Thirty adults with SLE who were willing to take part in the study and were able to give written informed consent were recruited for this criterion validity study. All patients had to fulfill the American College of Rheumatology 1997 revised criteria for SLE (9). Patients were excluded if they had severe cardiopulmonary pathology, history of ischemic heart disease, severe myositis, active nephritis, active neurologic disease, or pregnancy. Clearance for participation in the study was given by consultant rheumatologists from the rheumatology departments in Ysbyty Gwynedd, Ysbyty Glan Clwyd, and Maelor Wrexham Hospital. The study was approved by the North West Wales Local Research Ethics Committee.
Exercise tests and procedures.
All participants attended 2 testing sessions in the physiology laboratory of the School of Sport, Health and Exercise Sciences, University of Wales-Bangor. These 2 visits were separated by 1–2 weeks and were scheduled for the same time of the day. Participants were instructed to avoid any strenuous activity for 24 hours prior to testing and to avoid a heavy meal, caffeine, or nicotine within 3 hours of testing. In the first visit, patients' height and weight were measured using standard equipment and procedures and body mass index was calculated as weight in kg/height in m2. Hemoglobin concentration was also measured (B-hemoglobin; Hemocue AB, Ängelholm, Sweden). After this evaluation, patients performed the SST described in detail elsewhere (8). Briefly, patients were required to step up and down a portable 10-inch bench for 3 minutes at a rate of 17 steps per minute, which was kept constant with the help of a metronome. HR was monitored continuously during the test by telemetry (model S810; Polar, Kempele, Finland) and was recorded at the end of the stage. If target HR (65% of the predicted [220 minus age] maximum HR) was reached, the protocol was terminated. Otherwise, a second stage (26 steps per minute) and, eventually, a third stage (34 steps per minute) were completed with 1 minute of rest between each stage. The steady-state absolute VO2 at each stage was calculated according to the following equations (8):
Then, each patient's uncorrected VO2max was estimated using the mathematical version of the original Astrand-Ryhming nomogram (10) developed by Steven Siconolfi (personal communication):
where %VO2max is calculated as (0.769 × stage HR) – 48.5 for men and (0.667 × stage HR) – 42 for women. Alternatively, the uncorrected VO2max can be predicted using the original graphic version of the Astrand-Ryhming nomogram (10).
Finally, the uncorrected VO2max calculated from the above equations was entered in the following equations developed by Siconolfi et al (8) to predict VO2max according to age and sex:
At the end of this first testing session, patients were familiarized with the maximal exercise test to be performed in the second visit.
On the second testing session, participants repeated the SST and, after a minimum of 30 minutes of rest, performed an incremental exercise test on a cycle ergometer (874E, Monark, Vansbro, Sweden) keeping a constant pedaling rate of 50 revolutions per minute. This protocol consisted of 2 minutes with no resistance followed by increments of 25W every 2 minutes until volitional exhaustion. Throughout the test, expired gases and flow were analyzed breath by breath using an automated system (600Ergo Test; ZAN Messgeräte, Oberthulba, Germany) from which VO2 and carbon dioxide production (VCO2) were calculated and averaged every 30 seconds. The highest VO2 recorded during the test (our criterion measure) was considered to be maximal when the criteria used in the original validation study by Siconolfi et al (8) were met. These criteria were 1) a plateau defined as a difference <0.25 liters/minute between the 2 final VO2 measurements obtained every 30 seconds, and 2) a respiratory quotient (VCO2/VO2) for the last measurement ≥1.0.
When describing patient characteristics, VO2max values are reported relative to body mass (ml/kg/minute) because this is the measure of physical fitness commonly used for comparative purposes. For validity and reliability analyses, however, we used absolute VO2max values (liters/minute) because these are the values directly measured by the automated breath-by-breath analysis system and predicted by the equations developed by Siconolfi et al (8).
Data are reported as the mean ± SD for normally distributed variables; otherwise, data are reported as the median (interquartile range). The criterion validity of VO2max predicted from the SST was assessed using the 95% limits of agreement (LOA) by Bland and Altman (12), paired t-test, and intraclass correlation coefficient (ICC) against VO2max measured during the cycling test. Standard error of the estimate (SEE) was also calculated. Reliability of predicted VO2max was assessed using 95% LOA, paired t-test, and ICC between the first and second visit SST. Coefficient of variation (CV) was also calculated. For validity and reliability, 2-way, mixed-effects, single-measure models of ICC were used. For validity, methods of VO2max assessment (measured during the cycling test versus predicted from the SST) were a fixed effect whereas the participants were a random effect. For reliability, the first and second visit SST were a fixed effect whereas the participants were a random effect. Significance level was set at 0.05 for all analyses.