Advertisement

Associations between presence of diabetes, mortality and fracture type in individuals with a hip fracture

Open AccessPublished:September 16, 2022DOI:https://doi.org/10.1016/j.diabres.2022.110084

      Highlights

      • Patients with diabetes display a greatly increased mortality risk after hip fracture.
      • This increased mortality is not explained by fracture type or comorbidities.
      • Hip fractures may be considered as yet another diabetes-associated complication.
      • In-and-out of hospital efforts in diabetes care are needed to reduce excess mortality.

      Abstract

      Aims

      An overlooked aspect of diabetes is an increased risk of hip fractures, with associated mortality. We investigated whether fracture type and/or burden of comorbidities explains the increased risk of mortality in diabetes after hip fracture.

      Methods

      For this cohort study, we used a de-identified data set of hip fracture patients registered in a quality-of-care registry (2017/2018) included in Maastricht University Medical Centre.

      Results

      Among 594 hip fracture patients, 90 (15.2 %) had diabetes. Median (IQR) age was 82 (71–87) years and 63.8 % were women. Compared to patients without, patients with diabetes had higher median Charlson Comorbidity Index [1 (0–2) vs 0 (0–2), P < 0.001)] and were more likely to sustain intertrochanteric/subtrochanteric fractures [54.4 vs 38.7 %, P = 0.02]. Over a median follow-up of 2.7 (1.6–3.3) years, crude mortality rate was 30.8 % in patients without and 50.0 % in patients with diabetes. This association remained unaltered after adjustment for age, sex, BMI, fracture type or burden of co-morbidities.

      Conclusion

      Individuals with diabetes display a greatly increased absolute mortality risk after hip fracture. This association was not attenuated after adjustment for fracture type or non-diabetes associated co-morbidity. These findings have important implications for diabetes care in hip fracture patients, and underline the importance of fracture prevention.

      Keywords

      1. Introduction

      Although the decreased life expectancy attributed to diabetes attenuates with older age [
      • Rawshani A.
      • Rawshani A.
      • Franzén S.
      • Sattar N.
      • Eliasson B.
      • Svensson A.-M.
      • et al.
      Risk factors, mortality, and cardiovascular outcomes in patients with Type 2 diabetes.
      ], an overlooked aspect of diabetes care in older age is the increased risk of hip fractures, and hip-fracture associated mortality [
      • Lipscombe L.L.
      • Jamal S.A.
      • Booth G.L.
      • Hawker G.A.
      The risk of hip fractures in older individuals with diabetes: a population-based study.
      ,
      • Shen Q.
      • Ma Y.
      Impact of diabetes mellitus on risk of major complications after hip fracture: a systematic review and meta-analysis.
      ]. This increased risk is often attributed to falls associated with clinical evidence of end-organ damage, such as retinopathy and neuropathy, and to a higher propensity to fractures due to diabetes-associated bone fragility and sarcopenia [
      • Formiga F.
      • Chivite D.
      • Ruiz D.
      • Navarro M.
      • Perez Castejon J.M.
      • Duaso E.
      • et al.
      Clinical evidence of diabetes mellitus end-organ damage as risk factor for falls complicated by hip fracture: a multi-center study of 1225 patients.
      ,
      • Hofbauer L.C.
      • Busse B.
      • Eastell R.
      • Ferrari S.
      • Frost M.
      • Müller R.
      • et al.
      Bone fragility in diabetes: novel concepts and clinical implications.
      ]. Therefore, it is conceivable that individuals with diabetes suffer from different subtypes of hip fractures than the general aging population. This is important, as subtrochanteric fractures are associated with a higher proportion of orthopedic device failure, while intertrochanteric fractures are associated with more blood loss [
      • LeBlanc K.E.
      • Muncie Jr, H.L.
      • LeBlanc L.L.
      Hip fracture: diagnosis, treatment, and secondary prevention.
      ]. In line with the above, Nolan et al. [
      • Horner N.S.
      • Samuelsson K.
      • Solyom J.
      • Bjørgul K.
      • Ayeni O.R.
      • Östman B.
      Implant-related complications and mortality after use of short or long gamma nail for intertrochanteric and subtrochanteric fractures: a prospective study with minimum 13-Year follow-up.
      ] reported 30 % mortality in individuals with inter- and subtrochanteric fractures after 1 year. In addition, 10 % required additional surgery in a case series evaluating the commonly used Gamma nail as the surgical technique of choice [
      • Horner N.S.
      • Samuelsson K.
      • Solyom J.
      • Bjørgul K.
      • Ayeni O.R.
      • Östman B.
      Implant-related complications and mortality after use of short or long gamma nail for intertrochanteric and subtrochanteric fractures: a prospective study with minimum 13-Year follow-up.
      ]. To our knowledge, it has not yet been investigated whether hip fracture type explains the mortality gap of individuals with diabetes and a hip fracture.
      Furthermore, individuals with diabetes also have a higher risk of post-surgical infectious, renal and cardiac complications, albeit it is not known whether these complications arise from perioperative hyperglycaemia or from non-diabetes associated comorbidities [
      • Frisch A.
      • Chandra P.
      • Smiley D.
      • Peng L.
      • Rizzo M.
      • Gatcliffe C.
      • et al.
      Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac surgery.
      ]. As such, it is unknown whether the surgical (fracture type) or the medical (burden of co-morbidity) aspects of diabetes-associated hip fractures drive excess risk of mortality. This is an important issue to address in order to improve patient outcomes.
      We investigated these key aspects, in order to explain the excess risk of mortality after hip fractures in individuals with diabetes, in a quality-of-care registry of a high-volume academic trauma centre in the Netherlands. Firstly, we assessed whether individuals with diabetes were more often admitted with high-risk fractures (i.e. intertrochanteric or subtrochanteric fractures vs femoral neck fractures) and whether this explained the increased associated mortality risk in diabetes. Secondly, we investigated whether non-diabetes associated comorbidity, as expressed by a diabetes-modified Charlson Comorbidity Index (CCI), explained excess mortality.

