The modified Borg scale was used to measure dyspnea and ULs RPE at rest and during each minute of the UULEX-M test. This valid and reliable scale ranged from 0 (no effort) to 10 (maximum effort); higher scores indicated increased RPE.22

The spirometer (Spirobank II-MIR) was used to characterize the sample. The test followed the American Thoracic Society recommendations and used reference values established by Pereira et al. (2002).32 The Jamar dynamometer (North Coast Medical) measured handgrip strength in kilograms. The individual was seated in a chair without UL support, with the spine erect, knees flexed at 90°, shoulders adducted and neutrally rotated, elbows flexed at 90°, forearms half-pronated, and wrists neutrally positioned and slightly extended. The tested UL was suspended, and the examiner supported the hand placed on the dynamometer. Three measurements were performed in both ULs, with a 60-second rest between them. The mean measurements of each UL were registered.33

The Post-COVID-19 Functional Status (PCFS) scale.34 was used to classify functional status based on individuals' post-COVID-19 syndrome history and reports. They were ranked as “no functional limitations, "negligible functional limitations", "slight functional limitations", "moderate functional limitations", or "severe functional limitations". This scale considered the ability to conduct activities at home, symptoms of pain, anxiety, and depression, and self-management of activities.33

Individuals attended one data collection session. Demographic data (e.g., height, weight, and age) were recorded. After that, spirometry and handgrip strength tests were performed, and the PCFS was completed. Then, individuals performed one UULEX test and one UULEX-M in a randomized order, in blocks, using a computer program (https://www.random.org/). SpO2, HR, BP, dyspnea, and ULs RPE were assessed before and immediately at the end of each test.33 Individuals rested 30 min between tests or until baseline recovery, consistent with the UULEX normative study.33 The time to complete the tests was recorded.

Sample size was initially estimated from a pilot study with 10 participants, using test time as the main variable (2% for SpO₂, 7 beats per minute for HR, and 2.07 min for test time), yielding a minimum sample size of 36. However, since a shorter time was expected with the modified protocol, equivalence-based calculations were performed using physiological variables (SpO₂ and HR), which confirmed the adequacy of the protocol. In addition, a complementary calculation for the primary validation outcome, assuming r = 0.75 (α = 0.05, β = 0.20), indicated that 29 participants would be sufficient, consistent with Portney (2020, Chapter 29, p. 432).35 Thus, the final sample of 64 participants exceeded all requirements. According to the COSMIN checklist, this corresponds to a good sample size for methodological studies.

Data were presented as mean and standard deviation unless otherwise stated, and the Shapiro-Wilk test verified data normality. Comparisons between sexes for demographic, spirometric, PCFS classification, handgrip strength, and clinical variables were performed using Student's t-test, depending on the distribution of the characteristic and variable. t-tests or the Wilcoxon test were used to compare test times and physiological responses between UULEX and UULEX-M, depending on data normality. According to data normality, the Pearson or Spearman correlation coefficient was used to verify the correlation between tests and handgrip strength; correlations were classified as minimal or absent (<0.20), weak (from 0.25 to 0.50), moderate (from 0.50 to 0.75), or strong (from 0.75 to 1.00).36 Confidence intervals for correlation coefficients were calculated using Fisher’s Z transformation. In addition, a two-way repeated measures ANOVA was conducted to test the interaction between sex and protocol (UULEX vs. UULEX-M). Partial eta squared (η²p) was reported as a measure of effect size. The level of significance was set at 5%, and the SPSS version 19.0 (Chicago, IL, USA) was used for analyses.

To our knowledge, this is the first study to validate a UL test in individuals with post-COVID-19 syndrome. The results support that: 1) UULEX-M demonstrated performance similar to the original UULEX in individuals with post-COVID-19 syndrome; 2) the UULEX-M test provoked the same physiological responses and symptoms as UULEX, except for HR; 3) the time to perform the UULEX-M test was significantly shorter than in UULEX for the total sample and both sexes; and 4) men performed better on both tests than women.

The initial 0.2-kg load proposed by UULEX is compatible with very light ADLs, such as brushing teeth and combing hair, whereas the final 2-kg load approximates heavier tasks, including carrying groceries or lifting objects onto shelves. Takahashi et al. (2003) demonstrated that individuals completed an average UULEX test time of 7.95 ± 2 min, with most participants stopping at the initial 0.2-kg load. The warm-up time and late load progression (only after level eight) may explain these findings and limit the original UULEX's ability to represent ADLs. In contrast, UULEX-M eliminated the warm-up and applied a constant 1-kg load (five times the original starting point), which may better simulate ADLs that typically involve greater effort. Although the load did not increase, the constant load reflects endurance, as individuals performed repeated movements until fatigue. Accordingly, the moderate to strong correlations observed between the two protocols were supported by the scatter plot analysis, which showed that, despite differences in absolute duration, both tests exhibited a consistent linear relationship, further supporting concurrent validity.

