Accelerated decline in quadriceps area and Timed Up and Go test performance are associated with hip fracture risk in older adults with impaired kidney function

Highlights

•

A faster decline in quadriceps area and TUG test performance were significantly associated with increased hip fracture risk.

•

5-year change in quadriceps attenuation and knee extension strength were not associated with fracture risk.

•

TUG test, which can be easily included in routine medical examinations, may help identify high fracture risk individuals.

Abstract

Objective

This study aimed to examine whether an accelerated decline in quadriceps cross-sectional area (CSA), attenuation (a surrogate of quality), and strength, as well as lower limb muscular function, are associated with hip fractures in older adults with impaired kidney function.

Design

Prospective population-based study.

Setting

Community-dwelling old population in Reykjavik, Iceland.

Subjects

A total of 875 older adults (mean baseline age 76 years) from the Age, Gene/Environment Susceptibility (AGES)-Reykjavik Study with impaired kidney function.

Methods

Quadriceps CSA and density were determined using computed tomography (CT), knee extension strength was measured with an isometric dynamometer chair, and muscular function was assessed using the Timed Up and Go (TUG) test. All muscle-related measurements were assessed twice over a mean follow-up of 5.2 years. Data on hip fracture incidence was obtained from medical records during a maximum of 8.4 years of follow-up time.

Results

Fully adjusted cox-proportional hazard regression models showed that a faster decline in quadriceps CSA and TUG test performance were significantly associated with increased hip fracture risk (HR = 1.55, 95% CI = 1.02–2.36, and HR = 1.80, 95% CI = 1.19–2.72, respectively). A faster decrease in quadriceps density and isometric knee extension strength were not associated with fracture risk.

Conclusions

Accelerated decline in CT-derived quadriceps CSA and muscular function, as measured by the TUG test's performance, are predictive of hip fracture risk in older adults with impaired kidney function. TUG test is a simple measure and easily included in routine medical examinations, compared to CT scans, which seems to be useful for identifying a subgroup of individuals with high risk of fracture.

Keywords

Computed tomography
Chronic kidney disease
Estimated glomerular filtration rate
Muscular function
Fracture

Abbreviations

AGES
Age, Gene/Environment Susceptibility
CSA
cross-sectional area
CT
computed tomography
HU
Hounsfield units

1. Introduction

Chronic kidney disease (CKD) adverse effects on skeletal muscle mass and function are well-recognized (Carrero et al., 2016; Moorthi and Avin, 2017), contributing to a high prevalence of adverse clinical events, such as frailty, sarcopenia, protein-energy wastage, and death (MacKinnon et al., 2018). Even in an early stage of development, kidney dysfunction is associated with muscle impairment (Roshanravan et al., 2017) and a higher risk of fracture (Chen et al., 2018). In individuals with CKD, fractures are severe events due to higher periods of hospitalization and mortality rates than individuals without CKD (Tentori et al., 2014). However, studies on fracture risk assessment in early CKD stages are scarce and have focused on bone-related risk factors (Bucur et al., 2015).

Recent studies emphasize the potential role of muscle wasting in fracture risk prediction among older adults (Hars and Trombetti, 2017). Computed tomography (CT) can accurately assess skeletal muscle size (area) and composition because fat and muscle have widely different attenuation characteristics (Goodpaster et al., 2000a). Despite this, studies examining the relationship between quadriceps muscle loss based on CT imaging data and fracture risk are limited. Findings from a prospective study showed that lower creatinine clearance, a marker of kidney function, is associated with calf muscle atrophy (measured with peripheral CT) and more rapid declines in lower extremity strength over time (Roshanravan et al., 2015). However, the prediction of fractures from prospective data on muscle-related parameters in older adults with impaired kidney function remains unknown.

Routine assessment of physical functioning in CKD patients using field tests, such as the Timed Up and Go (TUG) test, is widely performed and encouraged in clinical care (Painter and Marcus, 2013). Although several physical capabilities are associated with TUG performance, muscular strength seems to play a determinant role (Coelho-Junior et al., 2018). However, only one early study assessed muscular function's contribution to fracture risk prediction in hemodialysis patients, but muscular strength testing was limited to grip strength (Jamal et al., 2006). No previous studies have explored the relationship between the decline in TUG performance and fracture risk in impaired kidney function subjects.

A better understanding of the contribution of decline in muscle-related characteristics such as quantity (area), quality (muscle density, reflecting fat infiltration on muscle tissue), strength, and function to hip fracture risk in older adults with impaired kidney function will contribute to improving treatment and prevention strategies of this high-risk group. Thus, this study aimed to examine whether a faster decline in TUG test performance (a surrogate of muscular function), quadriceps muscle size, density, and strength are associated with incident hip fractures in a cohort of community-dwelling older adults with impaired kidney function.

2. Methods

2.1. Study design and population

The present study is based on the Age, Gene/Environment Susceptibility (AGES)—Reykjavik Study, a single-center prospective population study of Icelandic older men and women. Specifically, data come from the baseline examination (2002 to 2006) and one follow-up examination (2007 to 2011), occurring on average, 5.2 years later (range 4.3–7.3 years). Design and recruitment have been described in detail elsewhere (Harris et al., 2007). The study sample includes 875 individuals with impaired kidney function, defined as estimated glomerular filtration rate (eGFR) < 60 ml/min/1.73 m2. Flow diagram for study participants is presented in Supplementary Fig. S1. Written informed consent was obtained from all participants. The study was approved by the Icelandic National Bioethics Committee (VSN: 00-063) and the Institutional Review Board of the Intramural Research Program of the National Institute on Aging.

2.2. Definition of kidney function

Baseline GFR was estimated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation (Levey et al., 2009). Serum creatinine was measured at the Icelandic Heart Association using the Roche-Hitachi P-Module instrument with Roche Creatininase Plus assay (Roche Diagnostics Corporation, Indianapolis, IN). Impaired kidney function was defined as eGFR < 60 ml/min/1.73 m2.

2.3. Hip fracture assessment

Information regarding fracture incidence was obtained from a fracture registry based on medical and radiological records (Siggeirsdottir et al., 2007). Hip fracture was defined according to ICD10 diagnostic codes S72.0, S72.1, and S72.2. We selected the hip fractures occurring between the date of participation in the AGES-Reykjavik follow-up examination (2007 to 2011) and April 10, 2016 (maximum 8.4 years of follow-up time).

2.4. Quadriceps size and density

CT measurements were performed at the mid-thigh using the same four-row detector CT system (Sensation; Siemens Medical Systems, Erlangen, Germany), as previously described (Lang et al., 2008). A single 10 mm thick axial image (120 kVp, 200 to 250 mA) was obtained at the mid-thigh after measuring the maximum length of the femur to find the center of the long axis. Quadriceps muscle size was assessed as the cross-sectional area (CSA, cm2) within the muscle contour, and muscle density as the mean of voxel attenuation in Hounsfield units (HU) within the range of 0 to 100 HU. Thus, we excluded the intermuscular and intramuscular adipose tissue lying interior to the deep fascial plane surrounding the muscle. Lower quadriceps muscle HU (density) indicates greater fat infiltration.

We analyzed data from the same thigh tested in the dynamometer chair. The baseline scan and the follow-up scan of each participant were analyzed together to minimize variability, and the analyses were done by a single observer.

2.5. Isometric knee extension (KE) strength

Quadriceps strength was measured with a dynamometer chair (Good Strength, Mettitur Ltd; Palokka, Finland) at baseline and follow-up as previously described in detail (Goodpaster et al., 2000b). The same testing protocol was used at baseline and follow-up examinations, which involved three trials, and peak torque (Nm) was derived as the product of the maximum value (N) of all possible filter windows and lower leg length (m). We analyzed the trial with the maximal value.

2.6. Timed Up and Go

TUG Test was used as a surrogate measurement of muscular function, using the same standardized protocol at baseline and follow-up examinations (Podsiadlo and Richardson, 1991). A stopwatch was used to measure the time it took the participant to stand from a sitting position, walk 3 m, turn around, walk back to the chair, and sit down. We used the time of the first complete trial in the analysis.

