & Mosekilde, 2007). Furthermore, among the survivors of a first hip fracture, the agestandardized risk of a SHF was 2.5 times higher in women and 4.6 times higher in men compared to the risk of a first hip fracture (Omsland et al., 2012). And worse still, patients had a greater mortality and more decreased dependence in their daily activities after SHF (Pearse, Redfern, Sinha, & A, 2004; Sawalha & Parker,
http://dx.doi.org/10.1016/j.archger.2014.02.012 0167-4943/ß 2014 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
2
2012). About 31.6% (Sawalha & Parker, 2012) died at the first year after the operation, 78.3% would have to survive with the aid of walking tools, and most would never recover to their pre-operative level (Galois, Dausse, Mainard, & Delagoutte, 2001). Besides the personal influence to patients, a heavy economic and social burden should be also considered, a proximally financial cost of over 16,000 dollar per patient for unilateral hip fracture and at least doubled for bilateral hip fracture (Van Balen et al., 2001), still not including postoperative care cost after discharge. Therefore, it is vital to identify risk factors to prevent the occurrence of SHF in those sustaining initial hip fractures. Many risk factors have been identified for the SHF, including advanced age (Angthong, Suntharapa, & Harnroongroj, 2004; Souder et al., 2012; Vochteloo et al., 2012), female (Gaumetou, Zilber, & Hernigou, 2011; Rodaro, Pasqualini, Iona, & Di Benedetto, 2004), osteoporosis (Angthong et al., 2004; Yamanashi et al., 2004), dementia (Mitani, Shimizu, Abo, Hiroshi, & Youichi, 2010; Saxena & Shankar, 2000; Yamanashi et al., 2004), Parkinson’s disease (Saxena & Shankar, 2000; Yamanashi et al., 2004), respiratory disease (Mitani et al., 2010) and lower vision (Angthong et al., 2004; Saxena & Shankar, 2000). However, these studies had some limitations such as a small sample size and the inclusion of a single or very few potential risk factors. In addition, some of the results obtained were controversial. For example, female and advanced age were both considered risk factors for SHF in some studies (Rodaro et al., 2004), however, were neither significant in other study (Mitani et al., 2010). Mary et al. (Egan, Jaglal, Byrne, Jennie, & Paul, 2008) made a systematic review of studies on SHF in 2006 with a qualitative description rather than a quantitative metaanalysis on the potential risks. Recently, Jesper et al. (Ryg, Rejnmark, Overgaard, Kim, & Peter, 2009) have conducted a nationwide population-based cohort study of 169, 145 cases to investigate the risks for SHF. Although with a greatest number of cases by far, the study had some certain limitations such as the misclassification of infections, periprosthetic fracture, alloplastic surgery and removal of osteosynthesis as SHF. Therefore, it was difficult for them to get the precise information with regard to the risk factors for SHF. To achieve a reliable and comprehensive conclusion, a metaanalysis was performed using data obtained from previous studies to evaluate the risk factors for SHF in elderly patients with initial hip fractures, and subsequently reduce the occurrence of SHF.
2.2. Data extraction All the data were carefully extracted from all eligible studies independently by the two reviewers, Chen and Sun. The following variables were extracted from each study: first author’s name, publication year, country, significant risk factors, definitions and numbers of case and control groups, numbers of citations for each potential risk factors for SHF. Any disagreement was settled by discussion and a consensus was reached for all data. 2.3. Quality of included studies The quality of the included studies was evaluated using the Newcastle–Ottawa Scale (Stang, 2004): based on the three main items: the selection of the study groups (0–4 points), the comparability of the groups (0–2 points) and the determination of either the exposure or the outcome of interest (0–3 points), with a perfect score of 9. 2.4. Meta-analyses For each risk factor, we extracted the adjusted OR or relative risk (RR), with its 95%CI. When the adjusted OR or the RR was not obtained, we computed a crude OR and pooled across studies to assess the associations between different variables and the risk of SHFs with a P < 0.05 indicating significant difference. Heterogeneity between the studies was tested by Q-test statistics with significance set at P < 0.10 (Lau, Ioannidis, & Schmid, 1997). The I2 statistics were used as a second measure of heterogeneity, with I2 more than 50% indicating inconsistency. A random effects model was used to calculate pooled ORs or SMDs in the case of significant heterogeneity (P < 0.10 or I2 > 50%); otherwise, a fixed-effects model was used (Wei, Yang, Luo, Qin, & Kong, 2013). The metaanalysis of risk factors was summarized graphically using a forest plot. Publication bias was assessed by Begg test and graphed by a funnel plot, a P < 0.10 was considered significant. Furthermore, we estimated pooled ORs including only studies presenting multivariate ORs. All analyses were performed using the software Stata 11.0 (Stata Corporation, College Station, TX).
3. Results 3.1. Literature search results
2. Methods 2.1. Literature search A computerized search was performed on Medline, Embase, and Cochrane central database (all through January 2014) for the studies exploring risk factors for SHF. The main key words were as follows: ‘‘factor’’ or ‘‘predictor’’ or ‘‘risk’’ AND ‘‘subsequent’’ or ‘‘second’’ or ‘‘contralateral’’ or ‘‘recurrent’’ or ‘‘bilateral’’ or ‘‘nonsimultaneous’’ AND ‘‘proximal femur’’ or ‘‘hip’’ or ‘‘intertrochanteric’’ or ‘‘femoral neck’’ AND ‘‘fracture’’. Also, a manual search of references in the identified articles and systematic reviews was performed for possible inclusion. Two reviewers independently evaluated the titles and abstracts of the identified papers. Only full-text articles published in English were included in this meta-analysis. The inclusion criteria were as follows: (1) only subsequent fracture at hip (including ipsilateral and contralateral hip), were considered SHF. (2) A case–control studies was performed to explore risk factors for SHF, controls and cases were defined based on the number of fractures (unilateral or SHF). (3) Sufficient data was published for estimating an OR or SMD with 95%CI.
A total of 242 initially selected references were retrieved, of which 22 studies altogether including 2154 cases and 24,859 controls were identified eligible (Fig. 1). The basic characteristics of included studies and participants are summarized in Table 1. Data on 19 potential risk factors for SHF were abstracted from the included studies, and number of citations for each risk factor were presented in Table 2. Initially selected and identified search (242) Excluded after screening for abstract and reduplicative publication(74) References selected for full text retrieval (168)
Non English studies (29) Non original studies e.g. letters, etc. (13) Irrelevant population and outcomes (63)! Insufficient quantitative data (37)
Included studies (22)
Unqualified methodological assessment
4
Fig. 1. Flow diagram of literature searching.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
3
Table 1 The basic characteristics of these 22 included studies and participants. First author
Publication year
Country
Control
Case
Total
Age (years)
Significant factors
Saxena Chayanin Sawalha Chiu Yamanashi Fukushima Gaumetou Anne Sugao Rodaro Dretakis Christopher Lee Dinah Lawrence Lee Chapurlat Lo¨nnroos Hagino Chang
2000 2009 2012 1992 2005 2006 2011 2012 2010 2004 1998 2012 2012 2002 2010 2013 2003 2007 2012 2013
UK Thailand UK China HK Japan Japan France Netherlands Japan Italy Greece USA Korea UK UK Korea USA Finland Japan Korea
530 97 4708 1479 669 741 215 1060 335 2771 1579 1089 483 164 4746 755 579 467 2251 141
53 28 633 35 45 94 26 169 49 210 106 88 34 22 280 71 53 34 77 47
583 125 5341 1514 714 835 241 1229 384 2981 1685 1177 517 186 5026 826 632 501 2328 188
No data 60 60 57 65 60 60 50 50 50 50 Mean, 80.0 56 Mean, 83.0 55 50 65 60 65 65
Kim Sarah
2012 2007
Korea USA
377 410
38 71
415 481
Alzheimer’s disease, cerebrovascular accident, blindness, syncope and collapsea Hypotrophic change, eye diseases and neurological diseasesb Institutionalized, female, older, lower mobility and mental test scoresa Neurological diseases, osteomalacia changea Dementia and Parkinsonsc Dementiaa Femalea Advanced agea Dementia and respiratory diseased Advanced age and sedentary womena Recurrent fallsa Older age, internal fixation, diabetes and bisphosphonatesa Bisphosphonates for osteoporosisa Femalea Agea Hematologic disease, renal disease, compliant use of bisphosphonatesd BMD, taking Tums, walking for exercise, weight, estrogenb Treatment of first hip fractured Body weightc Respiratory diseases, visual impairments, postoperative delirium, rehabilitation and ambulatory statusc Residency, fracture type, falls, cognitive impairment and diabetes mellitusd Older age and functional statusd
a b c d
60 Mean, 81
Univariate t-test or Chi-square test. Multivariate analyses (not specific). Multivariate logistic regression analyses. Multivariate cox regression analyses.
