Change in Free-Living Daily Steps does not Affect VO_2max Adaptation in Response to a Supervised Aerobic Walking Program in Females Ages 18–45

Abstract

We hypothesized that an increase in nonexercise physical activity (NEPA), assessed by daily steps outside of steps accrued during supervised exercise training sessions, would be positively correlated with the change in VO_2max. Females ages 18–45 yr (n = 44; 30 ± 7 yr; 67.7 ± 18.3 kg; 24.9 ± 6.4 kg/m²) completed 36 supervised training sessions on a motorized treadmill (3 sessions/week, 30 min/session) over 12 weeks, at 70% VO_2max (80% max heart rate). VO_2max was assessed at baseline, 4, 8, and 12 weeks, and steps outside of the supervised exercise sessions were recorded daily during each week of training with a hip-worn pedometer. For each participant, linear regression was used to determine the slope of steps/day during the 12 weeks of training. Mean VO_2max increased by 8.2% (30.5 ± 5.9 ml/kg/min to 33.0 ± 5.9 ml/kg/min; range −2.65 ml/kg/min to +6.24 ml/kg/min; P < 0.001). Although mean (± Standard Deviation) steps/day did not change during the 12-week intervention, 25 participants decreased daily steps (−1624 ± 1210) and 19 participants increased daily steps (+1713 ± 1402). The change in VO_2max was not different between the two groups (P = 0.74), and the correlation between Δ VO_2max and the slope of the regression (steps/day) over the 12 weeks was not significant (R² = 0.0005; P = 0.88 for ml/kg/min; R² = 0.008, P = 0.59 for l/min). Change in NEPA, assessed by daily steps, does not impact the VO_2max adaptation observed during a vigorous-intensity walking exercise program in females ages 18–45 yr.

Introduction

Maximum oxygen uptake (VO_2max) is a strong and consistent predictor of morbidity and mortality,1,2 and the American Heart Association has recommended that cardiorespiratory fitness be included as a vital sign in clinical practice.3 Aerobic exercise training is recommended for improving cardiorespiratory fitness.4 Although aerobic exercise training studies consistently show increases in mean VO_2max, there is considerable response heterogeneity.5–9

One potential factor that could contribute to the inter-individual variability in VO_2max response is nonexercise physical activity (NEPA), which generally represents ambulatory physical activity outside of a structured exercise program.10–12 Some reviews have indicated that NEPA does not change during exercise training,10–12 although four studies reported in one review indicated that NEPA may decrease with exercise training.13 However, these reviews focused on mean changes in NEPA for the entire study population. Large inter-individual differences could have occurred that would be undetected by examining mean changes alone. Both increases and decreases in NEPA have been reported in exercise training studies.14–16 Decreases in NEPA upon initiating a structured exercise program may be due in part to behavioral adaptations to constrain total daily energy expenditure, although the mechanisms and factors responsible for changes in NEPA remain unknown.17,18

Walking is a major contributor to NEPA.19 Because daily steps have been reported to be positively associated with cardiorespiratory fitness,20–25 changes in daily steps could conceivably affect the magnitude of the VO_2max response to aerobic exercise training. Two randomized controlled trials in adults with overweight or obesity have addressed this issue and produced contrasting results.26,27 In one of these, 6 months of aerobic exercise training combined with a goal of increasing daily steps by ~3000/day (actual increase ~2000 steps/day) improved VO_2max whereas aerobic exercise training alone (with no increase in steps/day) did not.27 In the other study, 12 weeks of aerobic exercise training alone increased VO_2max though, despite not being asked to increase daily steps, there was a mean increase of ~3400 steps/day among participants, which could have contributed to the increase in VO_2max.26 The group assigned to a combination of aerobic exercise training and increasing daily steps (mean increase ~5500 steps/day) increased VO_2max to a similar degree as exercise training alone, but in a third group that just increased steps alone (~3200/day, with no aerobic training), VO_2max was unchanged. Thus, the contribution of daily steps to the changes in VO_2max with exercise training is unclear. Both studies had small sample sizes (12–16 per group), making it difficult to appreciate how inter-individual variability in daily steps may influence VO_2max among participants engaged in a supervised aerobic exercise training program.

The purpose of our study was to determine whether changes in NEPA (daily steps) during the course of a 12-week supervised aerobic exercise program (treadmill walking) were correlated with changes in VO_2max. We anticipated large inter-individual variability in changes in daily steps, as well as changes in VO_2max. We hypothesized that in a 12-week supervised, vigorous-intensity walking exercise training program in females, the change in daily steps outside of those accrued during supervised exercise sessions would be significantly correlated with the change in VO_2max.

Methods

Participants

This is a secondary data analysis of a study designed to assess predictors of fat loss during an exercise training program in women.28 As such, no a priori power analysis was performed for this study. Participants included 44 non-pregnant, females ages 18 to 45 yr, not engaged in any other exercise program prior to or during the study, were non-smokers, and did not have any known cardiovascular, pulmonary, renal, or metabolic diseases. The study was approved by the Arizona State University Institutional Review Board and was conducted in a manner consistent with the Declaration of Helsinki. Written informed consent was obtained from each participant. This research was carried out fully in accordance with the ethical standards of the International Journal of Exercise Science.29 Baseline characteristics are shown in Table 1.

Table 1.

Baseline characteristics and changes maximal exercise variables and anthropometrics with exercise training (n = 44).

	Baseline	Week of Training			P - value
		4	8	12
Age (yr)	30 ± 7
Height (cm)	164.9 ± 7.7
Weight (kg)	67.7 ± 18.3	68.1 ± 18.6	68.1 ± 18.1	67.0 ± 18.3	0.24
BMI (kg/m2)	24.9 ± 6.4	25.1 ± 6.6	25.0 ± 6.4	25.0 ± 6.3	0.34
Body Fat (kg)	25.3 ± 9.0	25.5 ± 9.0	25.0 ± 8.8	25.2 ± 9.0	0.35
VO2max (ml/kg/min)	30.52 ± 5.88	32.17 ± 5.94	32.23 ± 5.95	33.00 ± 5.87	<0.001
VO2max (l/min)	2.00 ± 0.38	2.10 ± 0.40	2.12 ± 0.34	2.17 ± 0.38	<0.001
HRmax (bpm)	184 ± 13	182 ± 14	184 ± 12	185 ± 11	0.22
RERmax	1.10 ± 0.04	1.09 ± 0.06	1.09 ± 0.05	1.11 ± 0.04	0.10

Values are means ± SD. BMI, body mass index; HR, heart rate; RER, respiratory exchange ratio. P-values represent 1-way repeated measures ANOVAs.

Protocol

The details of the 12-week exercise intervention have been described previously.28 In brief, the study consisted of 36 supervised exercise training sessions over 12 weeks: 3 vigorous-intensity (70% VO_2max) treadmill walking sessions per week, 30 min per session. Participants who missed a training session were required to make up the missed session within the next week. Total daily steps were assessed in all participants using a hip-worn pedometer (YAMAX Digi-Walker SW200; New Lifestyles, Lee’s Summit, MO, USA). Validity and reliability of this pedometer has been established.30 V̇O_2max was assessed at baseline and after 4, 8, and 12 weeks of aerobic exercise training. Height was assessed by a wall-mounted stadiometer (Seca, Chino, CA, USA). Weight was taken using a calibrated electronic scale (COSMED, Rome, Italy) with participants wearing a bathing suit, and body composition was determined using dual-energy x-ray absorptiometry (DXA, Lunar Prodigy; GE Medical Systems Lunar, Madison, WI, USA).

All participants performed a modified Balke incremental treadmill exercise test to assess VO_2max on a motor-driven treadmill (Trackmaster TMX 425, Full Vision Inc., Newton, KS). The treadmill protocol started with 1 min of walking at 1.5 mph at a 0% grade, then increased to 3.3 mph for 1 min. After the second min, the grade increased to 2% and then increased by 1% each min until volitional exhaustion was reached despite verbal encouragement. Ventilation and gas exchange were measured continuously with a ParvoMedics Truemax 2400 metabolic cart (ParvoMedics, Sandy, UT, USA) for determination of VO₂, carbon dioxide production, and respiratory exchange ratio (RER). Heart rate was continuously measured with a Polar heart rate monitor (Polar, Lake Success, NY, USA). VO_2max was taken as the average of the 2 highest consecutive 15-s values. The following criteria were used to define VO_2max: 1) VO₂ plateau, or 2) maximal RER ≥1.05 and 95% of age-predicted heart rate max using the formula 208 – (0.7 × age).31 We defined a VO₂ plateau as an increase in VO₂ during the last stage of the treadmill test that was <50% of the expected VO₂ increase established by the slope of the VO₂-stage relationship for the submaximal exercise stages of the treadmill test.32,33