2.7. Covariate measures

Several potential confounding covariates were selected based on their biological relevance to muscle parameters and found to be associated with both incident fractures and main predictors in our sample. These variables collected at AGES baseline included: age, sex, weight, height, percent weight change from age 50, self-reported physical activity level (categorized as moderate/high physically active or occasionally physically active at most) (Marques et al., 2016), cognitive status (categorized as normal or impaired) (Vidarsdottir et al., 2014), self-reported history of fracture, diabetes (defined as self-reported history of diabetes, use of glucose-modifying medications, or fasting blood glucose of ≥7.0 mmol/l), and glucocorticoid use.

2.8. Statistical analysis

Continuous variables are described as median and first quartile (Q1), third quartile (Q3) or percentages for categorical variables were used to summarize subject characteristics. Between group comparison was performed by chi-square tests and Mann-Whitney test as appropriate. To estimate annual percent change (Δ%) in each measure, we divided the inter-visit difference relative to absolute baseline, divided by the number of years between the visits, as follows: [(follow-up value − baseline value) / baseline value ∗ time between measurements] ∗ 100. For descriptive purposes, we conducted within-subject comparisons between baseline and follow-up values using Wilcoxon rank-sum tests. Also, repeated measured ANCOVA was used to remove bias from possible confounding variables.

Time to incident hip fracture was calculated as the interval from the AGES follow-up visit (2007 to 2011) to the earliest date of first incident hip fracture, death, loss to follow-up, or end of fracture follow-up on April 10, 2016. Cox proportional hazards models were used to estimate the risk of incident hip fracture, associated with having a faster rate of loss for each main predictor (defined as being in the tertile of fastest muscle-related loss in our sample compared with the other two tertiles of the annual percent change). Results were expressed as hazard ratios (HRs) with 95% confidence intervals (CIs).

The proportional hazard assumption was tested by evaluating interaction terms with time, using the Schoenfeld residuals, and by examining complementary log-log plots (i.e., log(−log-(survival) versus log(time)). HRs reported here did not violate the proportionality assumption and thus were constant over follow-up time.

We conducted sensitivity analyses to account for the competing risk of death, using the Fine and Gray method to estimate unadjusted and multivariable adjusted sub-distribution relative hazards of fracture. In addition, we repeated the main analyses including both quadriceps CSA Δ% and TUG Δ% as independent variables in the same Cox's proportional hazard regression models.

In addition, participants were stratified into CKD stages based on eGFR as follows: stage 3A – moderate CKD (eGFR 45–59 ml/min/1.73 m2), and stages 3B & 4 – moderate to severe CKD (eGFR 15–44 ml/min/1.73 m2). We then performed both within- and between-subject comparisons for CKD stage 3A and Stages 3B&4 using Wilcoxon and Mann-Whitney tests, respectively.

Significance testing was two-sided and based on a 5% probability level. Analyses were conducted using SPSS version 25 (IBM, USA).

3. Results

The participants' average age at baseline was 76 years (SD 4.8 years), and 36.7% were men. Table 1 details the characteristics of the participants who did (at least one hip fracture) and did not fracture during the median follow-up period of 5.6 years (after the second AGES-Reykjavik Study examination). Participants who did fracture were more likely to be female and were on average older, with lower body mass index, and smaller weight increase from middle age than subjects that did not fracture during the follow-up time. Physical activity and cognitive function were similar among the two groups.

Table 1. Baseline demographic and clinical characteristics for total analytical sample (n = 875) and according to fracture status by 2016.

Total
n = 875		No hip fracture
n = 783		Hip fracture
n = 92		P-value
% (N)
Female63.5 (556)61.8 (484)78.3 (72).002
Low PA level81.3 (711)80.5 (630)88.0 (81).078
Impaired cognitive function8.9 (78)8.4 (66)13.0 (12).142
History of fracture52.5 (459)51.9 (406)57.6 (53).296
Diabetes11.1 (97)11.5 (90)7.6 (7).261
Glucocorticoid use3.1 (27)3.3 (26)1.1 (1).241