Pun, Luk, & Chow, 1992; Kim et al., 2012; Vochteloo et al., 2012) scored 6.
3.2. Methodological quality assessment The outcome of quality assessment for included studies was as follows: four studies scored 9 (Dinah, 2002; Gaumetou et al., 2011; Hagino et al., 2012; Lawrence, Wenn, & Boulton, 2010), eight studies (Angthong et al., 2004; Chapurlat, Bauer, Nevitt, Stone, & Cummings, 2003; Lee, Ha, Yoon, & Koo, 2013; Mitani et al., 2010; Rodaro et al., 2004; Sawalha & Parker, 2012; Saxena & Shankar, 2000; Yamanashi et al., 2004) scored 8; seven studies (Berry, Samelson, Hannan, Robert, & Mei Lu Adrienne, 2007; Chang et al., 2013; Dretakis, Dretakis, Papakitsou, P, & Steriopoulos, 1998; Fukushima, Sudo, & Uchida, 2006; Lee, Lee, & Cho, 2012; Souder et al., 2012; Vochteloo et al., 2012) scored 7; three studies (Chiu,
3.3. Pooled analysis of risk factors A meta-analysis of combinable data was conducted to analyze the risk factors for SHF, and the main results were summarized in Table 2. The combined ORs ranged from 1.32 to 10.02. Significant heterogeneity was observed among studies, when evaluating the potential risk factors including age, living in institutions, low vision, diabetes mellitus, respiratory disease, neurological diseases, fracture type, cognitive impairment and osteoarthritis. On the basis of the combined ORs and 95%CI, the significant risk
Table 2 Detailed data on 19 potential risk factors for the SHF and the outcome of meta-analyses. Potential risk Female Age BMI Institutionalized Dementia Low vision Cardiac disease Diabetes mellitus Hypertension Parkinson Respiratory disease Neurological diseases Fracture type (femoral neck) Surgical type (osteosynthesis) Cognitive impairment Osteoarthritis Osteoporosis (SI 1–3) Alcoholism Cancer
No. of studies 15 9 4 4 7 4 7 7 5 2 6 8 9 3 3 3 2 2 2
Pooled OR or SMD 1.46 0.08 0.10 2.23 1.89 2.09 1.32 1.35 0.97 2.90 1.97 1.17 1.06 0.75 2.01 1.47 10.02 1.93 0.71
LL 95%CI 1.29 0.11 0.29 1.29 1.47 1.06 1.02 0.91 0.74 1.41 1.16 0.90 0.82 0.43 0.93 0.50 5.41 0.61 0.32
UL 95%CI 1.66 0.27 0.08 3.83 2.43 4.12 1.70 2.00 1.26 5.95 3.32 1.52 1.35 1.31 4.37 4.26 18.57 6.18 1.57
P value a
<0.001 0.415b 0.144a 0.004b <0.001a 0.034b 0.032a 0.132b 0.806a 0.004 0.011b 0.244b 0.672b 0.311b 0.077b 0.483b <0.001a 0.266a 0.400a
Q-test (P)
I2 (%)c
0.303 <0.001 0.282 0.011 0.483 0.038 0.591 0.042 0.364 0.702 0.075 0.032 0.007 0.078 0.072 0.03 0.182 0.550 0.463
13.4 84.6 44.6 73.2 0 64.3 0 54 7.4 0 50.0 54.4 61.8 60.7 62.0 71.5 43.9 0 0
Abbreviations: LL, lower limit; UL, upper limit. a Fixed-effects model was performed. b Fandom-effects model was performed. c 2 I statistic was defined as the proportion of heterogeneity not due to chance or random error.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 4
Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
factors were female (OR, 1.46; 95%CI, 1.29–1.66), living in institutions (OR, 2.23; 95%CI, 1.29–3.83), osteoporosis (SI 1–3) (OR, 10.02; 95%CI, 5.41–18.57), low vision (OR, 2.09; 95%CI, 1.06– 4.12), dementia (OR, 1.89; 95%CI, 1.47–2.43), Parkinson (OR, 2.90; 95%CI, 1.41–5.95), cardiac diseases (OR, 1.32; 95%CI, 1.02–1.70) and respiratory disease (OR, 1.97; 95%CI, 1.16–3.32). No significant difference was found in admission age between patients with the unilateral hip fracture and the first hip fracture of SHF groups (SMD, 0.08; 95%CI, 0.11 to 0.27, P = 0.415). The outcome of analysis for female, dementia, cardiac diseases and respiratory disease, as significant risks were presented by forest plots (Fig. 2) and others, were presented in Table 2. Fracture types, BMI, alcoholism, surgical types, cognitive impairment, osteoarthritis, cancer, diabetes mellitus, hypertension and neurological diseases were identified not as the risk factors for SHF (P > 0.05). Funnel plot and Begg’ test showed that no significant publication bias was found in the studies concerning the risk of gender (P = 0.767) (Fig. 3). The detailed results of pooled ORs computed for only for studies presenting multivariate analysis are presented in Supplementary Table 1. The results revealed that the exclusion of the more biased crude ORs did not affect the statistical significance of the associations among the most potential risk factors including gender, dementia, cardiac disease, hypertension, Parkinson, respiratory disease, neurological diseases; hence, the metaanalytic associations reported in this review for these risk factors demonstrate robust to the extent. In terms of the risk factors for low vision (OR, 2.31; 95%CI, 0.86–6.22 for the multivariate subgroup) and cognitive impairment (OR, 1.59; 95%CI, 1.05–2.41
for the multivariate subgroup), the results were revealed to be inconsistent with those obtained from the crude ORs, however, significant heterogeneity was detected for low vision (P = 0.02) but not for cognitive impairment (P = 0.733). See Supplementary Table 1 as supplementary file. Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.archger.2014.02.012. 4. Discussion SHF after the initial hip fracture is not rare in elderly patients and the incidence of SHF was 2.3–13.8% (Chiu et al., 1992; Dinah, 2002; Finsen & Benum, 1986; Gaumetou et al., 2011; Schroder & Petersen, 1993; Vochteloo et al., 2012; Yamanashi et al., 2004), almost 80% occurred in the three years after IHF (Dretakis et al., 1998; Mitani et al., 2010; Sawalha & Parker, 2012). The present systematic review and meta-analysis revealed multiple risk factors associated with SHF in IHF elderly patients. Statistically significant associations were identified for female, osteoporosis (SI 1–3), low vision, dementia, cardiac diseases and respiration diseases. No significant difference was found in admission age among patients with the unilateral hip fracture and the first hip fracture of SHF groups, therefore, older age was not a negative factor for SHF and it was not advisable to screen for patients at risk of SHF relying solely on the age at which the initial hip fracture occurred. Living in institutions after IHF was analyzed (Lo¨nnroos, Kautiainen, Karppi, H, & Kiviranta, 2007; Mitani et al., 2010; Sawalha & Parker, 2012), but its potential role as a risk factor for SHF was inclusive. It is probably that significant differences in
Fig. 2. Forest plots of the meta-analysis of female (A), osteoporosis (SI 1–3) (B), low vision (C), dementia (D), cardiac diseases (E) and respiratory disease (F) as risk factors for the SHFs after initial hip fractures. The width of the horizontal line represents the 95%CI of the individual studies, and the square proportional represents the weight of each study. The diamond represents the pooled OR and 95%CI.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
5
Fig. 2. (Continued ).
