Reliability and validity of a full-body function Get-Up test in older adults

Abstract

Identifying deficiencies in physical function in older adults is critical to evaluate important health outcomes like sarcopenia, but current protocols are expensive and require complex equipment. This study evaluates the reliability and validity of an inexpensive, simple new Get-Up test in older adults. It involves participants moving quickly from standing upright, to lying flat, then rising to a standing position unassisted. A total of 293 relatively healthy older adults without severe health conditions (e.g., cardiovascular, psychological, degenerative, or physical impairments) completed the Get-Up test twice for familiarization and twice for timed trials on two separate days alongside numerous validated clinical tests commonly used to assess strength, function, and fitness in older adults. ANOVA with post-hoc analysis and intraclass correlation (0.928 (95% CI [0.914, 0.940])) indicated strong reliability, with the second timed trial comparable to trials on a separate day. The Get-Up test was significantly (p <.0001) negatively correlated with all referenced measures of strength (Biodex peak torque, r = −.41, 1-repetition maximum, r = −.26, handgrip, r = −.38) and function (Short Physical Performance Battery, r = −.49, gait speed, r = −.39) and significantly (p <.0001) positively correlated with fitness (400-m walk, r =.70), which strongly predicted Get-Up test performance, suggesting good validity. Poor performance was associated with baseline sarcopenia prevalence (bottom tertile vs. top tertile: odds ratio 3.99 (95% CI 1.64–9.67)) and sarcopenia incidence after 1-year follow-up (hazard ratio 3.47 (1.10, 10.98)), suggesting potential to evaluate sarcopenia. This simple and safe Get-Up test requires minimal equipment, personnel, and expertise, yet it has good reliability and validity as a potential novel tool for full-body physical function in older adults that is associated with sarcopenia prevalence and incidence.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11357-025-01643-4.

Introduction

Identifying deficiencies in physical functioning and ability is critical for accurately diagnosing current health status, which is a necessary step for establishing a probable prognosis and guiding effective treatment planning in older adults [1]. Age- and physical inactivity–related loss in muscle mass that leads to sarcopenia has been identified as a primary driver of these deficiencies in physical functioning. Almost one in five Americans is already 65 years or older [2]. With the growing expansion of this population, the additional costs of healthcare and lower quality of life associated with increased morbidity, mobility disability, and mortality in older adults [3] underscore the urgent need to focus on identifying deficiencies in physical function and strength [4], given their established relationship with these adverse outcomes [5].

The role of prevention is paramount when addressing decreased physical function, particularly given the close association between muscle strength, muscle mass, and sarcopenia. Sarcopenia, characterized by the progressive loss of muscle mass and strength [6], is challenging because physical activity interventions in older adults tend to have substantial drop-out as reported by a recent systematic review with a mean 24.6% drop-out and a max of 74% [7]. The insidious nature of sarcopenia, how it is typically masked by stable body weight [8], may only become apparent after a significant detrimental event such as a fall or the onset of a disability [9]. Maintenance of muscle mass requires drastically less effort than regaining substantial quantities after a loss of muscle mass has occurred [10]. The functional reserve capacity, which relates to additional physical ability above and beyond the activities of daily living (ADL)[11], should be built-up before old age and protected to avoid becoming dependent [12]. Individuals with little to no functional reserve capacity are hospitalized for longer on average, pay more in health care costs, and have higher risk for mortality than those with greater capacity [13]. Current recommendations on how to preserve skeletal muscle function include increasing physical activity, specifically resistance exercise, and improving the quality of diet, especially establishing good habits early in life [14].

However, determining who is most in need of prevention efforts is often difficult to ascertain. Physical screening is needed beyond simply surveying an individual’s activities of daily living (ADLs) [15], taking a sarcopenia survey screener such as SARC-F (a simple five-question survey on strength, assistance with walking, rising from a chair, climbing stairs, and falls) [16] or Mini Sarcopenia Risk Assessment (MSRA) [17], or combining a survey with a physical measurement, such as calf circumference [18]. These methods have been criticized for their limited sensitivity and specificity in diagnosing sarcopenia, as they may fail to detect early or mild cases and are prone to misclassification [19, 20]. Given what is known about human movement, we posit that a physical function test may be more appropriate to screen for sarcopenia, as it would directly assess aspects of skeletal muscle function. A valid and reliable, objective screening measure should have favorable qualities: it should be safe, have an established cutoff, be cost-effective, and be easy to perform for healthy adults [21]. Various screening tools have been proposed to aid in early identification and prevention of sarcopenia, which ought to reduce the diagnostic burden of existing definitions, but further research is needed before use in the general population [10]. Although handgrip strength provides a proxy of well-being and has modest predictive validity for decline in cognition, mobility, functional status, and mortality [22], a simple and easy whole-body measure of functional ability may still be more applicable to diagnose and prevent sarcopenia at the population level. Such a test could alleviate much of the expense and tedium in diagnosing sarcopenia, which currently requires advanced technology to assess muscle mass (e.g., bioelectrical impedance analysis, dual-energy X-ray absorptiometry, computed tomography, or magnetic resonance imagery). Until a suitable sarcopenia screening test is developed, it will continue to be challenging to identify at-risk populations and thus improve prevention efforts.