      2. Methods

      2.1 Data collection

      For this cohort study, we used a de-identified data set of hip fracture patients registered in a quality-of-care registry derived from hospital records, based on the Hospital Standardized Mortality Ratio (HSMR). Included were all hip fracture patients who presented to the Maastricht University Medical Centre, the Netherlands, between January 1, 2017 and December 31, 2018.
      The data set contained demographic, clinical and follow-up data, including age, sex, fracture type, comorbidity, body mass index (BMI), Katz activities of daily living (ADL) index, recorded surgical (pressure ulcer, hemorrhage, wound site infection, dislocation, implant failure), and medical complications (delirium, infections, myocardial infarction, acute kidney injury (AKI) and heart failure), hospital length of stay (LOS), and discharge destination. Additional data on bisphosphonate use and antidiabetic therapy were obtained using hospital medical records. Data were collected by trained personal tasked with data collection for care evaluation purposes. The institutional medical ethical review board approved the study with a waiver of consent (METC 2020–2285).

      2.2 Modified Charlson comorbidity index

      The CCI is commonly used in clinical practice, relatively simple and well-validated method to estimate global burden of comorbidity by assigning points to and adding up all diagnosed comorbidities in an individual [
      • D'Hoore W.
      • Bouckaert A.
      • Tilquin C.
      Practical considerations on the use of the Charlson comorbidity index with administrative data bases.
      ]. However, since diabetes is included in this index, we subtracted one point from the CCI in individuals with diabetes without complications and two points in individuals with diabetic complications (retinopathy, diabetic kidney disease and neuropathy). This in order to avoid over-adjustment for diabetes in the Cox regression models.

      2.3 Katz ADL index

      The Katz ADL index was used to assess limitations in activities of daily living (ADL) [
      • Katz S.
      • Ford A.B.
      • Moskowitz R.W.
      • Jackson B.A.
      • Jaffe M.W.
      Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function.
      ]. Six ADL items (bathing, dressing, toileting, transferring, maintaining continence and feeding) were scored on functional dependence or independence, with total scores ranging from 0 (completely dependent) to 6 (completely independent).