methodology, including sample sizes and a broad range of institutions examined, contributed to significant heterogeneity in the analytical results. Clinically, many factors (including family dynamics and chronic pathological conditions) could influence the need for IHF patients to be admitted to these institutions. Furthermore, living in institutions was a significant risk for not regaining pre-fracture level of mobility and resulted in worse functional outcome, as reported by Dharmarajan (2001) and Beaupre et al. (2007). Therefore, whether living in institutions as an independent factor influencing SHF needed to be further prospectively investigated. In this review, due to the significant heterogeneity after sensitive and subgroup analysis, pooled result was unreliable. Likewise, significant heterogeneity was found for studies of recurrent falls, indicating not a proper quantitative risk assessment. This review also extends previous work by showing that cardiac disease is a strongly significant (P = 0.032) factor associated with the occurrence of SHF, which benefited directly from pooled samples of individual studies. As a matter of fact, the negative impact of cardiac diseases is complex. Firstly, this can be explained by the common pathophysiological mechanisms that cardiac diseases and hip fracture share, including calcification regulatory hormones, sex steroids, oxidative stress and chronic inflammation (McFarlane et al., 2004; Sennerby et al., 2009). Secondly, this may be partly due to the greater propensity for falls in those with cardiac diseases, because balance disturbances might occur due to sudden myocardial ischemia and the adverse effect of multiple medications (Hartikainen, Lo¨nnroos, & Louhivuori, 2007). Further studies are required to elucidate the mechanism whereby cardiac disease influenced the SHF. Undoubtedly, effective strategies for Begg's funnel plot with pseudo 95% confidence limits
2
log[var2]
1
0
-1 0
.2
.4
.6
.8
s.e. of: log[var2] Fig. 3. Begg’s funnel plot for publication bias (with 95% pseudoconfidence limits) of the case–control studies that investigated female and risk factor for the SHF (P = 0.767).
management and secondary prevention of hip fracture among patients with initial hip fracture must be developed, for instance through implementing routine home visits for medications by physiotherapists to help prevent falls. It is notable that, the majority of above-motioned risk factors (gender, low vision, dementia, cardiac diseases and respiration diseases) could not be modified, however, these factors should be kept in mind to prepare prevention strategy or reminder with the aim to reduce the risk of SHF for patients with such factors. The osteoporosis as a significant risk factor was potentially amenable. As reported by Bischoff et al. (Bischoff-Ferrari et al., 2009), a vitamin D high dose of 700–800 IU/d could significantly reduce the RR of the hip fracture. Likewise, the compliant use of biphosphonates have been identified to reduce 60% of the occurrence of SHF (4.2% vs. 10.9%) (Lee et al., 2013). Lower bone mineral density have been undoubtedly regarded as the most important factors associated with SHF, although relevant studies were not pooled and calculated in present study due to the inconsistent quantitative criteria (mean or count) and measurement in different sites (calcaneus, total hip or vertebra) for each study. Likewise, American Society of Anesthesiologists (ASA) physical status score were not pooled for the non-uniform quantitative criteria. Because we could not obtain the original raw data, it was impossible to perform a meta-analysis of BMD or ASA. In addition, there were other risk factors that could not be combined because studies were reported in fewer studies. For example, hematologic disease (Lee et al., 2013), renal disease (Lee et al., 2013) and smoking (Souder et al., 2012) were reported in only a single study. This meta-analysis has some limitations. Firstly, a weakness exists in the analyses that not all the ORs on the potential risk factors for the meta-analysis are adjusted because a lot of reports could only provide the univariate rather than multivariate statistics; likewise, some studies might choose not to report the insignificant results or results of no interest, potentially resulting in a considerable amounts of missing data. Hence, our overall effect estimates may be an over-estimate but the multivariate subgroup analysis was performed and the corresponding pooled results are reliable. Secondly, all the included studies were case–control studies with unavoidable recall and interviewer biases, which might affect the associations between the risk and SHF. Thirdly, the measurements of various risk factors differed much from each other, either by standard grading forms or by patient-reported questionnaires. Similarly, follow-up periods ranged widely from several months to more than 10 years. Therefore, a significant heterogeneity was unavoidable in this review. Finally, all the subjects in 22 papers were Asian, European and US; therefore, our conclusions might not extrapolate to African and Australasian populations.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 6
Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
Despite these limitations, some advantages should be mentioned. First and the most important, cardiac disease as a risk factor was identified, which might contribute to complementing prevention strategies for patients involved. Secondly, a comprehensive search strategy based on computer-assisted and manual searching avoided any eligible study to omission. Finally, no publications bias was found, suggesting a robust result. In summary, the present meta-analysis suggests that females, living in institutions, osteoporosis (SI 1–3) low vision, dementia, Parkinson, respiration diseases and cardiac diseases were significant risk factors for SHF. Furthermore, cardiac disease as the significantly increased risk for SHF was firstly reported and related prevention strategies should be formulated. Conflict of interest All the authors declare that they have no conflict of interest with any organization that sponsored the research. Acknowledgements We are grateful to S.L. and Z.W. of the Department of Orthopedics, and to X.C. and G.W. of the Department of Statistics and Applications for their kind assistance. References Angthong, C., Suntharapa, T., & Harnroongroj, T. (2004). Major risk factors for the second contralateral hip fracture in the elderly. Acta Orthopaedica et Traumatologica Turcica, 43, 193–198. Beaupre, L. A., Cinats, J. G., Jones, C. A., Angela, V. S., Johnston, D. W., Ambikaipan, S., et al. (2007). Does functional recovery in elderly hip fracture patients differ between patients admitted from long-term care and the community? Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 62, 1127–1133. Berry, S. D., Samelson, E. J., Hannan, M. T., Robert, R. M., Lu, M., Adrienne, C., et al. (2007). Second hip fracture in older men and women: The Framingham Study. Archives of Internal Medicine, 167, 1971–1976. Bischoff-Ferrari, H. A., Willett, W. C., Wong, J. B., Stuck, A. E., Staehelin, H. B., Orav, E. J., et al. (2009). Prevention of nonvertebral fractures with oral vitamin D and dose dependency: A meta-analysis of randomized controlled trials. Archives of Internal Medicine, 169, 551–561. Chang, J.-D., Yoo, J.-H., Reddy, P., Lee, S.-S., Hwang, J.-H., & Kim, T.-Y. (2013). Risk factors for contra-lateral hip fracture in elderly patients with previous hip fracture. Injury, 44, 1930–1933. Chapurlat, R., Bauer, D., Nevitt, M., Stone, K., & Cummings, S. R. (2003). Incidence and risk factors for a second hip fracture in elderly women. The Study of Osteoporotic Fractures. Osteoporosis International, 14, 130–136. Chiu, K., Pun, W., Luk, K., & Chow, S. (1992). Sequential fractures of both hips in elderly patients – A prospective study. Journal of Trauma and Acute Care Surgery, 32, 584– 587. Dharmarajan, T., Tankala, H., Patel, B., Sipalay, M., & Norkus, E. P. (2001). Outcome in ambulatory status immediately following hip fracture surgery in the acute setting: A comparison of nursing home residents and community older adults. Journal of the American Medical Directors Association, 2, 115–119. Dinah, A. (2002). Sequential hip fractures in elderly patients. Injury, 33, 393–394. Dretakis, K., Dretakis, E., Papakitsou, E., Psarakis, & Steriopoulos, K. (1998). Possible predisposing factors for the second hip fracture. Calcified Tissue International, 62, 366–369. Egan, M., Jaglal, S., Byrne, K., Jennie, W., & Paul, S. (2008). Factors associated with a second hip fracture: A systematic review. Clinical Rehabilitation, 22, 272–282. Finsen, V., & Benum, P. (1986). The second hip fracture an epidemiologic study. Acta Orthopaedica, 57, 431–433. Fredman, L., Magaziner, J., Hawkes, W., JR, H., Linda, P. F., Judith, K., et al. (2005). Female hip fracture patients had poorer performance-based functioning than communitydwelling peers over 2-year follow-up period. Journal of Clinical Epidemiology, 58, 1289–1298. Fukushima, T., Sudo, A., & Uchida, A. (2006). Bilateral hip fractures. Journal of Orthopaedic Science, 11, 435–438.
Galois, L., Dausse, Y., Mainard, D., & Delagoutte, J. P. (2001). Bilateral noncontemporary hip fractures. 31 cases reported. European Journal of Orthopaedic Surgery & Traumatology, 11, 219–223. Gaumetou, E., Zilber, S., & Hernigou, P. (2011). Non-simultaneous bilateral hip fracture: Epidemiologic study of 241 hip fractures. Orthopaedics & Traumatology: Surgery & Research, 97, 22–27. Hagino, H., Sawaguchi, T., Endo, N., Yasuyo, I., Tetsuo, N., & Yoshinobu, W. (2012). The risk of a second hip fracture in patients after their first hip fracture. Calcified Tissue International, 90, 14–21. Hartikainen, S., Lo¨nnroos, E., & Louhivuori, K. (2007). Medication as a risk factor for falls: Critical systematic review. Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 62, 1172–1181. Kim, S.-M., Moon, Y.-W., Lim, S.-J., Yoonb, B.-K., Minc, Y.-K., Lee, D.-Y., et al. (2012). Prediction of survival, second fracture, and functional recovery following the first hip fracture surgery in elderly patients. Bone, 50, 1343–1350. Lau, J., Ioannidis, J. P., & Schmid, C. H. (1997). Quantitative synthesis in systematic reviews. Annals of Internal Medicine, 127, 820–826. Lawrence, T., Wenn, R., & Boulton, C. (2010). Age-specific incidence of first and second fractures of the hip. Journal of Bone & Joint Surgery British, 92, 258–261. Lee, S. H., Lee, T. J., & Cho, K. J. (2012). Subsequent hip fracture in osteoporotic hip fracture patients. Yonsei Medical Journal, 53, 1005–1009. Lee, Y.-K., Ha, Y.-C., Yoon, B.-H., & Koo, K.-H. (2013). Incidence of second hip fracture and compliant use of bisphosphonate. Osteoporosis International, 24, 2099–2104. Lo¨nnroos, E., Kautiainen, H., Karppi, P., Hartikainen, & Kiviranta, R. (2007). Sulkava Incidence of second hip fractures. A population-based study. Osteoporosis International, 18, 1279–1285. McFarlane, S. I., Muniyappa, R., Shin, J. J., John, J. S., Gul, B., James, R. S., et al. (2004). Osteoporosis and cardiovascular disease. Endocrine, 23, 1–10. Mitani, S., Shimizu, M., Abo, M., Hiroshi, H., & Youichi, K. (2010). Risk factors for second hip fractures among elderly patients. Journal of Orthopaedic Science, 15, 192–197. Omsland, T. K., Holvik, K., Meyer, H. E., Jacqueline, R. C., Nina, E., Tell, G. S., et al. (2012). Hip fractures in Norway 1999–2008: Time trends in total incidence and second hip fracture rates. A NOREPOS study. European Journal of Epidemiology, 27, 807–814. Pearse, E., Redfern, D., Sinha, M., & A, J. E. (2004). Outcome following a second hip fracture. Injury, 34, 518–521. Penrod, J. D., Litke, A., Hawkes, W. G., Kenneth, J., Koval, M. D., John, T., et al. (2007). Heterogeneity in hip fracture patients: Age, functional status and comorbidity. Journal of the American Geriatrics Society, 55, 407–413. Rodaro, E., Pasqualini, M., Iona, L., & Di Benedetto, (2004). Functional recovery following a second hip fracture. Europa Medicophysica, 40, 179–183. Ryg, J., Rejnmark, L., Overgaard, S., Kim, B., & Peter, V. (2009). Hip fracture patients at risk of second hip fracture: A nationwide population based cohort study of 169,145 cases during 1977–2001. Journal of Bone and Mineral Research, 24, 1299–1307. Sawalha, S., & Parker, M. (2012). Characteristics and outcome in patients sustaining a second contralateral fracture of the hip. Journal of Bone & Joint Surgery British, 94, 102–106. Saxena, P., & Shankar, J. (2000). Contralateral hip fractures – Can predisposing factors be determined? Injury, 31, 421–424. Schroder, H. M., Petersen, K. K., & Erlandsen, M. (1993). Occurrence and incidence of the second hip fracture. Clinical Orthopaedics and Related Research, 289, 166–169. Sennerby, U., Melhus, H., Gedeborg, R., Liisa, B., Hans, G., Anders, A., et al. (2009). Cardiovascular diseases and risk of hip fracture. Journal of the American Medical Association, 302, 1666–1673. Souder, C. D., Brennan, M. L., Brennan, K. L., Juhee, S., Johnathan, W., & Christopher, C. (2012). The rate of contralateral proximal femoral fracture following closed reduction and percutaneous pinning compared with arthroplasty for the treatment of femoral neck fractures. Journal of Bone & Joint Surgery, 94, 418–425. Stang, A. (2004). Critical evaluation of the Newcastle–Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. European Journal of Epidemiology, 25, 603–605. Van Balen, R., Steyerberg, E., Polder, J., Ribbers, T., Habbema, J., & Cools, H. (2001). Hip fracture in elderly patients: Outcomes for function, quality of life, and type of residence. Clinical Orthopaedics and Related Research, 390, 232–243. Vestergaard, P., Rejnmark, L., & Mosekilde, L. (2007). Has mortality after a hip fracture increased? Journal of the American Geriatrics Society, 55, 1720–1726. Vochteloo, A. J., van der Burg, B. L. B., Ro¨ling, M. A., Ro¨ling, M. A., van Leeuwen, D. H., van den Berg, P., et al. (2012). Contralateral hip fractures and other osteoporosisrelated fractures in hip fracture patients: Incidence and risk factors. An observational cohort study of 1229 patients. Archives of Orthopaedic and Trauma Surgery, 132, 1191–1197. Wei, J., Yang, T.-B., Luo, W., Qin, J.-B., & Kong, F.-J. (2013). Complications following dorsal versus volar plate fixation of distal radius fracture: A meta-analysis. Journal of International Medical Research, 41, 265–275. Yamanashi, A., Yamazaki, K., Kanamori, M., Kuniyoshi, M., Shozo, O., Youichi, K., et al. (2004). Assessment of risk factors for second hip fractures in Japanese elderly. Osteoporosis International, 16, 1239–1246. Zhang, Y. (2012). Clinical epidemiology of orthopedic trauma (1st ed.). New York and Stuggart: Thieme.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
AGG-2967; No. of Pages 6 Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Archives of Gerontology and Geriatrics journal homepage: www.elsevier.com/locate/archger
Review
Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients Yanbin Zhu a,b, Wei Chen a,b, Tao Sun a,b, Qi Zhang a,b, Jiaxiang Cheng a,b, Yingze Zhang a,b,* a b
Department of Orthopaedic Surgery, The Third Hospital of Hebei Medical University, Shijiazhuang, Hebei 050051, PR China Key Laboratory of Biomechanics of Hebei Province, Shijiazhuang, Hebei 050051, PR China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 2 December 2013 Received in revised form 24 February 2014 Accepted 27 February 2014 Available online xxx
This study aims to quantitatively summarize the risk factors for the incidence of SHF. A meta-analysis was performed with the data obtained from 22 relevant papers published in Pubmed, Embase and Cochrane central database (all through January 2014) following strict selection. The pooled odds ratios (ORs) or standardized mean difference (SMD) with 95% confidence intervals (CIs) were calculated for potential risk factors associated with SHF. Our meta-analysis indicated the significant risk factors for SHF were female (OR, 1.46; 95%CI, 1.29–1.66), living in institutions (OR, 2.23; 95%CI, 1.29–3.83), osteoporosis (Singh index (SI) 1–3) (OR, 10.02; 95%CI, 5.41–18.57), low vision (OR, 2.09; 95%CI, 1.06–4.12), dementia (OR, 1.89; 95%CI, 1.47–2.43), Parkinson (OR, 2.90; 95%CI, 1.41–5.95), cardiac diseases (OR, 1.32; 95%CI, 1.02–1.70) and respiratory disease (OR, 1.97; 95%CI, 1.16–3.32). Related strategies must be implemented on those involved with above-mentioned medical conditions to effectively prevent a SHF. ß 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Second hip fracture Elderly patients Risk factor Meta-analysis
Contents 1. 2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Literature search . . . . . . . . . . . . . . . . Data extraction . . . . . . . . . . . . . . . . . 2.2. Quality of included studies. . . . . . . . 2.3. Meta-analyses . . . . . . . . . . . . . . . . . . 2.4. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Literature search results . . . . . . . . . . 3.1. Methodological quality assessment . 3.2. Pooled analysis of risk factors . . . . . 3.3. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
1. Introduction Hip fracture is the most common fracture in elderly patients and accounts for 20% of all the fractures in this population (Zhang,
* Corresponding author at: Department of Orthopaedics, The Third Hospital, Hebei Medical University, No. 139 Ziqiang Road, Shijiazhuang 050051, PR China. Tel.: +86 311 88603682; fax: +86 311 87023626. E-mail addresses: [email protected], [email protected] (Y. Zhang).