To improve the screening for decreased muscular strength and physical function, which are components of sarcopenia, a simple Get-Up test was developed, which requires only a modest allotment of time, space, equipment, personnel, and expertise. The newly developed Get-Up test may improve upon several issues seen with existing screening, specifically, their reliance on self-report questionnaires (SARC-F, MSRA), existence of floor and ceiling effects with score-based tests (Short Physical Performance Battery), and focus on a specific type of movement or limited muscle groups that restrict a test’s generalizability (Sitting-Rising test [23]). Therefore, the purpose of this study is to (1) determine the reliability and validity of the test compared to existing measures of strength and function that are commonly used in research among older adults, acknowledging that further research may be necessary to fully establish its diagnostic utility for sarcopenia, and (2) explore the predictive capacity of the test to identify future sarcopenia cases.

Methods

Participants

Participants were recruited from a Midwest university town and surrounding communities through postings, advertisements, and informational talks at local organizations. Between October 2015 and May 2016, 293 community-dwelling older adults, ages 65 years and older, were recruited. Participants were ambulatory and free of self-reported severe conditions (major heart, psychological, degenerative, or physical impairments) at baseline.

Study design

Baseline data were collected in a series of laboratory visits over the course of 2 weeks involving physical activity measurement, medical history questionnaires, body composition, fitness, function, and strength tests and were repeated a year later. The Institutional Review Board at Iowa State University approved this study (15–430). Procedures were fully explained to all participants. Prior to participation, all individuals provided written informed consent.

Measures

Height was measured using a standard stadiometer in centimeters. Weight was measured with a digital scale (Cardinal Detecto 758 C Digital Scale, Webb City, MO, USA) in kilograms. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. Waist circumference was measured in a standing position at the level of the umbilicus using a measuring tape after exhalation. Measurements were taken while participants wore scrubs with no shoes.

A comprehensive survey (lifestyle, physical activity, and medical history) questionnaire was administered. The detailed self-report survey asked for the average aerobic and resistance physical activity and sitting time in a week over the past 3 months in the four domains (work, transportation, household, and leisure time). The survey was developed based on national and international physical activity questionnaires such as National Health and Nutrition Examination Survey (NHANES) and International Physical Activity Questionnaire (IPAQ). Falling was assessed via self-reported number of falls in the past 12 months.

The 7-day step count was measured using an Omron accelerometer-based pedometer (HJ- 321, Omron Healthcare Co. Ltd, Kyoto, Japan). The Omron triaxial pedometer has been validated and used in large clinical trials [24, 25]. The multiple piezoelectric sensors capture acceleration waveforms to determine steps and are not restricted to specific positions on the participant to accurately count steps. The monitor stores a running 7 days of step data. Participants were instructed to wear the pedometer on the waist.

The Short Form Health Survey (SF- 36) was used as a generic measure to compare the relative burden of diseases and to differentiate health benefits produced by a wide range of treatments. Areas of health measured include two general measures: physical and mental health, which are broken down into further categories. Physical health scales include physical functioning, role-physical, bodily pain, and general health, whereas mental health scales are vitality, social functioning, role-emotional, and mental health [26]. The average score for each of the eight subcategories were summed to get a global score (ranging from 0 to 800, a higher score indicates higher perceived health). The survey showed high reliability (Cronbach’s α > 0.85) and construct validity with other valid measures [27].

The Geriatric Depression Scale is a 15-item scale for epidemiological research developed by the National Institute for Mental Health. This survey measured the frequency of depressive symptoms. The questions were answered using a yes or a no. A higher score (ranging from 0 to 15) indicates a more depressive state [28]. The scale showed high reliability (Cronbach’s α = 0.92) and area under the curve of 0.89 in a receiver operating characteristic (ROC) analysis indicating usefulness in screening and diagnosing clinical depression [29].

A full-body scan was performed on the Hologic Horizon W model DXA (Hologic Inc., Bedford, MA) and Apex Software (Version 5.5.3). A trained technician performed the scan with the participant supine in the standard fashion and wearing scrubs. Software automatically defined regions on the trunk and appendages, which were then adjusted manually by one trained technician. DXA software measured whole and regional body composition measures, including fat mass (FM), percentage body fat (%BF), fat-free mass (FFM), and appendicular lean mass (ALM). The DXA machine was calibrated daily using the manufacturer-provided phantom spine segment. DXA showed high reliability (Cronbach’s α > 0.99) and good concordance across body composition compartments with computed tomography [30].

A 3-day diet record was collected from two weekdays and one weekend day and was entered into The Food Processor SQL version 10.14.1 (ESHA Research, Salem, OR). Exclusion criteria for diet-records included inadequate nutrient information, improper recording of weekday to weekend-day ratio, and ineligible handwriting.

Sarcopenia was defined following the European Working Group on Sarcopenia in Older People (EWGSOP) guidelines due to Caucasian population and frequent use in published literature [31]. The EWGSOP diagnosis of sarcopenia required low appendicular lean mass (ALM/ht² female ≤ 5.67, male ≤ 7.23) and either gait speed ≤ 0.8 m/s or low handgrip strength (female < 20 kg, male < 30 kg).