      2.4 Assessment of diabetes status

      Presence of diabetes was derived from medical records. If no specific mention of diabetes type was made, we assumed type 2 diabetes mellitus, based on the low prevalence of type 1 diabetes in the older population. Only 1 individual was recorded to have type 1 diabetes mellitus.

      2.5 Outcomes after hip surgery

      Our primary outcome was all-cause mortality after hip fracture diagnosis. Secondary outcomes included length of hospital stay and discharge destination. Follow-up for all-cause mortality was conducted to 31 December 2020.

      2.6 Statistical analyses

      We compared patient characteristics stratified according to the presence of diabetes, with data presented as mean ± standard deviation (SD), median and interquartile range (IQR), or percentages as appropriate. Differences were compared with a student’s-Test, Chi-square test. Mann-Whitney U test, and analysis of variance (ANOVA)-test, as appropriate. We plotted a Kaplan-Meier curve, and used a log-rank test to compare the difference between patients with and without diabetes.
      Cox proportional hazard regression analysis was performed to compare mortality between hip fracture patients with and without diabetes, with adjustments for age, sex, BMI, fracture type, time to surgery (hours), Katz ADL index and adjusted CCI. Inspection of log(−log(survival)) curves showed parallel lines, satisfying the proportional hazards assumption. In the multivariable analyses we excluded individuals with missing values listwise (n = 18), and we performed additional analyses using multiple imputation of missing covariates in the full data-set. Furthermore, we investigated whether the association between the presence of diabetes and mortality differed between men and women by adding an interaction term to the adjusted model (sex*diabetes).

      3. Results

      3.1 Demographics

      Table 1 shows the baseline characteristics of hip fracture patients stratified according to the presence of diabetes. Among the 594 hip fracture patients, 90 (15.2 %) had diabetes. Patients with diabetes had significantly less often a femoral neck fracture, and more often a intertrochanteric- or a subtrochanteric fracture (Table 1). In addition, individuals with diabetes had a higher modified CCI, and tended to have a lower Katz ADL index. Furthermore, individuals with diabetes tended to be older and have a higher BMI. We observed no difference in time between admission and surgery, length of hospital stay and discharge to the own living environment between patients with- and without diabetes.
      Table 1Demographic and clinical characteristics.
      Diabetes
      1 individual with T1D.
      (n = 90)
      No Diabetes (n = 504)Total cohort (n = 594)P-value
      Age, years, median (IQR)82 (76–87)82 (71–87)82 (71–87)0.289
      Women (%)63.363.963.80.906
      BMI (kg/m2), mean (SD)25.0 (4.3)24.1 (4.4)24.2 (4.4)0.082
      Type of fracture (%)0.046
      Femoral neck fracture48.960.758.9
      Intertrochanteric fracture46.735.937.5
      Subtrochanteric fracture4.43.43.6
      Adjusted CCI, median (IQR)1 (0–2)0 (0–2)0 (0–2)<0.001
      Katz ADL-index, median (IQR)3 (2–6)5 (2–6)4 (2–6)0.098
      Length of stay days, median (IQR)9 (5.8–13)8 (5–13)8 (5–13)0.528
      Time to surgery hours, median (IQR)19.7 (15.3–24.8)18.4 (12.9–25.7)18.6 (13.3–25.4)0.226
      Discharge to own living environment (%)25.633.932.70.119
      Bisphosphonate use (%)9.013.312.60.303
      Glucose lowering therapy (%)N/A
      Diet26.7
      Metformin monotherapy30.0
      Sulfonylurea monotherapy4.4
      DPP-4i monotherapy1.1
      Insulin monotherapy22.2
      Double oral antidiabetics
      1 patient with metformin + sulfonylurea and 1 patient with metformin + DPP-4i,
      2.2
      Oral antidiabetics + insulin
      9 patients with metformin + insulin, 2 patients with sulfonylurea + insulin, 1 patient with metformin + sulfonylurea + insulin.
      13.3
      Data were presented as mean (SD), median (IQR) or percentage and T-test, Chi-square test, Mann-Whitney U test, and analysis of variance (ANOVA)-test were used as appropriate.
      *Adjusted Charlson Comorbidity Index: minus 1 for diabetes without and minus 2 for diabetes with complications.
      ¥ 1 individual with T1D.
      α 1 patient with metformin + sulfonylurea and 1 patient with metformin + DPP-4i,
      β 9 patients with metformin + insulin, 2 patients with sulfonylurea + insulin, 1 patient with metformin + sulfonylurea + insulin.
      Table 2 shows the (post-)surgical complications. Overall, we observed no statistically significant differences between surgical complications and individuals with- and without diabetes. Individuals with diabetes tended to a higher risk of recorded medical complications, although only the difference in incidence of AKI was significant when analyzed separately, while the trends towards higher incidences of delirium, myocardial infarction, heart failure in individuals with diabetes were not significant.
      Table 2Occurrence of in-hospital adverse events after surgery.
      Diabetes¥