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
. . . . . . . . . . . . .
000 000 000 000 000 000 000 000 000 000 000 000 000
2012). Consequently, patients involved with hip fractures suffered functional deterioration, limited mobility and increased morbidity, especially at the first year post operation (Fredman et al., 2005; Penrod et al., 2007; Vestergaard, Rejnmark, & Mosekilde, 2007). Furthermore, among the survivors of a first hip fracture, the agestandardized risk of a SHF was 2.5 times higher in women and 4.6 times higher in men compared to the risk of a first hip fracture (Omsland et al., 2012). And worse still, patients had a greater mortality and more decreased dependence in their daily activities after SHF (Pearse, Redfern, Sinha, & A, 2004; Sawalha & Parker,
http://dx.doi.org/10.1016/j.archger.2014.02.012 0167-4943/ß 2014 Elsevier Ireland Ltd. All rights reserved.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
2
2012). About 31.6% (Sawalha & Parker, 2012) died at the first year after the operation, 78.3% would have to survive with the aid of walking tools, and most would never recover to their pre-operative level (Galois, Dausse, Mainard, & Delagoutte, 2001). Besides the personal influence to patients, a heavy economic and social burden should be also considered, a proximally financial cost of over 16,000 dollar per patient for unilateral hip fracture and at least doubled for bilateral hip fracture (Van Balen et al., 2001), still not including postoperative care cost after discharge. Therefore, it is vital to identify risk factors to prevent the occurrence of SHF in those sustaining initial hip fractures. Many risk factors have been identified for the SHF, including advanced age (Angthong, Suntharapa, & Harnroongroj, 2004; Souder et al., 2012; Vochteloo et al., 2012), female (Gaumetou, Zilber, & Hernigou, 2011; Rodaro, Pasqualini, Iona, & Di Benedetto, 2004), osteoporosis (Angthong et al., 2004; Yamanashi et al., 2004), dementia (Mitani, Shimizu, Abo, Hiroshi, & Youichi, 2010; Saxena & Shankar, 2000; Yamanashi et al., 2004), Parkinson’s disease (Saxena & Shankar, 2000; Yamanashi et al., 2004), respiratory disease (Mitani et al., 2010) and lower vision (Angthong et al., 2004; Saxena & Shankar, 2000). However, these studies had some limitations such as a small sample size and the inclusion of a single or very few potential risk factors. In addition, some of the results obtained were controversial. For example, female and advanced age were both considered risk factors for SHF in some studies (Rodaro et al., 2004), however, were neither significant in other study (Mitani et al., 2010). Mary et al. (Egan, Jaglal, Byrne, Jennie, & Paul, 2008) made a systematic review of studies on SHF in 2006 with a qualitative description rather than a quantitative metaanalysis on the potential risks. Recently, Jesper et al. (Ryg, Rejnmark, Overgaard, Kim, & Peter, 2009) have conducted a nationwide population-based cohort study of 169, 145 cases to investigate the risks for SHF. Although with a greatest number of cases by far, the study had some certain limitations such as the misclassification of infections, periprosthetic fracture, alloplastic surgery and removal of osteosynthesis as SHF. Therefore, it was difficult for them to get the precise information with regard to the risk factors for SHF. To achieve a reliable and comprehensive conclusion, a metaanalysis was performed using data obtained from previous studies to evaluate the risk factors for SHF in elderly patients with initial hip fractures, and subsequently reduce the occurrence of SHF.
2.2. Data extraction All the data were carefully extracted from all eligible studies independently by the two reviewers, Chen and Sun. The following variables were extracted from each study: first author’s name, publication year, country, significant risk factors, definitions and numbers of case and control groups, numbers of citations for each potential risk factors for SHF. Any disagreement was settled by discussion and a consensus was reached for all data. 2.3. Quality of included studies The quality of the included studies was evaluated using the Newcastle–Ottawa Scale (Stang, 2004): based on the three main items: the selection of the study groups (0–4 points), the comparability of the groups (0–2 points) and the determination of either the exposure or the outcome of interest (0–3 points), with a perfect score of 9. 2.4. Meta-analyses For each risk factor, we extracted the adjusted OR or relative risk (RR), with its 95%CI. When the adjusted OR or the RR was not obtained, we computed a crude OR and pooled across studies to assess the associations between different variables and the risk of SHFs with a P < 0.05 indicating significant difference. Heterogeneity between the studies was tested by Q-test statistics with significance set at P < 0.10 (Lau, Ioannidis, & Schmid, 1997). The I2 statistics were used as a second measure of heterogeneity, with I2 more than 50% indicating inconsistency. A random effects model was used to calculate pooled ORs or SMDs in the case of significant heterogeneity (P < 0.10 or I2 > 50%); otherwise, a fixed-effects model was used (Wei, Yang, Luo, Qin, & Kong, 2013). The metaanalysis of risk factors was summarized graphically using a forest plot. Publication bias was assessed by Begg test and graphed by a funnel plot, a P < 0.10 was considered significant. Furthermore, we estimated pooled ORs including only studies presenting multivariate ORs. All analyses were performed using the software Stata 11.0 (Stata Corporation, College Station, TX).
3. Results 3.1. Literature search results
2. Methods 2.1. Literature search A computerized search was performed on Medline, Embase, and Cochrane central database (all through January 2014) for the studies exploring risk factors for SHF. The main key words were as follows: ‘‘factor’’ or ‘‘predictor’’ or ‘‘risk’’ AND ‘‘subsequent’’ or ‘‘second’’ or ‘‘contralateral’’ or ‘‘recurrent’’ or ‘‘bilateral’’ or ‘‘nonsimultaneous’’ AND ‘‘proximal femur’’ or ‘‘hip’’ or ‘‘intertrochanteric’’ or ‘‘femoral neck’’ AND ‘‘fracture’’. Also, a manual search of references in the identified articles and systematic reviews was performed for possible inclusion. Two reviewers independently evaluated the titles and abstracts of the identified papers. Only full-text articles published in English were included in this meta-analysis. The inclusion criteria were as follows: (1) only subsequent fracture at hip (including ipsilateral and contralateral hip), were considered SHF. (2) A case–control studies was performed to explore risk factors for SHF, controls and cases were defined based on the number of fractures (unilateral or SHF). (3) Sufficient data was published for estimating an OR or SMD with 95%CI.