The Get-Up test assessed full-body musculoskeletal performance through evaluation of the ability to move from a standing position to lying on your back and then returning to standing as quickly and safely as possible. No prescribed method of lowering to or rising from the ground was given, but individuals were encouraged to use their whole body both quickly and safely through the movements. The test was administered on a non-slippery and flat surface, on a large, padded mat, with the participant in non-restrictive clothing without shoes. The evaluator explained the movement and instructed: “Starting standing upright, with your arms across your chest, move as quickly as you are able to safely from standing to a lying position and touch your hands behind your neck (to prevent a potential head or neck injury), then get up off the ground and return to a standing position and place your arms across your chest. Before the test is timed, you will practice the test twice to get familiar with the movements.” Once the participant was familiar with the movement, the evaluator said: “3, 2, 1, GO,” and started the stopwatch. During the movement a spotter remained next to the participant for safety. A 1-min break was given between trials. The participant completed the movement, and once their body was upright and their arms were across their chest, the evaluator stopped the stopwatch and recorded the time in seconds. Based on preliminary findings from pilot testing, we attempted to avoid a learning effect by having the participants practice the movement twice before they would be timed. This allowed them to determine a movement strategy that they would use during the timed trials. Participants who would have been unable to complete a test like the Sitting-Rising test [23] due to a joint issue were able to successfully complete the Get-Up test due to the freedom to perform the test as needed. See Fig. 1 for a graphical representation of the movement. To establish test–retest reliability of the Get-Up test, the protocol was repeated again 3 days later for the entire sample. A single rater gave instructions for the test.

Fig. 1.

Start (standing upright), middle (lying down), and finish (standing upright) position for the Get-Up test

The Biodex System 3 (Biodex Medical Systems, Shirley, NY, USA) is an isokinetic dynamometer, which allows constant velocity with accommodating resistance throughout the range of motion of the joint. Resistance was maintained through an electric servo-controlled mechanism at a constant speed. It is considered the gold standard for strength measurement [32]. According to the manufacturer’s instructions, the device was properly warmed up and calibrated before data collection. The protocol was explained to participants before using the equipment. The individual was seated and secured with chest straps (see Supplemental Fig. 1). Their arm was positioned immediately adjacent to the lever arm. The arm was nearly straight, and the epicondyles of the arm were aligned with the axis of the dynamometer. The grip was positioned such that it rests comfortably in the hand when in neutral position (between supine and prone). The individual pulled using their biceps through 90° of motion towards their shoulder with their hand supine and then extended their arm using their triceps with the hand prone until they returned to the starting position. Similarly, the individual’s leg was positioned immediately adjacent to the lever arm. The leg was nearly straight, and the epicondyles of the femur were aligned with the axis of the dynamometer. The leg was strapped in just above the ankle joint on the lower shin. The participant curled their leg with their hamstring muscles through 90° of motion and then extended their leg using their quadriceps until they returned to the starting position. Each individual was encouraged to use maximal effort for each repetition. The protocol consisted of one set of five maximal repetitions at 60°/s on their dominant side using the biceps, triceps, quadriceps, and hamstring muscles. Allowing for five repetitions reduced the learning effect that might have been present in the first few repetitions using a new machine. The peak torque generated for each muscle group during the set was recorded and summed to create a measure of whole-body strength. The system showed high reliability (ICC = 0.99–1.00) across trials and days and near perfect agreement between measures and associated criterion at < 300°/s [33].

American College of Sports Medicine guidelines were followed to determine muscular strength from one repetition maximum (1RM) chest and leg press [34]. The participant was given three progressive warm-up sets each followed by a short rest. The 1RM was determined within five trials with rest periods of 2 min between trials. The initial weight was selected to be within the individual’s perceived capacity, and then resistance was progressively increased until the participant could not complete the repetition throughout the whole range of motion. The final weight lifted successfully was the individual’s 1RM. Total weight from 1RM chest press and leg press tests were summed to get a single score of upper and lower body strength. The 1RM test had high reliability (Kappa index from 0.93 to 0.95, intraclass correlation coefficient = 0.99) and detected change between 1 and 3% indicating changes less than 1 kg could be identified [35].

Handgrip strength was measured using a calibrated, digital dynamometer (Jamar Plus +, Lafayette Instrument, Lafayette, IN, USA). The width of the device was adjusted to the size of the participant’s hand such that the middle phalanx rests on the inner handle [36]. Participants sat and held their elbow joint at a right angle and gripped the dynamometer for 2 s using each hand with maximal effort. The best of three trials with 1-min rest between trials was recorded for each hand [37]. Handgrip strength using the Jamar Plus + showed high reliability (ICC = 0.94–0.99) and is considered a gold standard device [38].

The Short Physical Performance Battery (SPPB) is a short series of tests (balance, 4-m gait speed, chair stand) that gives a composite score [39]. Tests are weighted equally with a score between zero and four, and a higher score indicating higher function (total ranging from 0 to 12). The balance test consisted of holding three standing positions for up to 10 s for each position: (1) side by side, (2) heel of foot beside toe of the other foot, (3) heel of foot in front of other foot. The participant’s usual gait speed was measured over 4 m. Participants were instructed to walk at their usual pace. Participants began walking at the start line and were instructed not to decelerate before crossing the finish line. Walking aids were allowed if used in everyday life. The time to complete the 4-m course was recorded. The repeated chair stand test was the time it took for the participant to stand up from a seated position five times in a row without stopping and keeping their arms folded across their chest. The SPPB has high reliability and validity [40].

The 400-m walk test was the time required to walk 10 laps as fast as possible on a 20-m-long course [41]. Rests of up to 60 s while standing were permitted. The test was conducted in a long hallway that is approximately five feet wide. Cones marked the ends of the course. The researcher told the participant the number of laps completed and remaining each round. The 400-m walk test has high reliability (ICC = 0.95) and validity with measured maximum volume of oxygen uptake [42].

Statistical analysis

Data are presented as mean ± standard deviation or n (%). Participant characteristics are displayed for the whole sample as well as compared between Get-Up test groups. Shapiro–Wilk tests were conducted to test normality, and histograms to visualize normality. Differences between Get-Up test tertiles are compared using the analysis of variance and chi-squared tests for continuous and categorical variables, respectively. The reliability of the Get-Up test was examined using Pearson’s correlation coefficient and repeated measures analysis of variance with post hoc t-tests. Test–retest reliability of the Get-Up was measured using a two-way intraclass correlation coefficient (ICC) model with consistency definition to focus on the stability of each participant’s rank-over-time, allowing for learning effects without penalizing reliability. The ICC estimate and 95% confidence interval were calculated to quantify the degree to which participants’ rank orders remained consistent across the four trials. Validity of the Get-Up test was explored through Pearson’s correlation coefficient, simple linear regression, and multiple linear regression with the Get-Up test as the dependent variable in comparison with commonly used strength, function, and fitness tests as the independent variables. Multiple logistic regression and Cox proportional hazard regression were used to identify associations between independent variables (Get-Up test or SPPB and confounding variables) and sarcopenia as the dependent variable. The ability of the Get-Up to discriminate between individuals with and without sarcopenia was determined using a ROC approach. Specifically, the “pROC” package (version 1.18.5) in R was used to generate the ROC curve and estimate the area under the curve (AUC). The sarcopenia dichotomous classification served as the reference standard. To determine an optimized cutoff for the Get-Up measure, the Youden Index (sensitivity + specificity − 1) was maximized. Significance was set at a p-value of < 0.05. Statistical analysis was performed using SAS 9.4 (SAS Institute Inc., Cary, NC, USA) and R 3.4.1 (R Foundation for Statistical Programming, Vienna, Austria) [43]. Graphs were prepared in ggplot2 3.1.1 [44].

Results

Descriptive statistics

Participants (Table 1) had a mean age of 72.8 ± 5.8 years (range 65–96), were nearly all Caucasian (> 99%), were 58% female, and had slightly elevated body mass index (27 ± 4.9 kg/m²). Thirty-three percent were former smokers (only 1% current smokers), accumulated a low number of daily steps (4943 ± 2632 steps/day), reported high levels of physical activity (PA), were high function (94% scored ≥ 10 on SPPB, ranging from 5 to 12), and reported good to excellent health (99%) at baseline. The discrepancy between daily steps and reported PA may reflect differences in measurement, as the pedometer captures some but not all PA reported in the survey, while also missing non-ambulatory activity and PA intensity, and surveys include recall and social desirability bias. Seventy percent of the sample had diet records of sufficient quality to be analyzed. There was a low prevalence of comorbidities (72% without major chronic diseases) and low levels of depressive symptoms. Eighty-eight percent were college-educated, and 81% earned more than $50,000 annual household income. Across the Get-Up test tertiles, the groups differed in all variables except for weight, waist circumference, moderate and light aerobic and resistance activity, depressive symptoms, education, income, and comorbidities.

Table 1.

Participant characteristics by Get-Up test speed

		Get-Up test (tertiles)
	All	Low (fast)	Moderate	High (slow)	p-value
Age (years)	72.8 (5.8)	70.0 (4.6)	72.2 (5.6)	74.3 (6.3)	<.0001
Sex (% female)	56.7 (166)	59.8 (58)	44.4 (44)	25.8 (25)	<.0001
Height (cm)	168.4 (9.6)	171.3 (9.6)	168.9 (9.7)	165.7 (8.9)	0.0002
Weight (kg)	76.9 (16.6)	76.1 (15.1)	75.9 (15.8)	78.2 (18.8)	0.57
Waist circumference (cm)	93 (14)	90 (12)	93 (13)	95 (17)	0.10
Smoking status					0.002
Never	65.7 (199)	74.2 (72)	52.5 (52)	71.9 (69)
Current	1.0 (3)	0.0 (0)	3.0 (3)	0.0 (0)
Former	33.3 (102)	25.8 (25)	44.4 (44)	28.1 (27)
Daily alcoholic drinks	0.5 (0.6)	0.6 (0.7)	0.5 (0.6)	0.3 (0.5)	0.01
Heavy alcohol consumption	5.3 (16)	6.2 (6)	6.1 (6)	3.1 (3)	0.54
Average daily calories intake (kcal)	1792 (469)	1922 (577)	1725 (382)	1737 (422)	0.02
Dietary protein intake (g/kg/day)	1.0 (0.4)	1.1 (0.5)	1.0 (0.3)	1.0 (0.3)	0.01
Body composition
Body mass index (kg/m2)	27.0 (4.9)	25.8 (3.8)	26.5 (4.0)	28.4 (5.9)	0.0004
Percentage body fat (%)	39.7 (7.7)	35.4 (7.0)	39.1 (6.6)	43.5 (6.9)	<.0001
Fat mass (kg)	30.2 (9.7)	26.4 (7.3)	29.2 (7.8)	33.8 (11.3)	<.0001
Fat free mass (kg)	42.9 (10.0)	45.6 (10.2)	43.0 (9.9)	40.5 (9.8)	<.0001
Appendicular lean mass (kg/m2)	6.5 (1.2)	6.9 (1.2)	6.5 (1.2)	6.2 (1.2)	0.002
Female	5.8 (0.8)	5.7 (0.7)	5.7 (0.6)	5.9 (1.0)	0.50
Male	7.5 (0.9)	7.6 (0.9)	7.5 (0.9)	7.4 (0.9)	0.77
Strength
1 rep. max chest press (kg)	34 (17)	42 (18)	34 (16)	27 (13)	<.0001
1 rep. max leg press (kg)	83 (35)	98 (36)	82 (33)	72 (31)	<.0001
Grip strength (kg)	30 (10)	36 (10)	31 (9)	24 (8)	<.0001
Female	24 (5)	22 (4)	24 (5)	27 (4)	<.0001
Male	39 (9)	35 (10)	37 (7)	41 (7)	0.006
Function
SPPB	11.4 (1.1)	11.8 (0.6)	11.7 (0.7)	11.0 (1.3)	<.0001
Gait speed (m/s)	1.1 (0.2)	1.2 (0.2)	1.1 (0.2)	1.1 (0.1)	<.0001
Female	1.1 (0.2)	1.1 (0.2)	1.2 (0.2)	1.2 (0.1)	0.001
Male	1.2 (0.2)	1.1 (0.2)	1.1 (0.2)	1.2 (0.2)	0.003
Fitness (400-m walk time in min)	4.5 (0.8)	4.0 (0.4)	4.4 (0.5)	5.0 (0.8)	<.0001
Daily steps	4943 (2632)	5734 (2711)	4970 (2498)	4366 (2522)	0.001
Physical activity (MET-hours/week)
Vigorous aerobic	14.1 (18.7)	19.4 (20.9)	15.9 (18.9)	8.0 (14.1)	<.0001
Moderate aerobic	81.7 (65.5)	84.2 (60.0)	81.9 (64.9)	80.2 (70.1)	0.91
Light aerobic	67.3 (36.4)	63.2 (34.5)	64.5 (36.8)	73.3 (37.3)	0.11
Resistance	17.7 (23.1)	20.1 (20.7)	17.5 (19.7)	14.8 (22.4)	0.21
Sedentary time (hours/day)	11.9 (5.0)	11.4 (3.7)	11.1 (4.6)	12.9 (5.4)	0.01
Self-reported health					0.001
Excellent	77.0 (25.3)	0.0 (0)	0.0 (0)	1.0 (1)
Very good	162.0 (53.3)	39.2 (38)	25.3 (25)	14.4 (14)
Good	59.0 (19.4)	50.5 (49)	53.4 (53)	57.7 (56)
Fair	4.0 (1.3)	10.3 (10)	21.2 (21)	22.7 (22)
Poor	0.0 (0.0)	0 (0.0)	0 (0.0)	4.1 (4)
Health-related Quality of Life (SF- 36)	725 (92)	751 (67)	731 (89)	695 (99)	<.0001
Geriatric Depression Scale (GDS)	0.9 (1.4)	0.7 (1.1)	1.0 (1.6)	1.0 (1.5)	0.07
Education					0.73
High school	12.2 (37)	10.3 (10)	13.1 (13)	13.4 (13)
College	39.5 (120)	35.1 (34)	41.4 (41)	40.2 (39)
Graduate	48.3 (147)	54.6 (53)	45.5 (45)	46.4 (45)
Income					0.43
Low (< $50,000)	16.1 (49)	8.5 (8)	18.7 (18)	21.9 (21)
Moderate ($50,000–74,999)	22.0 (67)	20.2 (19)	20.8 (20)	25.0 (24)
High (≥ $75,000)	58.6 (178)	71.3 (67)	60.4 (58)	53.1 (52)
Comorbidities					0.79
0	71.7 (218)	79.0 (75)	73.7 (70)	73.3 (66)
1	18.4 (56)	14.7 (14)	21.1 (20)	21.1 (19)
2	5.2 (16)	6.3 (6)	5.3 (5)	5.6(5)
3	0.0 (0)	0 (0.0)	0 (0.0)	0 (0.0)
4	0.3 (1)	0 (0.0)	0 (0.0)	0 (0.0)
Sarcopenia (EWGSOP)	10.2 (30)	1.0 (3)	9.1 (9)	18.6 (18)	<.0001

Data are presented as mean (SD) or % (n). Heavy alcohol consumption was defined as > 14 and > 7 alcohol drinks per week for males and females, respectively. SPPB, Short Physical Performance Battery; MET, metabolic equivalent. Comorbidities include myocardial infarction, congestive heart failure, heart arrhythmia, stroke, abnormal ECG, type 2 diabetes, and chronic obstructive pulmonary disease. EWGSOP: European Working Group on Sarcopenia in older people. The p-value reflects differences in Get-Up tertile group, determined using analysis of variance or chi-square tests

Reliability of the Get-Up test

The two timed attempts of the Get-Up protocol were recorded (Table 2). See Fig. 2 for the mean and standard deviation of Get-Up test times for each age and sex group. Pearson’s correlation coefficient and partial coefficient (controlling for age and sex) showed a very strong relationship between trials completed on the same day (r = 0.96–0.97) as well as on different days (r = 0.89–0.91). A repeated-measures analysis of variance uncovered a significant main effect for both day (p < 0.0001) and trial (p = 0.009), but post hoc t-tests showed that only the Day 1:Trial 1 differed significantly from the Day 1:Trial 2 (p = 0.045), Day 2:Trial 1 (p = 0.0004), and Day 2:Trial 2 (p < 0.0001). No differences were revealed between Day 1:Trial 2 and Day 2:Trial 1 (p = 0.46), and Day 2:Trial 2 (p = 0.052). The four Get-Up trials (Day 1:Trial 1, Day 1:Trial 2, Day 2:Trial 1, and Day 2:Trial 2) demonstrated excellent internal consistency. The two-way ICC(C,1) was 0.928 (95% CI [0.914, 0.940]), indicating that participants’ relative performance on the Get-Up was highly consistent across the repeated assessments. An F-test for H₀: ICC = 0 was significant (F(286, 858) = 52.2, p < 0.001).

Table 2.

Get-Up test performance (seconds)

		Mean	SD
Monday PM	Trial 1	8.1	3.7
	Trial 2	7.8	3.3
Thursday PM	Trial 1	7.5	3.6
	Trial 2	7.4	3.3

Repeated-measures analysis of variance revealed significant main effects for day (p < 0.0001) and trial (p = 0.009). Follow-up pairwise comparisons showed Monday Trial 1 differed from all others, but Monday Trial 2 did not differ from either Thursday trial

Fig. 2.

Mean Get-Up times and standard deviations by sex (column 1 female, column 2 male) and age groups

Concurrent criterion validity of the Get-Up test

The Get-Up test was compared to a wide range of existing tests for strength (Biodex isokinetic dynamometer, 1RM chest and leg press, handgrip strength), function (SPPB balance, SPPB total, usual gait speed), fitness (400-m walk test completed as quickly as possible) and falls (Table 3). We found weak to moderate negative correlations for Biodex (r = − 0.41, p < 0.0001), 1RM chest and leg press (r = − 0.26, p < 0.0001), handgrip strength (r = − 0.38, p < 0.0001), SPPB (r = − 0.49, p < 0.0001), SPPB balance component (r = − 0.26, p < 0.0001), and usual gait speed (r = − 0.39, p < 0.0001). Therefore, as the time to complete the Get-Up test increases (slow Get-Up), strength/function decreases. We found moderate positive correlations with the 400-m walk test (r = 0.70, p < 0.0001). Therefore, as the time to complete the Get-Up test goes up, so does the time to complete the 400-m walk. We observed no correlation with the number of falls in the past 12 months (r = 0.06, p = 0.35). When controlling for age and sex, all correlations showed a minimal change or decreased slightly, except 1-repetition maximum strength test, which was no longer significant (r = − 0.09, p = 0.16).

Table 3.

Pearson’s correlation and partial correlation of Get-Up test and reference tests

	Biodex peak torque (kg/s)	1 rep. max. chest/leg press (kg)	Handgrip strength (kg)	SPPB balance (s)	SPPB score (0–12)	Usual gait speed (m/s)	400-m walk (s)	Falls in past 12 months
Get-Up test (s)	r(287) = −.41, p <.0001	r(267) =  −.26, p <.0001	r(291) = −.38, p <.0001	r(291) = −.26 p <.0001	r(291) = −.49, p <.0001	r(291) = −.39, p <.0001	r(291) =  .70, p <.0001	r(272) =.06, p =.35
Controlling age and sex	r(285) = −.29 p <.0001	r(265) = −.09 p =.16	r(289) = −.30 p <.0001	r(289) = −.23 p <.0001	r(289) = −.48 p <.0001	r(289) = −.33 p <.0001	r(289) =  .64 p <.0001	r(270) =.11 p =.07

Simple linear regressions indicate that small to moderate amounts of variance in Get-Up test are explained by the individual strength, function, and fitness tests (Table 4). Further assessment using a full set of reference tests indicates a sizable amount of variance in the Get-Up test is explained by the predictors (adjusted R² = 0.56) (Table 4). Finally, to understand which predictors were most influential, a stepwise regression including all predictors resulted in a strong simpler model consisting of handgrip strength, SPPB total, and 400-m walk test with the same adjusted R² value, indicating that the Get-Up test is related to strength, function, and fitness (Table 4). Adding relevant predictors (age, sex, smoking status, heavy alcohol consumption, and percentage body fat) did not significantly improve the model fit (Supplemental Table 1). Modest changes in the beta estimates and R² values were observed when the models were run for females and males separately (Supplemental Table 1).

Table 4.

Linear regression of reference tests on Get-Up test

Model	Parameter	β	Standard error	p value	Standard β	Model R2	Adj. model R2
Simple linear regression models
Get-Up test	Biodex peak torque (kg/s)	− 0.02	0.002	<.0001	− 0.41	0.17	0.17
	1 rep. maximum chest/leg press (kg)	− 0.01	0.002	<.0001	− 0.26	0.07	0.07
	Handgrip strength (kg)	− 0.12	0.02	<.0001	− 0.38	0.15	0.14
	SPPB score (0–12)	− 1.66	0.17	<.0001	− 0.49	0.24	0.24
	Usual gait speed (m/s)	− 7.17	0.98	<.0001	− 0.40	0.16	0.15
	400-m walk test (min to complete)	3.14	0.19	<.0001	0.70	0.49	0.49
Multiple linear regression (full model)
Get-Up test	Biodex peak torque (kg/s)	− 0.0002	0.004	0.94	− 0.01	0.57	0.56
	1 rep. maximum chest/leg press (kg)	0.003	0.002	0.22	0.10
	Handgrip strength (kg)	− 0.07	0.02	0.004	− 0.21
	SPPB score (0–12)	− 1.02	0.17	<.0001	− 0.27
	Usual gait speed (m/s)	− 0.77	0.81	0.35	− 0.04
	400-m walk test (min to complete)	2.52	0.26	<.0001	0.52
Multiple linear regression (stepwise variable selection)
Get-Up test	Intercept	9.37	2.78	<.001	0.00	0.56	0.56
	Handgrip strength (kg)	− 0.05	0.01	<.001	− 0.15
	SPPB score (0–12)	− 1.04	0.17	<.0001	− 0.28
	400-m walk test (min to complete)	2.57	0.24	<.0001	0.52

