Search tips
Search criteria 


Logo of jmniJournal of Musculoskeletal & Neuronal Interactions
J Musculoskelet Neuronal Interact. 2015 September; 15(3): 257–263.
PMCID: PMC5601238

Reliability of parameters during stair ascent measured with Leonardo Mechanograph® Stair A in healthy subjects



Stair climbing (SC) as daily activity is assessed with different SC-tests, but none directly measures ground reaction force over several steps. The Leonardo Mechanograph Stair A has five steps and four force sensors. This study aimed at investigating the reliability of the Stair A test for force, power and time to SC.


55 healthy participants (age: 48±14 years) were five times tested during SC with self-chosen and fast speed. 30 participants were examined for test-retest-reliability, calculated with the intraclass correlation coefficient (ICC). The variability was examined with the coefficient of variation (CV). To determine potential associations between SC and jumping performance or daily activity, squat and countermovement jumps were additionally performed and the International physical activity questionnaire (IPAQ) was completed.


The inter-visit ICCs of self-chosen and fast SC were good to excellent 0.63-0.77. The intra-visit ICCs were excellent after three trials (0.78-0.88). The CVs for SC with self-chosen speed were lower (2.1-6.6%) than those for fast SC (4.9-10.8%). There were no significant correlations between SC and jump parameters and only moderate correlations with the IPAQ.


The Stair A is a reliable tool for the assessment of SC.

Keywords: Stair Climbing Test, Force, Power, Activity


Stair climbing (SC) is a daily activity with an adaptable movement pattern, which produces greater ground reaction forces than walking and requires more energy[1]. The centre of gravity translates vertically, which is more challenging compared to walking[2]. The postural stability needs to be guaranteed in order to be able to SC, especially for the single stance phase to prevent falls[2]. It is known that many stair accidents occur, especially among elderly or frail people[3]. Not only the risk of falls is increased in elderly and frail people but also the ability for SC might get crucial due to reduced force. It is thereby well known that increasing age leads to decreased physical activity, as well as to sensory and cognitive constraints[4,5]. Altogether, it has been shown that age can affect the vertical ground reaction force, especially during SC[6]. In addition, several medical diseases are related to reduced SC performance[7]. The inability for SC is a large limitation in everyday life that reduces independency. Measurement of ground reaction force during SC with SC tests (SCT) offers the option to investigate the actual condition of an individual and leafs room for an intervention if necessary. In this regard, SCT are widely used in patients with cardiopulmonary[8,10] and orthopaedic diseases[11] to test their skeletal muscle (dys-) function, which is often reduced, within a daily activity.

At the moment three different SCT exist. First, participants climb the steps of a normal stair and time is stopped manually[8,12-14]. Second, participants climb a stair construction (three steps) built on two force plates that are integrated into the floor[15]. Third, participants climb a staircase with an integrated force plate on single steps only[6,11,16,17]. The latter measurement tool is a construction of six steps, with an integrated force plate on the third and the fourth step. Studies have shown that the curve of the vertical ground reaction force during a single step stair ascent is similar to the one of a gait cycle. These conventional SCT suffer from several weaknesses. Manual stoppage of time during SC lacks scientific accuracy and suffers from a large inter-tester variation. The other two SCT assess force, power and time but the assessment is limited to maximal two steps. Hence, the assessment of force and power is only possible for one step per leg. The progression of force and power during SC, which is mostly composed of more than two steps, are not possible.

A novel measurement device to determine force and power during SC is the Leonardo Mechanograph Stair A (Stair A). It works with four force sensors underneath the steps at the whole stair. Hence, this new instrument measures not only a single step, but also the performance during the whole SC with its five steps. The Stair A offers a promising option to investigate force and power values during SC. The measurement of force during SC is especially important in elderly and frail people as well as patients. In these population groups, muscle force might be reduced to such an extent that it is simply not possible to climb a stair. Timed SCT are questionable in these cohorts because these people are not interested in SC as fast as possible. They are rather interested in SC per se. Hence, it is important for these people to maintain a certain level of force to allow SC and maintain thereby their independency during everyday life. Repeated measurement of ground reaction force during SC represents a promising option to investigate the progression of muscle force during an important daily task. However, up to date, the reliability of the Stair A was not investigated so far.