A total of 242 initially selected references were retrieved, of which 22 studies altogether including 2154 cases and 24,859 controls were identified eligible (Fig. 1). The basic characteristics of included studies and participants are summarized in Table 1. Data on 19 potential risk factors for SHF were abstracted from the included studies, and number of citations for each risk factor were presented in Table 2. Initially selected and identified search (242) Excluded after screening for abstract and reduplicative publication(74) References selected for full text retrieval (168)
Non English studies (29) Non original studies e.g. letters, etc. (13) Irrelevant population and outcomes (63)! Insufficient quantitative data (37)
Included studies (22)
Unqualified methodological assessment
4
Fig. 1. Flow diagram of literature searching.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
3
Table 1 The basic characteristics of these 22 included studies and participants. First author
Publication year
Country
Control
Case
Total
Age (years)
Significant factors
Saxena Chayanin Sawalha Chiu Yamanashi Fukushima Gaumetou Anne Sugao Rodaro Dretakis Christopher Lee Dinah Lawrence Lee Chapurlat Lo¨nnroos Hagino Chang
2000 2009 2012 1992 2005 2006 2011 2012 2010 2004 1998 2012 2012 2002 2010 2013 2003 2007 2012 2013
UK Thailand UK China HK Japan Japan France Netherlands Japan Italy Greece USA Korea UK UK Korea USA Finland Japan Korea
530 97 4708 1479 669 741 215 1060 335 2771 1579 1089 483 164 4746 755 579 467 2251 141
53 28 633 35 45 94 26 169 49 210 106 88 34 22 280 71 53 34 77 47
583 125 5341 1514 714 835 241 1229 384 2981 1685 1177 517 186 5026 826 632 501 2328 188
No data 60 60 57 65 60 60 50 50 50 50 Mean, 80.0 56 Mean, 83.0 55 50 65 60 65 65
Kim Sarah
2012 2007
Korea USA
377 410
38 71
415 481
Alzheimer’s disease, cerebrovascular accident, blindness, syncope and collapsea Hypotrophic change, eye diseases and neurological diseasesb Institutionalized, female, older, lower mobility and mental test scoresa Neurological diseases, osteomalacia changea Dementia and Parkinsonsc Dementiaa Femalea Advanced agea Dementia and respiratory diseased Advanced age and sedentary womena Recurrent fallsa Older age, internal fixation, diabetes and bisphosphonatesa Bisphosphonates for osteoporosisa Femalea Agea Hematologic disease, renal disease, compliant use of bisphosphonatesd BMD, taking Tums, walking for exercise, weight, estrogenb Treatment of first hip fractured Body weightc Respiratory diseases, visual impairments, postoperative delirium, rehabilitation and ambulatory statusc Residency, fracture type, falls, cognitive impairment and diabetes mellitusd Older age and functional statusd
a b c d
60 Mean, 81
Univariate t-test or Chi-square test. Multivariate analyses (not specific). Multivariate logistic regression analyses. Multivariate cox regression analyses.
Pun, Luk, & Chow, 1992; Kim et al., 2012; Vochteloo et al., 2012) scored 6.
3.2. Methodological quality assessment The outcome of quality assessment for included studies was as follows: four studies scored 9 (Dinah, 2002; Gaumetou et al., 2011; Hagino et al., 2012; Lawrence, Wenn, & Boulton, 2010), eight studies (Angthong et al., 2004; Chapurlat, Bauer, Nevitt, Stone, & Cummings, 2003; Lee, Ha, Yoon, & Koo, 2013; Mitani et al., 2010; Rodaro et al., 2004; Sawalha & Parker, 2012; Saxena & Shankar, 2000; Yamanashi et al., 2004) scored 8; seven studies (Berry, Samelson, Hannan, Robert, & Mei Lu Adrienne, 2007; Chang et al., 2013; Dretakis, Dretakis, Papakitsou, P, & Steriopoulos, 1998; Fukushima, Sudo, & Uchida, 2006; Lee, Lee, & Cho, 2012; Souder et al., 2012; Vochteloo et al., 2012) scored 7; three studies (Chiu,
3.3. Pooled analysis of risk factors A meta-analysis of combinable data was conducted to analyze the risk factors for SHF, and the main results were summarized in Table 2. The combined ORs ranged from 1.32 to 10.02. Significant heterogeneity was observed among studies, when evaluating the potential risk factors including age, living in institutions, low vision, diabetes mellitus, respiratory disease, neurological diseases, fracture type, cognitive impairment and osteoarthritis. On the basis of the combined ORs and 95%CI, the significant risk
Table 2 Detailed data on 19 potential risk factors for the SHF and the outcome of meta-analyses. Potential risk Female Age BMI Institutionalized Dementia Low vision Cardiac disease Diabetes mellitus Hypertension Parkinson Respiratory disease Neurological diseases Fracture type (femoral neck) Surgical type (osteosynthesis) Cognitive impairment Osteoarthritis Osteoporosis (SI 1–3) Alcoholism Cancer
No. of studies 15 9 4 4 7 4 7 7 5 2 6 8 9 3 3 3 2 2 2
Pooled OR or SMD 1.46 0.08 0.10 2.23 1.89 2.09 1.32 1.35 0.97 2.90 1.97 1.17 1.06 0.75 2.01 1.47 10.02 1.93 0.71
LL 95%CI 1.29 0.11 0.29 1.29 1.47 1.06 1.02 0.91 0.74 1.41 1.16 0.90 0.82 0.43 0.93 0.50 5.41 0.61 0.32
UL 95%CI 1.66 0.27 0.08 3.83 2.43 4.12 1.70 2.00 1.26 5.95 3.32 1.52 1.35 1.31 4.37 4.26 18.57 6.18 1.57
P value a
<0.001 0.415b 0.144a 0.004b <0.001a 0.034b 0.032a 0.132b 0.806a 0.004 0.011b 0.244b 0.672b 0.311b 0.077b 0.483b <0.001a 0.266a 0.400a
Q-test (P)
I2 (%)c
0.303 <0.001 0.282 0.011 0.483 0.038 0.591 0.042 0.364 0.702 0.075 0.032 0.007 0.078 0.072 0.03 0.182 0.550 0.463
13.4 84.6 44.6 73.2 0 64.3 0 54 7.4 0 50.0 54.4 61.8 60.7 62.0 71.5 43.9 0 0
Abbreviations: LL, lower limit; UL, upper limit. a Fixed-effects model was performed. b Fandom-effects model was performed. c 2 I statistic was defined as the proportion of heterogeneity not due to chance or random error.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 4
Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
factors were female (OR, 1.46; 95%CI, 1.29–1.66), living in institutions (OR, 2.23; 95%CI, 1.29–3.83), osteoporosis (SI 1–3) (OR, 10.02; 95%CI, 5.41–18.57), low vision (OR, 2.09; 95%CI, 1.06– 4.12), dementia (OR, 1.89; 95%CI, 1.47–2.43), Parkinson (OR, 2.90; 95%CI, 1.41–5.95), cardiac diseases (OR, 1.32; 95%CI, 1.02–1.70) and respiratory disease (OR, 1.97; 95%CI, 1.16–3.32). No significant difference was found in admission age between patients with the unilateral hip fracture and the first hip fracture of SHF groups (SMD, 0.08; 95%CI, 0.11 to 0.27, P = 0.415). The outcome of analysis for female, dementia, cardiac diseases and respiratory disease, as significant risks were presented by forest plots (Fig. 2) and others, were presented in Table 2. Fracture types, BMI, alcoholism, surgical types, cognitive impairment, osteoarthritis, cancer, diabetes mellitus, hypertension and neurological diseases were identified not as the risk factors for SHF (P > 0.05). Funnel plot and Begg’ test showed that no significant publication bias was found in the studies concerning the risk of gender (P = 0.767) (Fig. 3). The detailed results of pooled ORs computed for only for studies presenting multivariate analysis are presented in Supplementary Table 1. The results revealed that the exclusion of the more biased crude ORs did not affect the statistical significance of the associations among the most potential risk factors including gender, dementia, cardiac disease, hypertension, Parkinson, respiratory disease, neurological diseases; hence, the metaanalytic associations reported in this review for these risk factors demonstrate robust to the extent. In terms of the risk factors for low vision (OR, 2.31; 95%CI, 0.86–6.22 for the multivariate subgroup) and cognitive impairment (OR, 1.59; 95%CI, 1.05–2.41
for the multivariate subgroup), the results were revealed to be inconsistent with those obtained from the crude ORs, however, significant heterogeneity was detected for low vision (P = 0.02) but not for cognitive impairment (P = 0.733). See Supplementary Table 1 as supplementary file. Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.archger.2014.02.012. 4. Discussion SHF after the initial hip fracture is not rare in elderly patients and the incidence of SHF was 2.3–13.8% (Chiu et al., 1992; Dinah, 2002; Finsen & Benum, 1986; Gaumetou et al., 2011; Schroder & Petersen, 1993; Vochteloo et al., 2012; Yamanashi et al., 2004), almost 80% occurred in the three years after IHF (Dretakis et al., 1998; Mitani et al., 2010; Sawalha & Parker, 2012). The present systematic review and meta-analysis revealed multiple risk factors associated with SHF in IHF elderly patients. Statistically significant associations were identified for female, osteoporosis (SI 1–3), low vision, dementia, cardiac diseases and respiration diseases. No significant difference was found in admission age among patients with the unilateral hip fracture and the first hip fracture of SHF groups, therefore, older age was not a negative factor for SHF and it was not advisable to screen for patients at risk of SHF relying solely on the age at which the initial hip fracture occurred. Living in institutions after IHF was analyzed (Lo¨nnroos, Kautiainen, Karppi, H, & Kiviranta, 2007; Mitani et al., 2010; Sawalha & Parker, 2012), but its potential role as a risk factor for SHF was inclusive. It is probably that significant differences in
Fig. 2. Forest plots of the meta-analysis of female (A), osteoporosis (SI 1–3) (B), low vision (C), dementia (D), cardiac diseases (E) and respiratory disease (F) as risk factors for the SHFs after initial hip fractures. The width of the horizontal line represents the 95%CI of the individual studies, and the square proportional represents the weight of each study. The diamond represents the pooled OR and 95%CI.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
5
Fig. 2. (Continued ).