Predictive criterion validity of the Get-Up test

Validity was also explored by examining the association of the Get-Up test with sarcopenia at baseline and follow-up. To establish incidence of sarcopenia, the 33 individuals who were identified as sarcopenic at baseline were removed. Twenty-two new cases of sarcopenia were observed at follow-up (mean follow-up = 0.91 ± 0.1 years, range 0.7–1.4). A tentative cut-point based on the observed cases of sarcopenia at baseline was chosen for the Get-Up test that split the fastest tertile from the slowest two tertiles. As a comparison, the same analysis was conducted using a score of ≤ 9 on the SPPB, which splits high performance from intermediate and low performance [31] Although SPPB score is not typically used to diagnose sarcopenia, it is an alternative measure of muscle function proposed by the EWGSOP. The ability to identify sarcopenic participants was explored by SPPB and Get-Up test performance groups while controlling for age, sex, BMI, heavy alcohol consumption, smoking status, and number of chronic conditions.

Slow Get-Up performance and low SPPB scores were associated with increased prevalence of sarcopenia with an odds ratio (OR) 3.99 (95% CI 1.64, 9.67) and 5.15 (1.40, 18.93) at baseline, respectively, determined by multiple logistic regression. Cox proportional hazard regression explored the incidence of sarcopenia at follow-up and was predicted by slow Get-Up times (hazard ratio [HR] 3.47 (1.10, 10.98)), but not low SPPB score (1.98 (0.31, 12.52). Finally, a single unit increase in Get-Up time was associated with a HR of 1.19 (1.01, 1.40) for developing sarcopenia indicating an increased risk of 19% for each additional second needed to complete the Get-Up test, but not for an increase in SPPB score (0.91 (0.65, 1.26)). These significant associations indicate that the Get-Up test performance may be associated with both prevalence and incidence of sarcopenia.

The sensitivity and specificity ROC analysis indicated that the Get-Up had fair discriminative ability, with an AUC of 0.703 (95% CI 0.62 to 0.79) (see Fig. 3). By maximizing the Youden Index, an optimal cutoff value of 6.75 s on the Get-Up scale was identified. At this threshold, the sensitivity and specificity were 76.7% and 58.2%, respectively, corresponding to a Youden Index of 0.348. These findings suggest that the Get-Up can moderately distinguish between those with and without sarcopenia status at the time of the Get-Up test in our sample.

Fig. 3.

ROC curve and optimal cutoff point for the Get-Up based on the Youden Index

Discussion

The findings of the present study suggest that the Get-Up test is reliable both between trials and across time. Conducting two practice trials and two timed trials was necessary and sufficient to eliminate a learning effect in this population. Therefore, the time recorded for Trial 2 should be used when assessing Get-Up test ability, as it did not differ significantly from the trials conducted on a separate day. The four Get-Up trials demonstrated excellent internal consistency, indicating high reliability across repeated assessments. Compared to common measures of strength, function, and fitness, the Get-Up test shows low to moderate correlations. These individual tests also accounted for small to moderate amounts of variance in the Get-Up test. Combining several common tests to predict the Get-Up test time resulted in an acceptable model fit. Adding additional predictor variables like age, sex, lifestyle and health status did not improve the model fit, which suggests Get-Up performs equivalently across the demographics generalized in this study. Furthermore, Get-Up demonstrated fair discriminative ability on the basis of sensitivity and specificity. Finally, slow Get-Up performance was significantly associated with higher prevalence and incidence sarcopenia, offering face validity and predictive in addition to criterion validity.

The findings of this study support the validity of the Get-Up test in identifying sarcopenia, as slower performance times were associated with significantly higher odds of sarcopenia prevalence and higher hazard of incidence of sarcopenia after 1-year follow-up. These results align with previous research on mobility and functional tests. For instance, the Short Physical Performance Battery (SPPB) has demonstrated its predictive validity in identifying older adults at risk of disability, with low scores (4–6 out of 12) associated with significantly higher risk of disability in ADLs and mobility-related disability [45]. Several other studies have attempted to address earlier identification of sarcopenia in older adults. Strategies include developing a prediction equation for DXA appendicular lean mass but its complexity and small sample (n = 43) limit its practicality [46]. Phu and colleagues identified an SPPB cutoff of < 8 to be moderately predictive for diagnosing sarcopenia (AUC = 0.68) in community-dwelling older adults [47]. Echeverria and colleagues [48] reported that TUG and gait speed showed some predictive capacity but were insufficient for identifying sarcopenia in hospitalized older adults (n = 84), whereas Martinez and colleagues [49] demonstrated that a TUG > 10.85 s had good accuracy (AUC = 0.80) in a similar population (n = 68). In contrast, Filippin and colleagues [50] found a TUG > 7.5 s to be less precise (AUC = 0.66) despite high sensitivity in community-dwelling older adults (n = 211). Lin and colleagues [51] highlighted that the 8-foot Up-and-Go Test (AUC = 0.87) effectively discriminated between sarcopenic and non-sarcopenic individuals in a large sample (n = 745), though no cut-point was provided. Each second longer required to complete the Get-Up test was associated with a higher hazard of developing sarcopenia after adjusting for age, sex, BMI, heavy alcohol consumption, smoking status, and number of chronic conditions. Similarly, in another study on sarcopenia incidence in community-dwelling, elderly Japanese women, each 1-s increase in the Timed Up and Go (TUG) test was associated with higher odds of sarcopenia at a 4-year follow-up [52].

