Background
Rheumatoid arthritis (RA) is a chronic inflammatory disease with primary symptoms of joint pain, fatigue and major impact on functioning and health. People with RA have an increased risk of cardiovascular events which might result from an interaction between traditional risk factors, those related to chronic inflammation and possibly to physical inactivity [
1,
2].
Considering that physical activity (PA) is an important part of treatment and care in patients with RA, it is recommended that clinicians promote PA in this group [
3]. In line with evidence based practice, the effectiveness of a period of PA should be assessed and evaluated. Since self-monitoring has been identified as an effective technique of increasing PA, clinicians should encourage patients to regularly self-monitor PA progress outside of the clinic [
4]. The Fox-walk test is a novel method to estimate maximal oxygen uptake (VO
2max) by walking on an outdoor track. It is easy to perform, self-administered and requires no expensive equipment. The Fox-walk test is highly reliable in people with RA with an intra class correlation (ICC) of 0.98 (95% Confidence Interval, CI: 0.95-0.99) and the reliability is not influenced by disease-related factors [
5]. Moreover, the Fox-walk test is also a reliable method to monitor improvements in VO
2max. On a group level, the smallest detectable differences should be an increase of >1 ml · kg
−1 · min
−1 (or 2.4%) to show a clinically relevant difference, whereas on an individual level, an increase of >2.8 ml · kg
−1 · min
−1 (or 9.4%) indicates a clinically relevant difference in VO
2max[
5]. However, the test still needs to be validated in people with RA.
Assessment of aerobic fitness usually takes place in a clinical setting and is commonly supervised by a health professional as a test leader. One of the most commonly used submaximal cycle ergometry tests, suitable for a clinical setting, is the Åstrand test [
6]. The estimation of the VO
2max from the test is based on a linear relationship between mechanical load, oxygen uptake and heart rate (HR) obtained during the test. Although the test is recommended as an assessment method in physiotherapy guidelines in the management of patients with RA, it has not yet been tested for validity in this group [
7].
If true maximal HR is not known, a prediction of the individual’s maximal HR is usually needed for estimation of VO
2max by the Åstrand test. To do this an age-correction factor was incorporated in 1960 to account for the decrease in maximal HR with age [
8]. However, both this age correction factor and the most wide spread formula for age predicted maximal HR (220-age), developed in 1971 [
9] may underestimate maximal HR in an elderly healthy population [
10]. In order to increase the precision of the predicted maximal HR, and consequently the estimated VO
2max, it has recently been suggested that the formula should be modified [
10,
11]. Neither of these formulas have been evaluated in submaximal tests in people with RA.
Discussion
This is the first study to examine the criterion validity of the submaximal Fox-walk test, a self-monitoring test aiming to estimate VO2max. The results showed that the test overestimated VO2 max substantially, which should be taken into account when interpreting the results. However, despite this limitation the test could be useful for self-monitoring of aerobic fitness. To the best of our knowledge, this is also the only study to date which has examined the criterion validity of the submaximal Åstrand test in a population with RA and the test is considered to be a valid instrument to estimate VO2max in physically active people with RA.
The Fox-walk test overestimated VO2max by almost 30%, independent of participants’ levels of fitness. The overestimation could be explained by several factors. Some participants rated a low perceived exertion (five rated lower than 13) indicating that they should have performed the test running, as recommended for individuals with a high VO2max. However, this could consequently have lead to an underestimation of VO2max and not an overestimation, as was the case with the Fox-walk test. Pain in lower limbs is likely to affect performance in a population with RA and could have had an impact on the test results. This was probably not a limiting factor for the participants in the present study, indicated by the low rating of lower limb pain after walking the track and it is therefore unlikely that this could have influenced the associations between the two methods. Another factor explaining the discrepancy between the two methods could have been that the measured VO2max test was not performed with maximal exhaustion. However, a majority of the participants met the criteria for a maximal test and therefore a systematic interruption of the test at a submaximal level of exhaustion is unlikely and could not explain the large difference between the Fox walk test and the measured VO2 max test.
The submaximal Åstrand test showed a strong correlation with the VO
2max test when corrected for age expressed in l/min (r = 0.82) but weaker correlation when expressed in ml · kg
-1·min
−1 (r = 0.68). The slightly lower relative (ml · kg
−1 · min
−1 ) compared to the absolute (l/min) value should be regarded as a mathematic consequence of weight index giving a lower range in relation to the mean and thereby less good prerequisites for getting high r-values. In the present study, the Åstrand test underestimated VO
2max by 10%, which is in accordance with previous studies on healthy individuals [
22,
23], although an overestimation also has been shown [
24]. The assumption of a linear relationship between heart rate and VO
2max, makes the estimation of VO
2max from a submaximal test strongly dependent on the accuracy of the age-predicted maximal HR. Tanaka’s (208–0.7 · age) [
10] and Nes (211–0.64 · age) [
11] formulas turned out to better predict maximal HR compared to the Fox-Haskell formula (220-age) [
9], (99%, 104% and 108%, respectively, of assessed maximal HR). When age-correction was made with the use of these three alternative age-predicted HR max formulas, the widespread Fox-Haskell formula underestimated VO
2max by the same degree as Åstrand corrected for age [
8,
25], whereas the two formulas by Tanaka and Nes seem to come closer to the measured VO
2max .
Some limitations associated with this study need to be considered. The population in the present study participated in an intervention promoting physical activity and they had exercised regularly during the past year, and were well-trained. Additionally the participants in this study had low disease-activity compared to people with RA in general. In addition, a majority of the individuals included in the present study were females which also could have hampered the generalizability of the results. Three subjects used low-dose beta-adrenergic antagonists for treatment of hypertension. This could have influenced the study results according to beta blockers side-effects on HR response. However, all subjects in this study reached a maximal HR between 98% and 120% (median 107%) of the estimated age-predicted maximal HR. According to the normal HR response in these subjects, use of β-blocker antihypertensive treatment had no or limited effects on HR response in relation to work load. When performing a cycle test for the first time, anxiety and inexperience with the test situation could have an impact on the test result. The work efficiency could be lower and the ratio between the HR and the work load could be higher, consequently leading to an underestimation of VO2max. However, this was probably not the case in this study as all participants had performed the test at least twice and were familiar with exercise testing. A strength in our study was that the same biomedical scientist (T Ö) conducted all cycle ergometry tests. With regards to the Fox-walk test, the test was performed on a single track and no other tracks were tested, thus future studies should consider that different results may be obtained on other tracks.
Conclusions
The Fox-walk test cannot be used confidently for estimating VO
2max on the bases of the correlation and agreement analyses. However, the test may still be used but with consideration of its limitations when interpreting the results. We strongly recommend and encourage further development of the test, since it is a promising test for self-monitoring VO
2max by individuals outside of a clinical setting, and could also be used by professionals in the clinic. Provided that the Åstrand test is standardized according to the test manual, it should be considered as highly valid and feasible [
26] in physically active people with RA and is recommended for use by health professionals in both clinical and research settings. The newly developed formulas by Tanaka and Nes for predicting maximal heart rate according to age are preferable [
10,
11], but the Åstrand test is still valid with the use of its own age prediction VO
2max or with the Fox-Haskell formula for predicting maximal heart rate.
Competing interests
The authors declare they have no competing interests.
Authors’ contributions
BN developed and planned the study design, coordinated the study, recruited the participants, participated in acquisition of data, performed data analyses and drafted the manuscript. CF participated in the study design, acquisition and analyses of data, and drafting the manuscript. EJ participated in study design, analyses of data and drafting the manuscript. TÖ conducted all laboratory tests, participated in acquisition and analyses of data. WJG participated in analyses of data and drafting the manuscript. CO participated in the study design and participated in drafting the manuscript. AR was responsible for all laboratory tests, participated in study design, acquisition and analyses of data and drafting the manuscript. All authors participated in discussions and the revising of the manuscript and approved the final manuscript.