      (n = 90)
      No Diabetes

      (n = 504)
      Total cohort (n = 594)P-value
      Surgical complications (%)6.27.66.90.152
      Medical complications (%)51.141.743.10.096
      Delirium (%)36.727.028.50.061
      Infections (%)17.019.419.10.566
      Myocardial infarction (%)5.62.63.00.129
      Acute Kidney Injury (%)17.810.311.40.041
      Heart Failure (%)5.66.76.20.225
      Data are presented as percentages and differences were tested with a Chi-square test.

      3.2 Mortality

      Over a median (IQR) follow-up of 2.7 (1.6 – 3.3) years, 200 patients (33.7 %) died. The crude mortality rate was 30.8 % in patients without and 50.0 % in patients with diabetes, respectively (P < 0.001, Fig. 1). Following adjustment for age, sex, modified CCI and fracture type, patients with diabetes still had a higher mortality risk (HR 1.55, 95 %CI: 1.11–2.17). Age (per year, HR 1.07, 95 %CI: 1.05–1.08), male sex (HR 1.85, 95 %CI: 1.37–2.48), and CCI (per point, HR 1.30, 95 %CI: 1.21–1.39) all remained statistically significantly associated with mortality, while fracture type itself did not (Femoral neck fracture is reference, HR 0.99, 95 %CI: 0.77–1.27). Upon further adjustment for time to surgery, Katz ADL index and BMI, we observed, if anything, a higher mortality risk in individuals with diabetes (HR 1.78, 95 %CI: 1.23–2.58), while both a higher BMI (per kg/m2, HR 0.90, 95 %CI: 0.86–0.94) and a higher Katz ADL index (per point, HR 0.91, 95 %CI: 0.84–0.98) score were associated with a lower mortality risk.
      Figure thumbnail gr1
      Fig. 1Kaplan-Meier survival curve stratified according to the presence of diabetes. Legend: Log-rank test compared the difference between patients with and without diabetes (P < 0.001).

      3.3 Additional analyses

      When we excluded the individual with recorded type 1 diabetes mellitus, or when we imputed rather than excluded missing covariates, the analyses remained unaltered (data not shown). Furthermore, the associations did not differ between men and women (Pinteraction = 0.97).