The aim of the current study was to assess (absolute and relative) reliability of the Stair A in healthy participants. In addition, we aimed investigating whether the results of the Stair A differ as a function of age. Furthermore, to determine potential associations between SC and jumping performance or daily activity, the Stair A derived values were compared to the results of two-legged jumping manoeuvers and to the results of a questionnaire on daily physical activity.



Test-retest reliability was explored by executing the identical protocol of the SCT on two occasions within two to seven days. The study protocol was approved by the local Ethics Committee. The primary endpoint was the reliability of the maximal SC force relative to bodyweight (Fmax,rel) with self-chosen speed. Exploratory endpoints were the reliability of the total time (ttot), the relative maximal power (Pmax,rel) and the Fmax,rel during SC with self-chosen and fast speed. Two-legged jumping manoeuvres and a physical activity questionnaire were completed to determine potential associations of these results with SC values.


Fifty-five subjects (Table 1) were recruited by placard in the University Hospital of Zurich. Having read the information leaflet, the participants signed the written informed consent form before being included in the study. Out of the fifty-five subjects, fifteen males and fifteen females, aged forty to fifty-nine years, were included to assess test-retest reliability analogously to Walter et al. (1998)[18]. Fourteen participants aged twenty to thirty-nine and eleven subjects aged sixty years or older were examined to investigate the influence of age during SC.

Table 1
Baseline characteristics of the participants (n=55).


Measurements were accomplished using the certified Stair A (adult version; Novotec Medical, Pforzheim, Germany). The staircase has a step height of 0.175 m and a step depth of 0.28 m. To guarantee the safety of the participants there is a handrail on both sides of the Stair A. Four integrated sensors enable measuring of force during SC. The manufacturer’s software calculates the centre of mass related variables velocity and power. The sampling rate per sensor is 1000 Hz with a maximum force per sensor of 1.3 kN, using a 16 bit analogue-digital-converter. For analysis, the software Leonardo Mechanography v4.3 RES (Novotec Medical, Pforzheim, Germany) was used. The force platform was adjusted to zero (out of -0.03 kg and +0.03 kg) before any subject had stepped on the staircase.


Stair climbing test

According to previous studies with Leonardo Mechanography, as well as other studies testing other kinds of stairs, the participants SC five times[16,17,19,20]. Participants were asked to SC five times with their self-chosen speed, since the variability of gait parameters is lowest with habitual speed[17], and five times as fast as possible without running. The participants had to wear anti-slippery socks provided by the researcher. At the beginning there was one sample measurement which was not included in the statistical calculations to reduce a possible familiarisation effect. Stair descent was not assessed due to its higher variability[17]. We included all five trials of the participants in the calculations.

International physical activity questionnaire

In addition to the measurement with the Stair A, all subjects received a questionnaire about their regular physical activity. For this assessment the International Physical Activity Questionnaire (IPAQ) was chosen, including several questions about daily activities such as sports and work[21].

Two-legged jumping manoeuvres

For the determination of jumping force and power during countermovement jumps (CMJ) and squat jumps (SJ), five vertical jumps (separated by 30 s of rest) were performed on a Leonardo Mechanograph force plate (Novotec Medical, Pforzheim, Germany). For the measurements, storage and analysis of data, we used the manufacturer’s software (Leonardo Mechanography GRFP version 4.2, Novotec Medical, Pforzheim, Germany). CMJs were performed with freely moving arms, whereas the SJs were performed with the hands resting on the waist.


Relative reliability (intra- and inter session) was calculated using the intraclass correlation coefficient (ICC) to explain the measurement error relative to total variance[22]. For the ICC, the consistency, two-way random mixed model was used. We calculated the ICC over two, three, four and five trials in one session. Values >0.75 are considered to be excellent, 0.6-0.74 good, 0.4-0.59 moderate and <0.4 poor[23]. Absolute reliability was expressed by the coefficient of variation (CV), calculated with the formula:

An external file that holds a picture, illustration, etc.
Object name is JMNI-15-257-g001.jpg

where SDj represents standard deviation (calculated for each participant of all five trials), and [x with macron]j is the mean of all xij [24]. To measure the within-subject standard deviation the standard error of measurement (SEM) was calculated[22,25]:

An external file that holds a picture, illustration, etc.
Object name is JMNI-15-257-g002.jpg

The smallest detectable change (SDC) of the measurement was calculated as[26]:

An external file that holds a picture, illustration, etc.
Object name is JMNI-15-257-g003.jpg

To compare the Stair A results with the results of the jumping manoeuvres and the IPAQ scores (ordinal scale) with the Pearson correlations (r) and the Spearman correlations (rs), were calculated, repsectively.