Participants underwent 12 weeks of exercise training consisting of treadmill walking 3 days per week for 30 min each session. Each participant was allowed to select the treadmill speed based on personal preference (range 2.6–3.4 mph). The treadmill grade was adjusted to elicit a heart rate corresponding to 70% VO₂ (~80% of heart rate max) for each of the 36 training sessions. Heart rate during exercise sessions was monitored continuously with a Polar heart rate monitor and was recorded every 5 min. Each session included a 5-min warm-up and cool-down. The target heart rate was adjusted based on VO_2max changes at 4 and 8 weeks. Exercise sessions were completed on non-consecutive days, and all participants completed 36 exercise training sessions. Pedometers were not worn during the training sessions.

Statistical Analysis

All analyses were performed with SPSS 28.0 (IBM, Armonk, NY) with significance set at P < 0.05. All variables are presented as mean ± standard deviation (SD). A 1-way repeated measures analysis of variance was conducted to assess changes in VO_2max and daily steps during the 12-week intervention, as well as changes in weight, BMI, total and percent body fat, HR_max, and RER_max. We assessed whether daily steps changed over the course of the 12-week intervention by two methods. First, for each participant, linear regression was used to determine the slope of the change in daily steps (steps/day) over the 84 days of pedometer data. Second, we compared the average daily steps during the first 3 weeks of training with the average daily steps during the last 3 weeks of training. Pearson correlation coefficients were used to assess whether the change in VO_2max from baseline to 12 weeks was correlated with the change in daily steps over 12 weeks. A two-sided unpaired t-test was used to compare changes in VO_2max between participants who had positive and negative slopes for steps/day during the 12-week intervention.

Results

Of the 81 participants in our original study, 21 did not complete pedometer diaries, leaving 60 participants eligible for inclusion in this analysis. Of these 60 participants, only 44 met the criteria for VO_2max. Of the 88 tests (baseline and 12-week) used to assess a training response for these 44 participants, only 13 (15%) achieved a VO₂ plateau. Therefore, VO_2max for the remaining 75 (85%) of the tests was established using maximal RER ≥1.05 and 95% of age-predicted heart rate max. Of the 44 participants included in the analysis, 10 had zero missing days of pedometer data, 23 had between 1–7 days of missing pedometer data, 6 had between 8–14 days of missing pedometer data, and 5 had between 15–25 days of missing pedometer data. All but 1 participant had at least 65 days of pedometer data. All 44 participants included in the data analysis completed all 36 supervised training sessions.

Mean VO_2max increased significantly (P < 0.001) from 30.52 ± 5.88 ml/kg/min (2.00 ± 0.38 l/min) at baseline to 33.00 ± 5.87 ml/kg/min (2.17 ± 0.38 l/min) after 12 weeks of exercise training (Table 1). There was considerable heterogeneity in VO_2max responses (−2.65 ml/kg/min to +6.24 ml/kg/min; −0.11 l/min to +0.47 l/min), with the majority of participants increasing VO_2max after 12 weeks (Figure 1).

Figure 1.

Individual changes in VO_2max after 12 weeks of supervised exercise training, in ml/kg/min (A) and L/min (B). Individual changes in steps/day from baseline to 12-weeks (C and D), with subjects aligned in the same order as presented for corresponding VO_2max changes in panels A and B. The Δ steps is represented by the slope of the linear regression equation for each participant over the 12 weeks of training, in steps/day. N = 44

Mean daily steps did not change (P = 0.27) during the 12-week intervention (Table 2). Similar to the results for VO_2max, there was considerable heterogeneity in the changes in daily steps over the course of 12 weeks (Figure 1). Regression analysis indicated that 25 participants decreased daily steps (negative slope), and 19 participants increased daily steps (positive slope). When restricting the analysis to comparing the mean daily steps during the first 3 weeks and last 3 weeks of the study (i.e., weeks 10–12 vs. 1–3), 24 participants experienced a decrease in steps/day and 20 participants had an increase in steps/day. The correlation between the two methods for assessing change in daily steps was 0.98 (P < 0.00001). Because the regression method utilized 100% of the step data rather than just step data at the beginning and end of the exercise intervention, we used Δ step data from the regression analysis for our results and interpretation of data.

Table 2.

Mean steps/day for each week of training.

Week	1	2	3	4	5	6	7	8	9	10	11	12
Steps/day	7578 ± 2778	7892 ± 2663	8052 ± 2507	8085 ± 2569	7740 ± 2364	7590 ± 2560	7544 ± 2650	7378 ± 2434	7615 ± 2558	7400 ± 2527	7818 ± 3035	7990 ± 2448

Values are means ± SD. P = 0.27 for change in steps/day over time.

The distribution of positive and negative slopes for Δ daily steps did not coincide with the changes in VO_2max (Figure 1). Participants who increased their daily steps during the 12-week study (positive slope) were no more likely to increase VO_2max than those who decreased their daily steps (negative slope). The weak relationship between Δ V̇O_2max and Δ step daily steps is illustrated in Figure 2, which shows that changes in both relative (R² = 0.0005; P = 0.88) and absolute (R² = 0.008; P = 0.59) VO_2max were not significantly correlated with the changes in daily steps over the course of the intervention. The 25 participants with a negative slope (1624 ± 1210 fewer steps/day at the end of the intervention) had a mean increase in VO_2max of 2.39 ± 1.93 ml/kg/min, which was not different than the 2.59 ± 2.16 ml/kg/min increase observed in the 19 participants with a positive slope (1713 ± 1403 more steps/day at the end of the intervention) (P = 0.74). Daily steps during the 12 weeks of training did not differ for the participants who decreased their daily steps (negative slope) or increased their daily steps (positive slope) during the intervention (7655 ± 2397 steps/day vs. 7726 ± 1790 steps/day; P = 0.92).

Figure 2.

Change in VO_2max in ml/kg/min (A) and L/min (B) after 12 weeks of supervised exercise training versus change in daily steps from baseline to 12-weeks. The Δ steps is represented by the slope of the linear regression equation for each participant over the 12 weeks of training, in steps/day. N = 44

Discussion

Contrary to our hypothesis, change in daily steps was not correlated with changes in VO_2max. To be statistically significant, the R² for Δ VO_2max vs. Δ steps/day (Figure 2) would require a sample size of 989. Statistical significance, however, is greatly undermined by the fact that Δ steps/day explained <1% of the Δ VO_2max. This suggests that changes in NEPA do not contribute to between-subject variability in V̇O_2max response to aerobic exercise training.

Similar to previous studies,14–16 our participants demonstrated large variability in changes in daily steps during the study. This heterogeneity was not obvious from the mean data, which showed that daily steps for the entire sample did not change (Table 1). However, the individual data indicated large heterogeneity, with 57% of the participants decreasing daily steps over the course of 12 weeks, and 43% of participants increasing daily steps. The difference in mean daily steps between the two groups was ~3300 steps per day (i.e., −1624 vs. +1713). At a medium walking pace for adults of 90 steps/min,34 this difference could represent ~37 min of walking per day. Although we could not determine the intensity of the steps taken by our participants, this amount of NEPA could conceivably affect cardiorespiratory fitness.35–37 Despite this difference, changes in daily steps outside of the supervised walking sessions had no impact on the change in VO_2max as indicated by the very weak correlation between Δ steps/day and Δ VO_2max, and the fact that the increase in VO_2max did not differ between those with a positive or negative slope in daily steps during the 12-week intervention.