      4. Discussion

      This study underlines several important issues. First, we found that in hip fracture patients, diabetes remained associated with an at least 1.5 times increased risk of all-cause mortality, after adjustment for age, sex, comorbidity, BMI, Katz ADL index and fracture type. We also found that patients with diabetes were more likely to sustain ‘worse’ fractures, but this did not drive mortality. We also observed a trend towards more medical in-hospital complications. Therefore, excess mortality after hip fractures is not related to fracture type and seems to be directly diabetes-associated.
      Despite the fact that this is frequently hypothesized, we now show that although patients with diabetes indeed sustain ‘worse’ fractures, this does not independently contribute to mortality, while diabetes itself remains an independent risk factor associated with excess mortality itself, with a similar effect size as total other comorbidities combined.
      A major remaining issue is whether a tight or more lapsed glycaemic control regimen would reduce the risk of adverse outcomes after hip surgery. Because we used in-hospital data collected per usual care, we did not have access to glycaemic control at the time of hospital admission. Although this is a limitation of our study, we believe that tight glycaemic control is unlikely to improve mortality in individuals with diabetes and a hip fracture, based on previous studies on tight glycaemic control. This is especially the likely since glycaemic control tends to improve with older age [
      • Nanayakkara N.
      • Ranasinha S.
      • Gadowski A.M.
      • Davis W.A.
      • Flack J.R.
      • Wischer N.
      • et al.
      Age-related differences in glycaemic control, cardiovascular disease risk factors and treatment in patients with type 2 diabetes: a cross-sectional study from the Australian National Diabetes Audit.
      ], despite the accumulation of diabetic complications, while the adverse consequences of hypoglycaemia tend to increase with older age. The latter is further substantiated by the relatively low BMI of individuals with type 2 diabetes mellitus in our study, which is typical for diabetes at older age. As such, we speculate that our population on average likely had an expected good glycaemic control, which is further supported by the fact that in our study more than 60 % of the patients with diabetes were treated with oral monotherapy (metformin, sulfonylurea or DPP4-inhibitor) or had no glucose lowering therapy at all (diet). Also, in line with thisin the setting of arthroplasty, higher blood glucose levels are associated with more surgical infections [
      • Kheir M.M.
      • Tan T.L.
      • Kheir M.
      • Maltenfort M.G.
      • Chen A.F.
      Postoperative blood glucose levels predict infection after total joint arthroplasty.
      ], but to our knowledge, this has not been addressed in clinical trials, and even less is known about the effect of glycaemic control on mortality, delirium, as well as cardiovascular- and renal complications, especially in our older and very vulnerable population. Moreover, we believe that in the acute rather than elective setting of hip fractures, the risk to benefit ratio of tight glycaemic control is difficult to predict without a dedicated clinical trial, given the inconsistent results regarding the effect of tight glycaemic control on major outcomes in several surgical and critically ill patient populations, finding either protection, no effect or in fact harm by tight glucose control [
      • Kavanagh B.P.
      Glucose in the ICU–evidence, guidelines, and outcomes.
      ]. We are therefore skeptical to speculate whether glucose lowering interventions would abrogate the large mortality gap observed in the current study.
      More evidence is available on the efficacy of osteoporosis treatment in individuals with diabetes [
      • Mohsin S.
      • Baniyas M.M.
      • AlDarmaki R.S.
      • Tekes K.
      • Kalász H.
      • Adeghate E.A.