Correlations (Spearman and Pearson) of 0-0.25 were considered to be little, 0.25-0.50 fair, 0.50-0.75 moderate to good and >0.75 very good to excellent[27]. To check for associations between age as well as physical activity measured by the IPAQ and the results of SC one-way ANOVAs were conducted. If the statistical preconditions (test of homogeneity and normal distribution) were not complied, the non-parametric Kruskal-Wallis-Test was used to analyse the variances by ranks.

To check whether there was potential learning or fatigue effect, a paired t-test or a Wilcoxon test (if parameters were not normally distributed) was calculated between trial 1 and 5 of test day 1, trial 1 of the first and the second test day, as well as trial 1 and 5 of test day 2, respectively.

Statistical calculations were performed with SPSS Statistics 21.0 (SPSS, Chicago, IL, USA). The level of significance was set at p≤0.05. Data are given in means ± SD.



The SEM and SDC had similar values in self-chosen and fast SC ttot and Fmax,rel (Table 2). The SEM and SDC of Pmax,rel in fast SC were higher than in self-chosen SC (Table 2).

Table 2
Test-retest results, Standard error of measurement (SEM) and smallest detectable change (SDC) for stair climbing at self-chosen and fast speed.

Test-retest reliability

The values of the two different test days for self-chosen and fast SC are shown in Table 3. The ICC (2,1) for all stair variables (Table 3) assessed on two different days and within one test day ranged from 0.63 to 0.77 and from 0.70 to 0.89, respectively.

Table 3
Intraclass correlation coefficient (ICC (2,1)) and the coefficient of variation (CV) for Inter- and Intra-visit reliability.

The ICCs of the first session calculated for two, three, four, and five trials were similar (Table 4). For most variables, the highest ICC was achieved with two or three trials.

Table 4
Intraclass correlation coefficient (ICC (2,1)) and 95% confidence interval (CI) of the first session over two, three, four and five trial.


The CV, overall stair parameters, ranged from 2.1% to 10.8% (Table 3). The mean CVs for the self-chosen SC (2.1-6.6%) were lower than those for fast SC (4.9-10.8%).

Learning and/or fatigue effect

There was a significant difference in the parameter ttot at self-chosen speed between trial 1 and 5 in the first session (4.27 vs. 4.18 for the first vs. fifth trial, p=0.013). There was also a significant difference in the parameter Pmax,rel at self-chosen speed between trial 1 and 5 in the second session (6.00 vs. 5.82 W·kg-1 body mass for the first vs. fifth trial, p=0.045). For most variables, the highest ICC was achieved with two or three trials.

Daily activity and jumping performance

There was a correlation between total IPAQ score and Pmax,rel at self-chosen speed (y=0.266·x+20.6, rs=0.266, p=0.050). The active transportation part of the IPAQ correlated positively with Fmax,rel at self-chosen speed(y=0.300·x+19.6, rs=0.300, p=0.026) and with Pmax,rel at self-chosen speed (y=0.424·x+16.1, rs=0.424, p=0.001).

There was no significant correlation of the stair parameter means compared with the means of the variables determined during SJ or CMJ (r=-0.321-0.156, p>0.05).

Age differences

Age significantly influenced all stair parameters during fast SC and ttot during SC at self-chosen (Table 5). In particular, elderly participants (≥60 years) showed significant differences in ttot at self-chosen speed, fast ttot and fast Pmax,rel compared to the younger two groups. Fast Fmax,rel and Pmax,rel were also significantly different comparing the youngest group (20-39 years) and the middle aged group (40-59 years).

Table 5
Stair parameters of different age groups.

Sex did not have a significant influence on the measured stair parameters (data not shown).