methodology, including sample sizes and a broad range of institutions examined, contributed to significant heterogeneity in the analytical results. Clinically, many factors (including family dynamics and chronic pathological conditions) could influence the need for IHF patients to be admitted to these institutions. Furthermore, living in institutions was a significant risk for not regaining pre-fracture level of mobility and resulted in worse functional outcome, as reported by Dharmarajan (2001) and Beaupre et al. (2007). Therefore, whether living in institutions as an independent factor influencing SHF needed to be further prospectively investigated. In this review, due to the significant heterogeneity after sensitive and subgroup analysis, pooled result was unreliable. Likewise, significant heterogeneity was found for studies of recurrent falls, indicating not a proper quantitative risk assessment. This review also extends previous work by showing that cardiac disease is a strongly significant (P = 0.032) factor associated with the occurrence of SHF, which benefited directly from pooled samples of individual studies. As a matter of fact, the negative impact of cardiac diseases is complex. Firstly, this can be explained by the common pathophysiological mechanisms that cardiac diseases and hip fracture share, including calcification regulatory hormones, sex steroids, oxidative stress and chronic inflammation (McFarlane et al., 2004; Sennerby et al., 2009). Secondly, this may be partly due to the greater propensity for falls in those with cardiac diseases, because balance disturbances might occur due to sudden myocardial ischemia and the adverse effect of multiple medications (Hartikainen, Lo¨nnroos, & Louhivuori, 2007). Further studies are required to elucidate the mechanism whereby cardiac disease influenced the SHF. Undoubtedly, effective strategies for Begg's funnel plot with pseudo 95% confidence limits
2
log[var2]
1
0
-1 0
.2
.4
.6
.8
s.e. of: log[var2] Fig. 3. Begg’s funnel plot for publication bias (with 95% pseudoconfidence limits) of the case–control studies that investigated female and risk factor for the SHF (P = 0.767).
management and secondary prevention of hip fracture among patients with initial hip fracture must be developed, for instance through implementing routine home visits for medications by physiotherapists to help prevent falls. It is notable that, the majority of above-motioned risk factors (gender, low vision, dementia, cardiac diseases and respiration diseases) could not be modified, however, these factors should be kept in mind to prepare prevention strategy or reminder with the aim to reduce the risk of SHF for patients with such factors. The osteoporosis as a significant risk factor was potentially amenable. As reported by Bischoff et al. (Bischoff-Ferrari et al., 2009), a vitamin D high dose of 700–800 IU/d could significantly reduce the RR of the hip fracture. Likewise, the compliant use of biphosphonates have been identified to reduce 60% of the occurrence of SHF (4.2% vs. 10.9%) (Lee et al., 2013). Lower bone mineral density have been undoubtedly regarded as the most important factors associated with SHF, although relevant studies were not pooled and calculated in present study due to the inconsistent quantitative criteria (mean or count) and measurement in different sites (calcaneus, total hip or vertebra) for each study. Likewise, American Society of Anesthesiologists (ASA) physical status score were not pooled for the non-uniform quantitative criteria. Because we could not obtain the original raw data, it was impossible to perform a meta-analysis of BMD or ASA. In addition, there were other risk factors that could not be combined because studies were reported in fewer studies. For example, hematologic disease (Lee et al., 2013), renal disease (Lee et al., 2013) and smoking (Souder et al., 2012) were reported in only a single study. This meta-analysis has some limitations. Firstly, a weakness exists in the analyses that not all the ORs on the potential risk factors for the meta-analysis are adjusted because a lot of reports could only provide the univariate rather than multivariate statistics; likewise, some studies might choose not to report the insignificant results or results of no interest, potentially resulting in a considerable amounts of missing data. Hence, our overall effect estimates may be an over-estimate but the multivariate subgroup analysis was performed and the corresponding pooled results are reliable. Secondly, all the included studies were case–control studies with unavoidable recall and interviewer biases, which might affect the associations between the risk and SHF. Thirdly, the measurements of various risk factors differed much from each other, either by standard grading forms or by patient-reported questionnaires. Similarly, follow-up periods ranged widely from several months to more than 10 years. Therefore, a significant heterogeneity was unavoidable in this review. Finally, all the subjects in 22 papers were Asian, European and US; therefore, our conclusions might not extrapolate to African and Australasian populations.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012
G Model
AGG-2967; No. of Pages 6 6
Y. Zhu et al. / Archives of Gerontology and Geriatrics xxx (2014) xxx–xxx
Despite these limitations, some advantages should be mentioned. First and the most important, cardiac disease as a risk factor was identified, which might contribute to complementing prevention strategies for patients involved. Secondly, a comprehensive search strategy based on computer-assisted and manual searching avoided any eligible study to omission. Finally, no publications bias was found, suggesting a robust result. In summary, the present meta-analysis suggests that females, living in institutions, osteoporosis (SI 1–3) low vision, dementia, Parkinson, respiration diseases and cardiac diseases were significant risk factors for SHF. Furthermore, cardiac disease as the significantly increased risk for SHF was firstly reported and related prevention strategies should be formulated. Conflict of interest All the authors declare that they have no conflict of interest with any organization that sponsored the research. Acknowledgements We are grateful to S.L. and Z.W. of the Department of Orthopedics, and to X.C. and G.W. of the Department of Statistics and Applications for their kind assistance. References Angthong, C., Suntharapa, T., & Harnroongroj, T. (2004). Major risk factors for the second contralateral hip fracture in the elderly. Acta Orthopaedica et Traumatologica Turcica, 43, 193–198. Beaupre, L. A., Cinats, J. G., Jones, C. A., Angela, V. S., Johnston, D. W., Ambikaipan, S., et al. (2007). Does functional recovery in elderly hip fracture patients differ between patients admitted from long-term care and the community? Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 62, 1127–1133. Berry, S. D., Samelson, E. J., Hannan, M. T., Robert, R. M., Lu, M., Adrienne, C., et al. (2007). Second hip fracture in older men and women: The Framingham Study. Archives of Internal Medicine, 167, 1971–1976. Bischoff-Ferrari, H. A., Willett, W. C., Wong, J. B., Stuck, A. E., Staehelin, H. B., Orav, E. J., et al. (2009). Prevention of nonvertebral fractures with oral vitamin D and dose dependency: A meta-analysis of randomized controlled trials. Archives of Internal Medicine, 169, 551–561. Chang, J.-D., Yoo, J.-H., Reddy, P., Lee, S.-S., Hwang, J.-H., & Kim, T.-Y. (2013). Risk factors for contra-lateral hip fracture in elderly patients with previous hip fracture. Injury, 44, 1930–1933. Chapurlat, R., Bauer, D., Nevitt, M., Stone, K., & Cummings, S. R. (2003). Incidence and risk factors for a second hip fracture in elderly women. The Study of Osteoporotic Fractures. Osteoporosis International, 14, 130–136. Chiu, K., Pun, W., Luk, K., & Chow, S. (1992). Sequential fractures of both hips in elderly patients – A prospective study. Journal of Trauma and Acute Care Surgery, 32, 584– 587. Dharmarajan, T., Tankala, H., Patel, B., Sipalay, M., & Norkus, E. P. (2001). Outcome in ambulatory status immediately following hip fracture surgery in the acute setting: A comparison of nursing home residents and community older adults. Journal of the American Medical Directors Association, 2, 115–119. Dinah, A. (2002). Sequential hip fractures in elderly patients. Injury, 33, 393–394. Dretakis, K., Dretakis, E., Papakitsou, E., Psarakis, & Steriopoulos, K. (1998). Possible predisposing factors for the second hip fracture. Calcified Tissue International, 62, 366–369. Egan, M., Jaglal, S., Byrne, K., Jennie, W., & Paul, S. (2008). Factors associated with a second hip fracture: A systematic review. Clinical Rehabilitation, 22, 272–282. Finsen, V., & Benum, P. (1986). The second hip fracture an epidemiologic study. Acta Orthopaedica, 57, 431–433. Fredman, L., Magaziner, J., Hawkes, W., JR, H., Linda, P. F., Judith, K., et al. (2005). Female hip fracture patients had poorer performance-based functioning than communitydwelling peers over 2-year follow-up period. Journal of Clinical Epidemiology, 58, 1289–1298. Fukushima, T., Sudo, A., & Uchida, A. (2006). Bilateral hip fractures. Journal of Orthopaedic Science, 11, 435–438.