      An update on therapies for the treatment of diabetes-induced osteoporosis.
      ], with emerging awareness for a preference of bone-friendly glucose- and glycation lowering medications such as metformin. In recognition of the increased fracture risk the American Diabetes Association recommends patients with diabetes to discuss the need for bone densitometry, especially when additional risk factors for osteoporosis are present. An addition to reducing hip fracture-related mortality in diabetes, a greater emphasis on fracture prevention and screening- and treatment of osteoporosis in older individuals with diabetes may be needed. Furthermore, several factors have been identified that are associated with a great risk of falls in diabetes such as the number of medications, higher levels of pain, poorer self-perceived health, lower physical activity and grip strength, more limitations in ADLs, lower-extremity physical performance and cognitive impairment [
      • Pijpers E.
      • Ferreira I.
      • de Jongh R.T.
      • Deeg D.J.
      • Lips P.
      • Stehouwer C.D.A.
      • et al.
      Older individuals with diabetes have an increased risk of recurrent falls: analysis of potential mediating factors: the Longitudinal Ageing Study Amsterdam.
      ]. We propose that our current findings underline the need for interventional studies in at-risk diabetic populations evaluating the efficacy for a ‘screen and intervene’ approach on these key demographic characteristics.
      The overwhelming majority of individuals in our study had type 2 diabetes mellitus. We were therefore unable to assess whether these findings differ for type 1 diabetes mellitus. This is an important topic, since the relationship between hip fractures and diabetes seems to be more pronounced in type 1 diabetes [
      • Fan Y.
      • Wei F.
      • Lang Y.
      • Liu Y.
      Diabetes mellitus and risk of hip fractures: a meta-analysis.
      ,
      • Hamilton E.J.
      • Davis W.A.
      • Bruce D.G.
      • Davis T.M.E.
      Risk and associates of incident hip fracture in type 1 diabetes: the Fremantle Diabetes Study.
      ]. Furthermore, our study was not designed to address whether specific comorbidities mediate the associations between diabetes and mortality after a hip fracture.
      Interestingly, we confirmed the inverse association between BMI and mortality reported previously, which is also known as ‘the obesity paradox’ [
      • Bosello O.
      • Vanzo A.
      Obesity paradox and aging.
      ]. This suggests that for older and more vulnerable individuals, such as our population, a higher BMI may be more optimal [
      • Costanzo P.
      • Cleland J.G.F.
      • Pellicori P.
      • Clark A.L.
      • Hepburn D.
      • Kilpatrick E.S.
      • et al.
      The obesity paradox in type 2 diabetes mellitus: relationship of body mass index to prognosis: a cohort study.
      ,
      • Vosoughi A.R.
      • Emami M.J.
      • Pourabbas B.
      • Mahdaviazad H.
      Factors increasing mortality of the elderly following hip fracture surgery: role of body mass index, age, and smoking.
      ] and although the ‘obesity paradox’ is still debated, a protective nutritional status in overweight individuals probably contributes to this paradox, especially in our specific and vulnerable population at high risk for undernutrition.
      Although this study may be limited by its single-centre and observational design and inability to assess cause-specific mortality, based on these findings, future studies should urgently focus on increased risk of diabetes-associated complications such as hypoglycaemia, sarcopenia, infections and renal- and cardiac complications to improve outcomes in this especially vulnerable and growing patient population. Indeed, in the older population with diabetes, hip fractures may be considered as yet another diabetes-associated complication, and both in- and out- of hospital efforts in diabetes care may be needed to reduce the gap in excess mortality in individuals with diabetes and a hip fracture.