In the current study we showed a moderate to high reliability of the Stair A during SC in all of the assessed parameters. The primary endpoint, i.e. the reliability of the SC Fmax,rel, showed an excellent result at self-chosen speed. The ICCs (2,1) for Fmax,rel, Pmax,rel and ttot on two separate testing days were good to excellent for SC at self-chosen and good for fast SC. For all measured variables, an excellent reliability was already achieved after two or three trials within one session.

The ICCs (2,1) over one test day were better than over two test days. With respect to our ICCs (2,1) for Fmax,rel during SC at self-chosen speed, the ICCs of 0.65-0.84 measured with a force plate on two single steps described by Leitner et al. (2011)16 are very similar. In addition, the ICC in ttot at self-chosen during SC, which was 0.8-0.89 in their study, was comparable to our result[16]. However, in comparison with a study in chronic obstructive pulmonary disease (COPD) patients, in which participants had to ascend ten stairs as quickly and safely as possible[8], our correlations are lower. In the mentioned study, time was stopped manually and force as well as power, were then calculated. Patients with COPD showed lower power in the test than matched healthy people. The test-retest reliability was very high with an ICC (2,1) of 0.908. LeBrasseur et al. (2008) calculated higher ICCs of fast SC time (0.79-0.94), tested on 12 steps in different age categories of men, including also elderly men with mobility limitations, compared to the ICC (2,1) in fast SC ttot of our study (0.65-0.81)[28]. Overall the test-retest-reliability of the Stair A tended to be similar or only slightly lower than in previous studies[8,16,28]. Taken together, we can conclude that the current study is in line with previous SC assessments. Differences in ICC between the present study and previous studies can be explained by the determination methods (ground reaction force vs. manual hand stoppage of time) or the participant cohort (healthy participants vs. patients).

The individual calculations for the ICC (2,1) for two, three, four and five trials of the first session demonstrated a good repeatability already after two and three trials, which were even better than after five trials. Moreover, the intra-visit ICCs for all variables were already excellent after three trials in the first session (0.78-0.88). These results indicate that two trials would be sufficient to receive reliable measurement results with the Stair A. In addition to the increase in time efficiency, this result would also have an important practical benefit. The Stair A will mainly be applied in patients with a reduced exercise capacity. In these population groups, a fatigue effect might occur already after few trials, which would negatively influence test outcomes. Such a fatigue effect manifested itself already in our participants with a reduced Pmax,rel in trial 5 compared to trial 1 of the second testing day. Hence, especially in patients fewer trials represent a large benefit. However, one result might impair our proposal. For ttot at self-chosen speed during testing day 1, there was a significant better result for trial 5 compared to trial [1]. Therefore, the sample measurement at the beginning of the testing procedure, which was not included in the analysis, seemed not sufficient to eliminate the potential learning/familiarisation effect for all assessed variables. Based on our data, we suggest therefore that two familiarisation trials should be conducted in advance of two measurement trials to achieve reliable results with optimal time efficiency.

The CVs were higher during fast SC compared to SC at self-chosen speed. This can be explained by the higher variability for the assessment of ground reaction force during fast SC compared to SC at self chosen[29]. Considering the fact that Stacoff et al. (2005) reported CVs of 5-10% for ground reaction force during stair ascent, our results are promising[6]. Our CVs for SC at self-chosen are also lower than the CVs of Leitner et al. (2011) who calculated CVs of 2.52-3.87% for force variables, and 5.72-11.82% for time variables during SC[16]. The CV of the SC ttot at self-chosen speed is similar to Galvao and Taaffe (2005), who calculated a CV of 4.8% in community-dwelling older people tested on a flight of stairs consisting of eleven steps[30].