Each of the tests, Get-Up, SPPB, and TUG, require minimal equipment, expertise, and time to conduct. The Get-Up test, however, incorporates a more complex movement pattern that may engage a wider range of muscle groups compared to SPPB and TUG. Further research is warranted to explore whether the unique demands of the Get-Up test make it a more sensitive tool for predicting sarcopenia and related functional impairments in diverse cohorts.

This proposed test is a helpful addition to the existing tests of physical function for older adults due to its simplicity, requiring almost no equipment or expertise to conduct. Its instructions are flexible and allow individuals with a range of old injuries, joint replacement, and medical conditions to complete the test by varying their movement strategy accordingly. Since the outcome is time-based, it is easy to identify even small differences in performance between individuals and groups, as well as changes over time, which avoids measurement artifacts like ceiling and floor effects. It also mimics a natural movement that is a typical challenge for older adults to overcome: getting up off the ground. It requires the individual to conduct a reasonably complex, full-body movement, distinguishing it from other tests like the Timed Up and Go or the Sitting-Rising tests, many of which require a very specific movement pattern. This broader range of motion may complement existing assessments by capturing a wider spectrum of physical abilities.

Importantly, this test is safe and practical for older adults. Using a simple gymnastics-type high-density foam mat provided a stable, cushioned surface for the test. During the more than 2000 repetitions of the Get-Up test conducted in this study, only one rare and minor adverse event was reported (a small skin abrasion), suggesting that the test is generally safe when performed with appropriate precautions.

Strengths of this study include conclusively testing reliability by conducting multiple trials 3 days apart. A wide range of standard strength, function, and fitness measurements provided a thorough and varied set of comparisons to establish the validity of the Get-Up test. These tests were performed in a large group of older adults where differences between function varied and were described rigorously. Finally, the test is fast, flexible, can identify small changes in performance time, and mimics a typical challenge for older adults in rising from the floor that requires complex coordination to accomplish.

Limitations include the lack of racial diversity, high socioeconomic status, and general good health of the sample, which limits the generalizability to the general population. Although this sample characteristic does not invalidate our results, it is important to acknowledge how racial and socioeconomic disparities can influence both the interpretation and applicability of these outcomes. Additionally, despite self-reported good health, the prevalence of sarcopenia (10%) in the sample was typical for a group of community-dwelling older adults [53]. Future studies should aim to recruit more diverse populations—especially individuals from a variety of racial, socioeconomic, and health backgrounds—to enhance broader generalizability and applicability, for example to assess flexibility in movement strategies.

To further establish the Get-Up as a relevant and powerful test of strength, function, and fitness, its predictive ability should be tested with health outcomes and mortality over a longer follow-up period, controlling for or exploring performance along with other important variables associated with aging such as cognitive function and comorbidities. It would also be edifying to explore its relationship to balance and fall risk in greater detail. Further research should establish more substantial norms (see Supplemental Table 2 for normative data from the present sample) and compare Get-Up times of healthy to diseased populations and community-dwelling to institutionalized and hospitalized individuals. As the physical abilities of the individual diminish, for example when comparing community-dwelling and those who are institutionalized or hospitalized, more care should be given to appropriate spotting during the movement for safety and possibly avoiding the test when severe limitations are present, especially in institutionalized and hospitalized populations.

While sarcopenic research has had many successes, which have advanced interventions and pharmacotherapy, fewer strides have been made in the screening and prediction of sarcopenia. Here we tested the efficacy of a new Get-Up test, which is designed to mimic real-world movements that are commonly performed by able-bodied adults. However, it is important to acknowledge that some older adults, particularly those who are frail or have limited mobility, may naturally avoid such movements due to physical discomfort or fear of injury. This highlights the need for further research to determine the feasibility and acceptability of the test among frail individuals. Overall, there is convincing evidence for the reliability and validity of the Get-Up test as a measure of full-body strength, function, and fitness. Finally, a slow Get-Up performance time was associated with higher prevalence and incidence of sarcopenia compared to a fast Get-Up test performance. Based on our results here, we can suggest that moving forward with additional Get-Up test research is warranted as we progress towards deployment of the test in clinical settings.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

D.C. Lee and N.F. Meier designed and conducted the research. N.F. Meier and B.S. Klinedinst performed the statistical analysis. All authors contributed to manuscript writing and revision, reviewed the final manuscript, and approved its submission.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the ethical principles outlined in the Belmont Report and the regulations set forth in the Common Rule (45 CFR 46) for the protection of human research participants. Ethical approval for this study was obtained from the Iowa State University Institutional Review Board (IRB) (Approval Number 15–430). Written informed consent was obtained from all participants prior to their participation in the study.

Consent for publication

All authors have reviewed and approved the final manuscript and consent to its publication.

Conflict of Interest

The authors declare no competing interests.