Median (Q1, Q3)
Age, years76 (72, 80)75 (72, 79)78 (75, 82)<.001
Body mass index, kg/m227.2 (25.0, 29.8)27.2 (25.0, 29.9)26.3 (23.9, 29.1).049
Weight, kg75.3 (67.2, 84.6)76.4 (67.9, 85.2)71.9 (60.8, 76.0)<.001
Height, cm165.4 (159.3, 173.7)166.1 (159.6, 174.2)161.5 (157.4, 167.0)<.001
WC from age 50, %5.9 (−1.5, 13.4)6.2 (−1.2, 13.8)2.5 (−3.8, 12.2).015
Quad CSA, cm249.2 (42.0, 59.7)49.2 (42.6, 60.4)44.0 (37.8, 51.5)<.001
Quad density, HU45.2 (42.1. 48.2)45.4 (42.3, 48.3)43.9 (40.7, 46.2).002
KE strength, Nm117.4 (91.2, 161.7)119.9 (91.8, 166.0)103.8 (79.5, 129.6)<.001
TUG, s11.2 (9.8, 12.8)11.2 (9.8, 12.8)11.5 (9.8, 12.7).367

Abbreviations: CSA, cross-sectional area; Quad, quadriceps; KE, knee extension; PA, physical activity; TUG, Timed Up and Go test; WC, weight change; binary variables are reported as counts and proportions. For continuous variables, P-values were derived from Mann–Whitney U tests (univariate analysis), whereas binary variables were compared using chi-square tests.

At baseline assessment, quadriceps CSA, quadriceps density, and isometric KE strength were significantly lower in the participants who fractured compared to those who did not fracture. Baseline TUG performance was similar in both groups.

Regarding the longitudinal changes (Table 2), all participants experienced a significant loss in all muscle-related properties, while those who fractured their hip had a higher decrease in quadriceps CSA and TUG performance compared to those who did not fracture.

Table 2. Unadjusted longitudinal changes (% change per year) in muscle-related measures during the 5-year follow-up period.

Δ%/yearTotal
n = 875		No hip fracture
n = 783		Hip fracture
n = 92		P-value
Quadriceps CSA−1.2 (−2.1, −0.4)⁎−1.1 (−2.1, −0.3)−1.6 (−2.5, −1.1)<.001
Quadriceps density−0.6 (−1.5, 0.1)⁎−0.6 (−1.5, 0.1)−0.6 (−1.3, 0.2).575
Isometric KE strength−3.6 (−5.6, −1.1)⁎−3.5 (−5.7, −1.1)−3.8 (−6.2, −1.1).770
TUG1.4 (−0.6, 4.5)⁎1.3 (−0.7, 4.2)3.4 (0.5, 5.9).001

Values are median (Q1, Q3); abbreviations: CSA, cross-sectional area; TUG = Timed Up and Go; similar results were found using repeated measured ANCOVA to remove bias from possible confounding variables.

⁎

P < .001 for within-subject comparison between baseline and follow-up values using Wilcoxon test.

Using multivariable Cox proportional hazard models, a faster decrease in quadriceps CSA and a TUG performance were both associated with an increased risk of hip fracture, while a faster reduction in quadriceps muscle density and isometric KE strength were not associated with fracture risk (Table 3).

Table 3. Associations of accelerated muscle-related measures declinea with incident hip fracture.

Highest tertile of decline:ModelImpaired Kidney function
n = 875 (event, n = 92)
HR (95% CI)P value
Quadriceps CSA11.51 (1.00–2.29).050
21.55 (1.02–2.36).042
Quadriceps density10.97 (0.63–1.50).89
21.01 (0.66–1.56).96
KE strength11.16 (0.76–1.77).49
21.21 (0.79–1.86).38
TUG11.77 (1.17–2.68).007
21.80 (1.19–2.72).006

CSA = cross-sectional area, KE = knee extension, TUG = Timed Up and Go; Model 1 - adjusted for age and sex; Model 2 – additionally adjusted for weight, height, percentage of weight change from age 50, physical activity level, cognitive status, history of fracture, diabetes, and glucocorticoid use.

Bold values display significant hazard ratios.

a

The highest tertile of decline (high risk group) compared the two lowest tertiles (low risk group - reference).

Accounting for the competing risk of death did not substantially change the risk estimates. When we included quadriceps CSA and TUG (tertiles of Δ%) in the same model, the associations remained similar to results in the main analyses. Finally, changes in quadriceps size, density and strength, and lower limb muscular function did not differ by CKD-stage (stage 3A vs. stage 3B + 4; Supplementary Table S1).