To decrypt a potential association between SC and muscle function or daily physical activity, SC parameters were comapred with Fmax,rel, Pmax,rel and height of two different two-legged jumping manoeuvres or the IPAQ. No correlations were present between the variables of the SCT (Fmax,rel, Pmax,rel and ttot) and the variables of the two-legged jumping manoeuvers and only minor correlations with the scores of the IPAQ. These results stand in contrast to previously published data. Bean et al. (2007) showed a significant correlation between SC power and leg press power (r=0.47-0.52) in community-dwelling older adults with mobility limitations[13]. In another study, SC time correlated (r>0.6) with one repetition maximum and total strength of leg press, leg curl, and leg extension, respectively, but not with leg press endurance in older adults (≥60 years old)[31]. In this study, the SCT consisted of climbing 23 steps without using handrails, while time was recorded with a stopwatch[31]. In addition, Roig et al. (2010) showed a moderate correlation between the SCT and eccentric and isometric knee extension torque as well as for eccentric knee flexion torque in COPD patients[8]. Furthermore, SC power in COPD patients had a good correlation (r=0.68) with the 6 minute walking distance (6 MWD). A possible explanation for the discrepancy between our result and these previously published results might be the fact that we investigated healthy people and the other studies were on elderly people or patients. This difference in participant characteristic leads to at least two explanations. First, a ceiling effect might occur especially in measurements of activities in daily life[23]. In particular, we assume that patients or elderly participants approach their exercise limit already during fast SC, while this is not the case in healthy participants. Second, a potential explanation is the fact that healthy people jump as high as possible during two-legged jumping, while during SC, they only use the minimal force necessary to successfully climb the stair. This explanation is supported by the lower variances in Fmax,rel during SC (1.27-1.57 N·kg body mass-1) compared to Fmax,rel during CMJ (18.5-30.2 N·kg body mass-1).

The oldest age group needed significantly more time to SC as fast as possible, than the younger groups. This is in line with a previous study, which demonstrated that age influences SC especially between young (34 years) to middle-aged (64 years) people, as well as between young people and old (77 years) people. In contrast, there was no difference between middle-aged and older subjects[6]. The clinical relevance of this result is presented by Oh-Park et al. (2011) who stated that people who are slower in SC show a higher number of medical diseases, depressive symptoms and disability scores and lower gait velocity (in women those with fear of falling)7.


The SCT on the Stair A is reliable, especially during SC with self-chosen speed. Reliable results can be achieved with two familiarisation trials followed by two measurement trials. The Stair A has only a small association with parts of the IPAQ survey, but not with maximal values of two-legged jumping manoeuvres. Elderly (≥60 years) need significantly more time to SC than younger subjects.


