This was a cross-sectional study of 294 community-dwelling older adults based in Melbourne, Australia. Participants attending the falls and fractures clinic for clinical assessment between 2016-19 were included. Participants fulfilled the following criteria: aged over 65 years with a history of falls or at risk of falls, ability to mobilize independently with or without the use of gait aids, and no history of cognitive impairment. This study was approved by the Western Health local Human Ethics Research Committee (DB2017.13 and QA2018.80-46205).
Assessments
Participants undertook a comprehensive assessment which included the identification of risk for falls and fractures. Falls (in the past year) and fracture (in the past 5 years) history were determine by interview. Assessments of physical performance and balance, as well as body composition to determine ALM/h2 using DXA (Hologic Inc., Bedford, MA, USA) were also performed. Participant reported fear of falling was determined using the Falls Efficacy Scale – International.(11)
Physical performance measures conducted on participants included the assessment of height and weight for calculation of body mass index (BMI), handgrip strength, gait speed, timed up and go (TUG) and Short Physical Performance Battery (SPPB). Handgrip strength was assessed using a Jamar hydraulic dynamometer (Sammons Preston Inc.) with the participant seated with their arm resting on chair arms. Participants were instructed to squeeze the dynamometer at their maximal effort, with the test performed 3 times on each side and 30 seconds rest provided between each trial. The best score out of the 3 trials was recorded.
To assess gait speed, we used the GAIT Rite® (CIR Systems Inc, Havertown, PA) instrumented walkway system (580 cm × 89 cm × 0.625 cm, sample rate = 120 Hz). The GAIT Rite mat was used to record spatiotemporal data. It was positioned along a straight section of the walkway and participants were instructed to walk at their normal speed. Three trials were provided with the best gait speed recorded. Gait aids were used a necessary.
To assess TUG performance, we used a 3-meter course.(12) Participants began in a seated position and were instructed to stand, walk 3 meters to a marked area, then return to the starting seated position. The TUG test was performed twice at the participants normal speed with the best time of completion recorded. A gait aid was allowed for use as needed.
The SPPB is an assessment which includes 5 tests for lower limb function, including balance, strength and mobility.(10) Balance assessments were composed of 3 parts which progressed in difficulty (feet together stand, semi tandem and full tandem stand). The aim of the balance tests was to stand for 10 seconds unaided, with the test progressing in difficulty after successful completion. Gait speed was assessed as previously described (using GAIT Rite assessment results). The 5 times sit to stand test was performed with the participant starting in the seated position. After confirming ability to perform 1 sit to stand action, participants were then instructed to stand and sit 5 times as quickly as possible, ensuring feet were flat on the floor. Scores were allocated according to performance, with an overall maximum score of 12.
Classifying Sarcopenia
Sarcopenia status was classified based on the EWGSOP1 and EWGSOP2 definitions. This study focused on those who were classified as having severe sarcopenia and therefore fulfilled the 3 criteria for muscle strength, ALM/h2 and physical performance. Participants were classified as severely sarcopenic according to 7 different sarcopenia criteria. Using the EWGSOP1 definition, severe sarcopenia was defined as those with low ALM/h2, gait speed and handgrip strength. For the EWGSOP2 definition, 6 severe sarcopenia groups were created based on fulfillment of cut-points for muscle strength (handgrip strength or sit to stand test), ALM/h2 and physical performance (low gait speed, SPPB score or increased TUG time).
Statistical Analysis
Baseline characteristics for participants are presented as a median (IQR) or frequency (%). The diagnostic value of the SPPB was assessed by calculating the area under the receiver operating characteristic (ROC) curve separately for each sarcopenia definition. This was repeated with the addition of ALM/h2 to the model. Area under the ROC curve (AUC) of 0.7-0.8 was considered to be acceptable, with 0.8-0.9 excellent and >0.9 outstanding.(13) Optimal cut-points for the SPPB in diagnosing sarcopenia were defined by 3 different methods: maximizing the product of sensitivity and specificity (Liu),(14) maximizing their sum (Youden)(15) and cut-point on the ROC curve closest to perfect sensitivity and specificity (Nearest 0,1).(16) All analyses were performed using Stata 15.1 (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC).