The Chester step test is a valid tool to assess cardiorespiratory fitness in adults with hypertension: reducing the gap between clinical practice and fitness assessments

Cardiorespiratory fitness is a vital sign of cardiovascular health [1]. Despite its utmost importance due to its potential to improve risk prediction, the assessment of cardiorespiratory fitness in clinical practice is not routinely performed [1]. This is particularly relevant for highly prevalent conditions, such as hypertension, which is the major preventable cause of cardiovascular disease and all-cause mortality [2]. Moreover, a recent study following 6890 normotensive male subjects for an average period of 14.7 years highlighted the importance of routine assessments of cardiorespiratory fitness by showing that moderate to high levels of cardiorespiratory fitness are equally beneficial in preventing hypertension in subjects with and without a family history of hypertension [3].

The “gold standard” for assessing cardiorespiratory fitness is cardiopulmonary exercise testing (CPET); [4] although, several factors, such as the duration of the protocol, equipment cost, induction of high physical stress, and need for qualified professionals, make CPET impractical on a daily basis in a clinical practice [5]. Submaximal exercise tests, including step tests, could serve as practical and valid alternatives when time is limited and laboratory and specialized staff are unavailable [5]. Among these tests, the Chester step test (CST) has the advantage of requiring minimal and portable equipment and marginal space compared to tests utilizing treadmills, shuttle walks, or cycle ergometers [6].

To date, the validity of the CST to determine cardiorespiratory fitness in adults taking antihypertensive medication compared with CPET has not been examined. Hence, this study aims to (i) test the validity of the CST to estimate the maximal oxygen uptake (VO_2max) in adults with hypertension and (ii) assess the influence of different formulas for predicting the age-predicted maximal heart rate (HR_max) when estimating the VO_2max with the CST.

Fourteen patients (eight men) with essential hypertension [2] aged between 35 and 65 years were recruited. The exclusion criteria were as follows: diabetes or any contraindication to exercise. The power calculation was computed a priori; based on a beta error of 10% and an alpha error of 5%, a sample size of 12 patients was required to observe a 0.6 correlation between CPET and the CST VO_2max values. A target of 14 patients was identified to accommodate a dropout rate of 10%. The hospital ethics committee approved the study (N/Ref. 24-01-2018). The participants provided written informed consent, and all the procedures were conducted according to the Declaration of Helsinki.

For practical reasons, participants performed the CST and CPET on the same day in that order; the tests were separated by a 10-min rest or the time necessary for all physiological parameters to return to their basal values. The participants were instructed and familiarized with the rate of perceived exertion scale, the CST and CPET. The CST was performed according to test manual recommendations with a 15-cm step [7]. It has five stages, and each stage has a 2-min duration [8]. The step cadence was set by a tape and started at 15 steps/min and increased by 5 steps/min every 2 min. The test stopped when 80% of the age-estimated maximal heart rate was exceeded, a value above 14 on the perceived exertion scale was reached or the participant was unable to maintain the metronome-set pace. The HR_max was calculated with three different formulas: the Fox–Haskell (220-age) [9], Tanaka (220 − 0.7 × age) [10] and Nes (220 − 0.64 × age) formulas [11]. The examiner deriving the VO_2max was blinded to the study purpose. The VO_2max was measured on a maximal CPET performed on a cyclergometer (Cardiovit CS-200 Ergo-Spiro, Schiller, Baar, Switzerland); the test started at 50 W and increased 25 W every 2 min; the participant was instructed to keep a cadence of 60 revolutions per minute (rpm). The VO_2max was assessed by the following criteria: (i) a plateau in the VO₂ with increases in external work, (ii) a maximal respiratory exchange ratio (RER) > 1.10, and (iii) an HR_max exceeding 90% of the age-predicted maximum. The test was considered to be limited by the cardiorespiratory system when the participants could no longer maintain the targeted 60 rpm and at least the last two criteria were fulfilled. For this study, a maximal cycle test was chosen rather than a treadmill protocol because the cycle ergometer and step tests have been shown to yield similar VO_2max results [12]. All the tests were conducted by a cardiologist and physiotherapist. Pearson’s correlation and paired t-tests were used to test associations and compare mean differences, respectively. To assess agreement, Bland–Altman plots were constructed using the difference between the means of VO_2max in the CST and VO_2max in CPET and the standard deviations of the calculated differences.

Table 1 summarizes the characteristics of the participants. The VO_2max predicted by the CST was dependent on the formula used to determine the HR_max, i.e., a significantly lower VO_2max was obtained with the Fox–Haskell formula than with the Tanaka and Nes formulas (Table 1). The VO_2max measured during CPET was lower than the CST VO_2max predicted using the Tanaka [mean diff (95% CI): −1.06 (−1.84 to −0.28) ml kg⁻¹ min⁻¹, p = 0.012] and Nes formulas [−2.11 (−2.87 to −1.35) ml kg⁻¹ min⁻¹, p < 0.01]; no significant difference was found when using the Fox–Haskell formula [−0.35 (−1.06 to 0.35) ml kg⁻¹ min⁻¹, p = 0.30] (Table 1).

Table 1 Characteristics of the participants and main results of the exercise tests

The VO_2max predicted by the CST showed a strong, positive correlation with the VO_2max measured during CPET, with the strongest correlation obtained with the Fox–Haskell formula (r = 0.989, p < 0.001), followed by the Nes (r = 0.987, p < 0.001) and Tanaka formulas (r = 0.986, p < 0.001). The bias (95% limits of agreement) between the VO_2max measured during CPET and that estimated by the CST was −0.35 (−2.74 to 2.04) ml kg⁻¹ min⁻¹ with the Fox–Haskell formula, −1.06 (−3.72 to 1.60) ml kg⁻¹ min⁻¹ with the Tanaka formula, and −2.11 (−4.70 to 0.48) ml kg⁻¹ min⁻¹ with the Nes formula (Fig. 1). No bias trend was observed across the range of the VO₂ values studied.

Fig. 1

Bland–Altman plots comparing the agreement between the measured VO_2max during CPET and estimated by the CST using the a Fox–Haskell formula, b Tanaka formula, and c Nes formula

One of the assumptions of the CST is that a linear relationship between HR and VO_2max exists, making the result of the test dependent on the accuracy of the formula used to predict the individual HR_max. The CST showed a strong correlation with CPET, independent of the formula used to determine the age-predicted HR_max. This correlation is in agreement with previous studies; a systematic review [5] on the validity and reliability of submaximal step-test protocols to estimate VO_2max in healthy adults found correlations between 0.469 and 0.95; the best correlations belonged to the CST and the personalized step test.

In our study, the Fox–Haskell formula seemed to be the best formula when conducting the CST in this population. The 95% limits of agreement between the VO_2max predicted by the CST using the Fox–Haskell formula in our study were similar to those observed in 13 young healthy subjects in two trials (−2.8 ± 6.1 ml kg⁻¹ min⁻¹ and −1.9 ± 7.4 ml kg⁻¹ min⁻¹) [6]. In our study, the 95% limits of agreement oscillated from −2.7 to 2.0 (mean, −0.35) ml kg⁻¹ min⁻¹; this bias may not be significant when prescribing exercise training, but all health professionals must be aware of these limitations when using this test. Our results should be interpreted cautiously given the sample size; however, the existing validation studies on submaximal tests had similar sample sizes. Future studies should consider the assessment of blood pressure during the CST protocol, as is done in CPET, to determine if the blood pressure response to submaximal exercise exhibits the same pattern as it does in CPET. The lack of time and availability of participants to take part in multiple day assessments was also a limitation. Hence, the final results could have been influenced by the accumulated fatigue of the participants from one test to the other, even though there was a prudential time to rest.

In conclusion, our findings highlight that (i) the CST is a valid, easy, and inexpensive solution for assessing the VO_2max in individuals with hypertension and (ii) the Fox–Haskell formula is good for predicting the HR_max when using this test to estimate cardiorespiratory fitness. The CST provides a straightforward way to evaluate cardiorespiratory fitness during routine clinical visits.