Edited by: F. Rauch


1. Costigan PA, Deluzio KJ, Wyss UP. Knee and hip kinetics during normal stair climbing. Gait Posture. 2002;16:31–7. [PubMed]
2. Trew M, Everett T. Human Movement. 6th ed. Edinburgh London New York Oxford Philadelphia St. Louis Sydney Toronto: Churchill Livingstone Elsevier; 2010.
3. Archea JC. Environmental factors associated with stair accidents by the elderly. Clin Geriatr Med. 1985;1:555–69. [PubMed]
4. Kowal P, Naidoo N, Chatterji S. WHO Study on global AGEing and adult health (SAGE):Health status, chronic conditions and health care utilization in respondents aged 50 and older from 70 countries in the World Health Survey/SAGE Wave 0. SAGE. 2012;1
5. WHO. WHO global report on falls prevention in older age. France: World Helath Organization; 2007.
6. Stacoff A, Diezi C, Luder G, et al. Ground reaction forces on stairs:effects of stair inclination and age. Gait Posture. 2005;21:24–38. [PubMed]
7. Oh-Park M, Wang C, Verghese J. Stair negotiation time in community-dwelling older adults:normative values and association with functional decline. Arch Phys Med Rehabil. 2011;92:2006–11. [PMC free article] [PubMed]
8. Roig M, Eng JJ, MacIntyre DL, et al. Associations of the Stair Climb Power Test with muscle strength and functional performance in people with chronic obstructive pulmonary disease:a cross-sectional study. Phys Ther. 2010;90:1774–82. [PubMed]
9. Rejeski WJ, Foley KO, Woodard CM, et al. Evaluating and understanding performance testing in COPD patients. J Cardiopulm Rehabil. 2000;20:79–88. [PubMed]
10. Jairath N, Chapman JS, Fuller-Bey G. Hemodynamic responses and activity tolerance to stair climbing during the second week post-myocardial infarction. Can J Cardiovas Nurs. 1992;2:9–14. [PubMed]
11. Stacoff A, Kramers-de Quervain IA, Luder G, et al. Ground reaction forces on stairs. Part II:knee implant patients versus normals. Gait Posture. 2007;26:48–58. [PubMed]
12. Bennell K, Dobson F, Hinman R. Measures of physical performance assessments:Self-Paced Walk Test (SPWT), Stair Climb Test (SCT), Six-Minute Walk Test (6MWT), Chair Stand Test (CST), Timed Up &Go (TUG), Sock Test, Lift and Carry Test (LCT), and Car Task. Arthritis Care Res (Hoboken) 2011;63(Suppl 11):S350–70. [PubMed]
13. Bean JF, Kiely DK, LaRose S, et al. Is stair climb power a clinically relevant measure of leg power impairments in at-risk older adults? Arch Phys Med Rehabil. 2007;88:604–9. [PubMed]
14. Cataneo DC, Kobayasi S, Carvalho LR, et al. Accuracy of six minute walk test, stair test and spirometry using maximal oxygen uptake as gold standard. Acta Cir Bras. 2010;25:194–200. [PubMed]
15. Della Croce U, Bonato P. A novel design for an instrumented stairway. J Biomech. 2007;40:702–4. [PubMed]
16. Leitner M, Schmid S, Hilfiker R, et al. Test-retest reliability of vertical ground reaction forces during stair climbing in the elderly population. Gait Posture. 2011;34:421–5. [PubMed]
17. Luder G, Baumann T, Jost C, et al. Variability of ground reaction forces in healthy subjects during stair climbing. Physioscience. 2007;3:181–87.
18. Walter SD, Eliasziw M, Donner A. Sample size and optimal designs for reliability studies. Stat Med. 1998;17:101–10. [PubMed]
19. Cormack SJ, Newton RU, McGuigan MR, et al. Reliability of measures obtained during single and repeated countermovement jumps. Int J Sports Physiol Perform. 2008;3:131–44. [PubMed]
20. Fricke O, Witzel C, Schickendantz S, et al. Mechanographic characteristics of adolescents and young adults with congenital heart disease. Eur J Pediatr. 2008;167:331–6. [PubMed]
21. Patterson E. International Physical Activity Questionnaire. Secondary International Physical Activity Questionnaire. 2002. Available from:URL .
22. Atkinson G, Nevill AM. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med. 1998;26:217–38. [PubMed]
23. McDowell I. Measuring Health:A Guide to Rating Scales and Questionnaires. 3rd ed. New York;Oxford: Oxford University Press; 2006.
24. Gluer CC, Blake G, Lu Y, et al. Accurate assessment of precision errors:how to measure the reproducibility of bone densitometry techniques. Osteoporos Int. 1995;5:262–70. [PubMed]
25. Hopkins WG. Measures of reliability in sports medicine and science. Sports Med. 2000;30:1–15. [PubMed]
26. de Vet HC, Terwee CB, Knol DL, et al. When to use agreement versus reliability measures. J Clin Epidemiol. 2006;59:1033–9. [PubMed]
27. Dawson-Saunders B, Trapp RG. Basic & Clinical Biostatistics. 2nd ed. Appleton & Lange; 1994.
28. LeBrasseur NK, Bhasin S, Miciek R, et al. Tests of muscle strength and physical function:reliability and discrimination of performance in younger and older men and older men with mobility limitations. J Am Geriatr Soc. 2008;56:2118–23. [PMC free article] [PubMed]
29. Masani K, Kouzaki M, Fukunaga T. Variability of ground reaction forces during treadmill walking. J Appl Physiol (1985) 2002;92:1885–90. [PubMed]
30. Galvao DA, Taaffe DR. Resistance exercise dosage in older adults:single- versus multiset effects on physical performance and body composition. J Am Geriatr Soc. 2005;53:2090–7. [PubMed]
31. Vincent KR, Braith RW, Feldman RA, et al. Resistance exercise and physical performance in adults aged 60 to 83. J Am Geriatr Soc. 2002;50:1100–7. [PubMed]

Articles from Journal of Musculoskeletal & Neuronal Interactions are provided here courtesy of International Society of Musculoskeletal and Neuronal Interactions (ISMNI)