The Bland–Altman analysis showed that UULEX tended to present longer test times than UULEX-M (bias = 3.97 min). This difference was expected because UULEX-M used a constant heavier load (1 kg) and removed the warm-up phase. These modifications aimed to simplify the protocol and simulate ADLs, while also reducing the test's total time. Importantly, this study was planned as an equivalence study focused on physiological responses (SpO₂, HR, and symptoms), and not on absolute time. For this reason, the bias observed does not compromise the validity of the UULEX-M. The limits of agreement (0.54 to 8.48 min) were narrow and clinically acceptable, supporting the feasibility of UULEX-M as an alternative to UULEX in post-COVID-19 individuals.

Both protocols were primarily limited by ULs fatigue, as shown by RPE consistently higher than dyspnea. Similar findings were reported by Takahashi et al. (2003) in individuals with COPD performing UULEX. In the present study, men and women reported comparable fatigue levels (Borg ≈ 7), indicating that despite differences in absolute test duration, both groups reached a similar subjective limit of exertion. This supports UULEX-M's ability to capture functional limitations consistently across sexes.

Sex-related differences in test duration are consistent with normative data for UULEX, where men typically perform longer than women.21,33 Physiological differences such as body composition, muscle mass distribution, and aerobic/anaerobic capacity.37–39 likely explain this pattern. At the same time, the balanced distribution of men and women in our sample strengthens the interpretation of these analyses. Beyond confirming overall sex differences in test performance, the interaction analysis revealed that the gap between men and women was smaller in UULEX-M than in the original UULEX. This suggests that the modified test may attenuate sex-related disparities, potentially providing a more equitable assessment of ULs capacity. Such findings reinforce the clinical applicability of UULEX-M. The weak correlation between handgrip strength and test time reinforces that both UULEX and UULEX-M assess broader aspects of ULs endurance rather than isolated strength.39,40 Together, these findings corroborate normative and post-COVID-19 evidence.14,15,17,20 and support the validity of UULEX-M across sexes.

A practical advantage of UULEX-M is that it requires no specific equipment, unlike the original UULEX, which uses a board and weighted sticks. In UULEX-M, levels are determined anatomically, and the load can be reproduced with simple objects available in clinical or home environments. A 1-kg load was adopted to better reflect the functional demands of daily activities. While the original UULEX used 0.2 kg, this does not reflect real-life tasks. In contrast, 1-kg approximate common ADLs include carrying rice packages, cooking oil bottles, and others.2,24,41 Moreover, standardized ADLs employ even heavier objects, such as the Grocery Shelving Test (∼420 g) and the Gallon Jug Shelf Test (∼3.8 kg). In addition, studies in COPD populations have shown that arm activities elicit significant ventilatory and metabolic responses.25,30 and that resistance training with progressive loads up to 2-kg is safe and effective.4,42 Taken together, these considerations support the use of 1-kg weights as a functionally meaningful modification, which, although reducing test time, preserved physiological equivalence and strengthened the concurrent validity of UULEX-M.

This study has limitations. Test–retest reliability and measurement error indices (SEM, MDC) were not evaluated, and direct ventilatory and metabolic measures were not obtained. In addition, correlations with ADLs or participation questionnaires were not assessed. Future studies should address these aspects to strengthen evidence on reproducibility, interpretability, and clinical applicability.

Finally, the study was designed to validate the UULEX-M protocol. The choice of a post-COVID-19 population was informed by the pandemic context, which presented a unique opportunity to evaluate individuals with persistent pulmonary and muscular impairments and limited evidence on UL performance. While our findings support the validity of UULEX-M in this population, external validation in other chronic cardiopulmonary conditions is needed before broader generalization.

The UULEX-M test demonstrated moderate to strong correlations with UULEX, with comparable physiological and symptomatic responses. Test time was significantly shorter in UULEX-M, both in the total sample and across sexes, without compromising equivalence in physiological outcomes. These findings support the validity of UULEX-M as an alternative test for assessing ULs exercise capacity in individuals with post-COVID-19 syndrome. Nevertheless, further studies using gold-standard methods and including other chronic cardiopulmonary populations are needed to confirm and expand these results.