4. Discussion

In this population-based study, we found that having impaired kidney function is associated with a significant loss in quadriceps area (i.e., muscle atrophy), density (i.e., fat infiltration), isometric strength, and TUG performance. Further, we found that faster quadriceps atrophy and faster decline in lower limb muscular function were associated with increased hip fracture risk. These findings highlight the relevance of these two muscle-related parameters for predicting hip fracture risk in older adults with impaired kidney function.

Although the link between kidney impairment and muscle atrophy and associated weakness is known (Schardong et al., 2018), most of the data collected in humans have derived from cross-sectional studies or small experimental studies. We found that older adults with impaired kidney function had a significant loss in quadriceps CSA over 5 years of follow-up. Changes in the CSA of the thigh (measured with CT) over two years have only been reported in a small group of CKD patients (John et al., 2013). It is well established that kidney function loss causes several complex systemic alterations that affect muscular homeostasis, leading to loss of muscle mass, although the mechanisms are not fully understood. Our data indicate that quadriceps density, an indicator of fatty infiltration, was also significantly reduced after a 5-year follow-up period. Lower skeletal muscle attenuation has been related to greater fat infiltration (Goodpaster et al., 2000b), increasing age (Goodpaster et al., 2001; Johannesdottir et al., 2018), and muscle force reduction (Goodpaster et al., 2001; Rahemi et al., 2015).

Our data showed that isometric KE strength was also substantially altered over 5 years. Evidence from another prospective cohort study (Roshanravan et al., 2015) showed that creatinine clearance was associated with KE strength decline throughout a 9-year follow-up period. Other longitudinal studies examining muscle strength loss in older adults reported that isometric quadriceps strength is lost at a rate of 1.5% to 4% per year (Goodpaster et al., 2006; Pham et al., 2016), which is similar to the rate of decline from our study (3.6%). Thus, the annualized rates of strength decline in these impaired kidney function older adults were not higher than those reported for relatively healthy older adults. This finding may be explained by the fact that the majority of our study sample (76.1%) had GFR between 45 and 59 ml/min/1.73 m2 (stage-3a CKD), while only 2.3% were in stage 4 (severely impaired GFR). Also, our results concur with the current knowledge supporting that in healthy older adults, the rate of decline in strength is often more rapid than the concomitant loss of muscle mass (Mitchell et al., 2012).

Previous studies have shown that impaired kidney function is associated with an increased fracture risk (Chen et al., 2018; Dukas et al., 2010), being dramatically higher in dialysis patients (Tentori et al., 2014; Mathew et al., 2014). In the current study, the incidence of hip fracture (11%) was in line with the incidence reported in Canadian non-dialysis CKD patients (16%) during a 10-year follow-up period, including atraumatic and traumatic fractures (Prasad et al., 2019). Data on fracture incidence in osteoporotic older men and women with and without impaired kidney function is expectedly higher, as previously reported (33.1% and 22.9%, respectively) (Dukas et al., 2010).

Predictors of hip fracture in older adults have been extensively investigated, given the impact of bone fractures on frail individuals' functional independence and quality of life (Cauley, 2013). Although older adults with kidney impairment are potentially susceptible to an aggressive deterioration in muscle mass, muscle quality, and physical function, their potential contribution to fracture risk prediction in this population has not been adequately addressed. Studies have focused on the relationship between bone measures and fracture risk in the field of CKD due to the well-known CKD-induced changes in bone and mineral metabolism, thus making other predictors considerably less studied. In this study, we observed that an accelerated loss of quadriceps CSA and a faster decline in TUG performance were significant predictors of hip fracture risk. Moreover, we found that the rate of loss in muscle strength was similar in older adults who fractured than those who did not fracture during the follow-up period. This finding differs from a prior prospective study in men and women (aged 60 and older at baseline), which reported the decline in isometric quadriceps strength was associated with an increased risk of fracture (Pham et al., 2016). Small mid-thigh muscular area and low knee extensor strength were significantly associated with fractures in both sexes in a larger sample from the same population study (AGES-Reykjavik Study) (Johannesdottir et al., 2012). Several other studies have examined the contribution of muscle size and/or strength to fracture risk assessed only at baseline, thus not capturing the rate of loss (Lang et al., 2008; Harvey et al., 2018). This is the first study that had included four distinct muscle-related parameters measured on two occasions (change from baseline to follow-up) before hip fractures occurred.