      Conflicts of Interest

      N.M.J.H. has received honorarium from Boehringer Ingelheim and Novo Nordisk. The other authors have nothing to declare.

      Declaration of Competing Interest

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Acknowledgements

      NMJH is supported by a Senior Clinical Dekker grant by the Dutch Heart Foundation (grant number 2021T055 ) and a DFN- DON grant 2020 (Grant number 2020.10.002).

      References

        • Rawshani A.
        • Rawshani A.
        • Franzén S.
        • Sattar N.
        • Eliasson B.
        • Svensson A.-M.
        • et al.
        Risk factors, mortality, and cardiovascular outcomes in patients with Type 2 diabetes.
        N Engl J Med. 2018; 379: 633-644
        • Lipscombe L.L.
        • Jamal S.A.
        • Booth G.L.
        • Hawker G.A.
        The risk of hip fractures in older individuals with diabetes: a population-based study.
        Diabetes Care. 2007; 30: 835-841
        • Shen Q.
        • Ma Y.
        Impact of diabetes mellitus on risk of major complications after hip fracture: a systematic review and meta-analysis.
        Diabetol Metab Syndr. 2022; 14: 51
        • Formiga F.
        • Chivite D.
        • Ruiz D.
        • Navarro M.
        • Perez Castejon J.M.
        • Duaso E.
        • et al.
        Clinical evidence of diabetes mellitus end-organ damage as risk factor for falls complicated by hip fracture: a multi-center study of 1225 patients.
        Diabetes Res Clin Pract. 2015; 109: 233-237
        • Hofbauer L.C.
        • Busse B.
        • Eastell R.
        • Ferrari S.
        • Frost M.
        • Müller R.
        • et al.
        Bone fragility in diabetes: novel concepts and clinical implications.
        Lancet Diabetes Endocrinol. 2022; 10: 207-220
        • LeBlanc K.E.
        • Muncie Jr, H.L.
        • LeBlanc L.L.
        Hip fracture: diagnosis, treatment, and secondary prevention.
        Am Fam Physician. 2014; 89: 945-951
        • Horner N.S.
        • Samuelsson K.
        • Solyom J.
        • Bjørgul K.
        • Ayeni O.R.
        • Östman B.
        Implant-related complications and mortality after use of short or long gamma nail for intertrochanteric and subtrochanteric fractures: a prospective study with minimum 13-Year follow-up.
        JB JS Open Access. 2017; 2: e0026https://doi.org/10.2106/JBJS.OA.17.00026
        • Frisch A.
        • Chandra P.
        • Smiley D.
        • Peng L.
        • Rizzo M.
        • Gatcliffe C.
        • et al.
        Prevalence and clinical outcome of hyperglycemia in the perioperative period in noncardiac surgery.
        Diabetes Care. 2010; 33: 1783-1788
        • D'Hoore W.
        • Bouckaert A.
        • Tilquin C.
        Practical considerations on the use of the Charlson comorbidity index with administrative data bases.
        J Clin Epidemiol. 1996; 49: 1429-1433
        • Katz S.
        • Ford A.B.
        • Moskowitz R.W.
        • Jackson B.A.
        • Jaffe M.W.
        Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function.
        JAMA. 1963; 185: 914-919
        • Nanayakkara N.
        • Ranasinha S.
        • Gadowski A.M.
        • Davis W.A.
        • Flack J.R.
        • Wischer N.
        • et al.
        Age-related differences in glycaemic control, cardiovascular disease risk factors and treatment in patients with type 2 diabetes: a cross-sectional study from the Australian National Diabetes Audit.
        BMJ Open. 2018; 8: e020677https://doi.org/10.1136/bmjopen-2017-020677
        • Kheir M.M.
        • Tan T.L.
        • Kheir M.
        • Maltenfort M.G.
        • Chen A.F.
        Postoperative blood glucose levels predict infection after total joint arthroplasty.
        J Bone Joint Surg Am. 2018; 100: 1423-1431
        • Kavanagh B.P.
        Glucose in the ICU–evidence, guidelines, and outcomes.
        N Engl J Med. 2012; 367: 1259-1260
        • Mohsin S.
        • Baniyas M.M.
        • AlDarmaki R.S.
        • Tekes K.
        • Kalász H.
        • Adeghate E.A.
        An update on therapies for the treatment of diabetes-induced osteoporosis.
        Expert Opin Biol Ther. 2019; 19: 937-948
        • Pijpers E.
        • Ferreira I.
        • de Jongh R.T.
        • Deeg D.J.
        • Lips P.
        • Stehouwer C.D.A.
        • et al.
        Older individuals with diabetes have an increased risk of recurrent falls: analysis of potential mediating factors: the Longitudinal Ageing Study Amsterdam.
        Age Ageing. 2012; 41: 358-365
        • Fan Y.
        • Wei F.
        • Lang Y.
        • Liu Y.
        Diabetes mellitus and risk of hip fractures: a meta-analysis.
        Osteoporos Int. 2016; 27: 219-228
        • Hamilton E.J.
        • Davis W.A.
        • Bruce D.G.
        • Davis T.M.E.
        Risk and associates of incident hip fracture in type 1 diabetes: the Fremantle Diabetes Study.
        Diabetes Res Clin Pract. 2017; 134: 153-160
        • Bosello O.
        • Vanzo A.
        Obesity paradox and aging.
        Eat Weight Disord. 2021; 26: 27-35https://doi.org/10.1007/s40519-019-00815-4
        • Costanzo P.
        • Cleland J.G.F.
        • Pellicori P.
        • Clark A.L.
        • Hepburn D.
        • Kilpatrick E.S.
        • et al.
        The obesity paradox in type 2 diabetes mellitus: relationship of body mass index to prognosis: a cohort study.
        Ann Intern Med. 2015; 162: 610-618
        • Vosoughi A.R.
        • Emami M.J.
        • Pourabbas B.
        • Mahdaviazad H.
        Factors increasing mortality of the elderly following hip fracture surgery: role of body mass index, age, and smoking.
        Musculoskelet Surg. 2017; 101: 25-29