Galois, L., Dausse, Y., Mainard, D., & Delagoutte, J. P. (2001). Bilateral noncontemporary hip fractures. 31 cases reported. European Journal of Orthopaedic Surgery & Traumatology, 11, 219–223. Gaumetou, E., Zilber, S., & Hernigou, P. (2011). Non-simultaneous bilateral hip fracture: Epidemiologic study of 241 hip fractures. Orthopaedics & Traumatology: Surgery & Research, 97, 22–27. Hagino, H., Sawaguchi, T., Endo, N., Yasuyo, I., Tetsuo, N., & Yoshinobu, W. (2012). The risk of a second hip fracture in patients after their first hip fracture. Calcified Tissue International, 90, 14–21. Hartikainen, S., Lo¨nnroos, E., & Louhivuori, K. (2007). Medication as a risk factor for falls: Critical systematic review. Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 62, 1172–1181. Kim, S.-M., Moon, Y.-W., Lim, S.-J., Yoonb, B.-K., Minc, Y.-K., Lee, D.-Y., et al. (2012). Prediction of survival, second fracture, and functional recovery following the first hip fracture surgery in elderly patients. Bone, 50, 1343–1350. Lau, J., Ioannidis, J. P., & Schmid, C. H. (1997). Quantitative synthesis in systematic reviews. Annals of Internal Medicine, 127, 820–826. Lawrence, T., Wenn, R., & Boulton, C. (2010). Age-specific incidence of first and second fractures of the hip. Journal of Bone & Joint Surgery British, 92, 258–261. Lee, S. H., Lee, T. J., & Cho, K. J. (2012). Subsequent hip fracture in osteoporotic hip fracture patients. Yonsei Medical Journal, 53, 1005–1009. Lee, Y.-K., Ha, Y.-C., Yoon, B.-H., & Koo, K.-H. (2013). Incidence of second hip fracture and compliant use of bisphosphonate. Osteoporosis International, 24, 2099–2104. Lo¨nnroos, E., Kautiainen, H., Karppi, P., Hartikainen, & Kiviranta, R. (2007). Sulkava Incidence of second hip fractures. A population-based study. Osteoporosis International, 18, 1279–1285. McFarlane, S. I., Muniyappa, R., Shin, J. J., John, J. S., Gul, B., James, R. S., et al. (2004). Osteoporosis and cardiovascular disease. Endocrine, 23, 1–10. Mitani, S., Shimizu, M., Abo, M., Hiroshi, H., & Youichi, K. (2010). Risk factors for second hip fractures among elderly patients. Journal of Orthopaedic Science, 15, 192–197. Omsland, T. K., Holvik, K., Meyer, H. E., Jacqueline, R. C., Nina, E., Tell, G. S., et al. (2012). Hip fractures in Norway 1999–2008: Time trends in total incidence and second hip fracture rates. A NOREPOS study. European Journal of Epidemiology, 27, 807–814. Pearse, E., Redfern, D., Sinha, M., & A, J. E. (2004). Outcome following a second hip fracture. Injury, 34, 518–521. Penrod, J. D., Litke, A., Hawkes, W. G., Kenneth, J., Koval, M. D., John, T., et al. (2007). Heterogeneity in hip fracture patients: Age, functional status and comorbidity. Journal of the American Geriatrics Society, 55, 407–413. Rodaro, E., Pasqualini, M., Iona, L., & Di Benedetto, (2004). Functional recovery following a second hip fracture. Europa Medicophysica, 40, 179–183. Ryg, J., Rejnmark, L., Overgaard, S., Kim, B., & Peter, V. (2009). Hip fracture patients at risk of second hip fracture: A nationwide population based cohort study of 169,145 cases during 1977–2001. Journal of Bone and Mineral Research, 24, 1299–1307. Sawalha, S., & Parker, M. (2012). Characteristics and outcome in patients sustaining a second contralateral fracture of the hip. Journal of Bone & Joint Surgery British, 94, 102–106. Saxena, P., & Shankar, J. (2000). Contralateral hip fractures – Can predisposing factors be determined? Injury, 31, 421–424. Schroder, H. M., Petersen, K. K., & Erlandsen, M. (1993). Occurrence and incidence of the second hip fracture. Clinical Orthopaedics and Related Research, 289, 166–169. Sennerby, U., Melhus, H., Gedeborg, R., Liisa, B., Hans, G., Anders, A., et al. (2009). Cardiovascular diseases and risk of hip fracture. Journal of the American Medical Association, 302, 1666–1673. Souder, C. D., Brennan, M. L., Brennan, K. L., Juhee, S., Johnathan, W., & Christopher, C. (2012). The rate of contralateral proximal femoral fracture following closed reduction and percutaneous pinning compared with arthroplasty for the treatment of femoral neck fractures. Journal of Bone & Joint Surgery, 94, 418–425. Stang, A. (2004). Critical evaluation of the Newcastle–Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. European Journal of Epidemiology, 25, 603–605. Van Balen, R., Steyerberg, E., Polder, J., Ribbers, T., Habbema, J., & Cools, H. (2001). Hip fracture in elderly patients: Outcomes for function, quality of life, and type of residence. Clinical Orthopaedics and Related Research, 390, 232–243. Vestergaard, P., Rejnmark, L., & Mosekilde, L. (2007). Has mortality after a hip fracture increased? Journal of the American Geriatrics Society, 55, 1720–1726. Vochteloo, A. J., van der Burg, B. L. B., Ro¨ling, M. A., Ro¨ling, M. A., van Leeuwen, D. H., van den Berg, P., et al. (2012). Contralateral hip fractures and other osteoporosisrelated fractures in hip fracture patients: Incidence and risk factors. An observational cohort study of 1229 patients. Archives of Orthopaedic and Trauma Surgery, 132, 1191–1197. Wei, J., Yang, T.-B., Luo, W., Qin, J.-B., & Kong, F.-J. (2013). Complications following dorsal versus volar plate fixation of distal radius fracture: A meta-analysis. Journal of International Medical Research, 41, 265–275. Yamanashi, A., Yamazaki, K., Kanamori, M., Kuniyoshi, M., Shozo, O., Youichi, K., et al. (2004). Assessment of risk factors for second hip fractures in Japanese elderly. Osteoporosis International, 16, 1239–1246. Zhang, Y. (2012). Clinical epidemiology of orthopedic trauma (1st ed.). New York and Stuggart: Thieme.
Please cite this article in press as: Zhu, Y., et al., Meta-analysis of risk factors for the second hip fracture (SHF) in elderly patients. Arch. Gerontol. Geriatr. (2014), http://dx.doi.org/10.1016/j.archger.2014.02.012