Our results indicate that an accelerated decline in muscular function (TUG test performance) predicted fracture risk independent of other risk factors. This finding is particularly relevant as the TUG test does not require specialized expertise or equipment and is an inexpensive method often used as a clinical screening tool. Based on cross-sectional data from community-dwelling older adults, previous studies have reported that slow TUG performance is significantly associated with an increased risk of fractures (Zhu et al., 2011; Jeong et al., 2019). A cross-sectional study in CKD patients with stages 3–5 CKD showed that TUG and 6-minute walk test could discriminate fracture status (West et al., 2012). It is currently unclear why only the rate of loss in quadriceps size and TUG performance were independent predictors of hip fracture risk, and more mechanistic studies may be needed to clarify these aspects. There are still many questions unsolved about the pathophysiologic mechanisms associated with aging and CKD that can explain the changing rates in muscle-related characteristics and their potential contributions to fracture risk. It is also apparent that the progression and severity of the CKD will differently compromise muscle mass, strength, quality, and function and disrupt the metabolism.

Only two studies have examined the association between CT-derived muscle measurement (CSA and attenuation) and hip fracture in older adults (Lang et al., 2008; Lang et al., 2010). In the current study, we found a significant association between an accelerated decrease in quadriceps CSA, but not in density, and incident hip fracture risk. CT data capturing muscle structural features provide valuable information for a better understanding of hip fracture etiology. However, the use of CT for improving fracture prediction compared with existing clinical practice techniques deserves further research.

The major strengths of our study were: the exploration of changes over time in different muscle-related parameters, enrollment of community-dwelling older adults not selected based on kidney disease or risk for fracture, large sample size, long follow-up time, and the analyses accounted for common risk factors for falls and fracture (based on detailed baseline data). The AGES cohort offers the opportunity to study these associations in a sample that closely resembles the heterogeneity of health measures and outcomes in the older adult population. However, our research had some limitations, including its reliance on a single measurement to estimate GFR by an equation based on serum creatinine (as is common in epidemiological studies), lack of data on the incidence of falls during follow-up, and our results may not be generalizable to other populations, including other white groups with different characteristics or other ethnicities. Our main analysis was also not stratified by CKD stage as we did not find differences in the rate of decline of all muscle-related variables between stages.

In conclusion, we observed that accelerated declines in quadriceps area and lower limb muscle function predict hip fracture risk in older adults with impaired kidney function. Results also highlight the potential usefulness of a simple measurement that captures lower limb muscular function (i.e., TUG test), which can be easily implemented during routine clinical examinations. This study adds insight into the potential kidney-related changes in skeletal muscle parameters measured with different tools, thus improving our understanding of the mechanisms leading to physical frailty and fracture risk in this population.

CRediT authorship contribution statement

E.A.M. was responsible for the study concept and design, analysis, and interpretation of data, and drafted the manuscript. M.E. and J.L.V. participated in the analysis and interpretation of data, in the critical revision and final approval of the manuscript. T.A., T.L., and K.S. were responsible for the acquisition of participants and/or data and approved the final version. V.G., G.S., S.S., L.L., G.E., and T.B.H. were responsible for the study concept and design, acquisition of participants and data, critical revision of the manuscript or important intellectual content and approval of the final version.

Declaration of competing interest

The authors have no conflicts.

Acknowledgments

Financial disclosure

This work was supported by National Institutes of Health contract N01-AG-1-2100, the Hjartavernd (the Icelandic Heart Association), and the Althingi (the Icelandic Parliament). This research was supported in part by the Intramural Research Program of the NIH, National Institute on Aging.

Research Center in Sports Sciences, Health Sciences and Human Development (CIDESD) is supported by the Portuguese Foundation for Science and Technology (UID/04045/2020).

Appendix A. Supplementary data

Download : Download Word document (109KB)

Supplementary material

References

View Abstract
© 2021 Elsevier Inc. All rights reserved.