In a 6-month study of adults with obesity, aerobic training combined with a goal of increasing daily steps by ~3000/day (actual increase ~2000 steps/day) improved VO_2max whereas aerobic training alone (with no increase in steps/day) did not.27 It is surprising that V̇O_2max was not improved despite 3–4 exercise sessions per week in the group that did not increase steps/day during the study. These findings contrast with a larger body of literature that shows improvements in VO_2max of ~0.28 L/min or ~3.94 ml/kg/min with exercise training.38 The lack of effect of supervised aerobic exercise training on VO_2max could be due to the small sample size (N=12) or potentially from participants not achieving VO_2max during testing. The authors did not define what criteria were used to confirm attainment of VO_2max. It is also important to note that regression analysis in that study indicated that the change in steps/day during the 24 week study period was a weak, and not a statistically significant, predictor of the change in VO_2max after training.27 This is at odds with their main finding that VO_2max was increased only in the training group that also increased daily steps during the study. This result is, however, consistent with our finding that the change in steps/day during the 12 weeks of training was not significantly correlated with Δ VO_2max. That changes in steps/day are not likely to impact Δ VO_2max in response to training is also supported by the findings of Kozy Keadle et al,26 which demonstrated that increasing daily walking by ~3200 steps/day over 12 weeks in the absence of exercise training had no effect on VO_2max.

In the DREW study,39 mean daily steps did not change over the course of the 6-month study, which is similar to our findings. However, no analysis was performed in that study to see if changes in steps were correlated with changes in VO_2max. In Ross et al.,40 sedentary adults with overweight or obesity were encouraged to increase daily steps by 2000/day over 24 months. Estimated VO_2max did not change, suggesting that increasing steps by this amount was not sufficient to increase cardiorespiratory fitness in the absence of a moderate or vigorous exercise training program. But individual changes were not reported in that study. Consequently, changes in steps alone may not provide information necessary to assess between-participant heterogeneity. Our results, however, suggest that changes in daily steps, as a proxy for NEPA, may not impact the effectiveness of an aerobic exercise program for improving VO_2max if the training intensity is vigorous. During the supervised training sessions our participants performed uphill walking at 70% of VO_2max, which corresponded to ~21 ml/kg/min and ~80% of maximum heart rate. The preferred walking speeds of our participants during their training sessions ranged between 2.6 and 3.4 mph, which corresponds to a VO₂ of ~10–16 ml/kg/min.41,42 If most of our participants’ walking outside of training sessions occurred at close to their preferred walking speed, this may be too low an intensity to provide sufficient stimulus to improve VO_2max beyond that of the vigorous-intensity training sessions.

Our study has several strengths. All participants completed the 36 training sessions, demonstrating 100% adherence. All training sessions were supervised with heart rate monitoring to ensure exercise intensity was at the target intensity for each training session. We monitored steps daily throughout the 12-week training program. Moreover, we calculated regression equations for each participant to assess changes the steps/day during the training program. Results were the same regardless of whether we used the slope of the regression of steps/day during the entire 84 days of the intervention or just used data from the first 3 and last 3 weeks of the 12-week intervention. Previous studies did not explicitly state how the change in steps/day during the course of the study was calculated. Also, we had participants remove their pedometer during each training session, so our step data include only steps taken outside of training sessions. We adhered to strict criteria for VO_2max, and only included participants who met the criteria for both pre- and post-training tests.

Our study has some limitations. We did not record baseline steps prior to starting the exercise program. However, this would likely have had negligible impact on the slope of the steps/day calculated during the 12 weeks of training. We did not measure other activity outside of the supervised sessions, nor were we able to assess the intensity of the daily steps. Our study did not have a no-exercise control group; thus, we could not assess the potential impact on V̇O_2max of changes in daily steps in the absence of exercise training. We did not include a verification test to confirm VO_2max,32,33 but this is also a limitation of previous studies on this topic.20–22,26,27,35,39 Average daily steps of our participants (~7000–8000 steps/day; Table 2) is much greater than that of females in the United States (~4500–5000 steps/day).43 Thus, our findings may not be generalizable to females with much lower daily step counts. Finally, most of our participants had BMI ≤ 25 kg/m², with only 7 having a BMI > 30 kg/m². Thus, our results may not be generalizable to females with obesity, or to older females and males. 43

In conclusion, our results indicate that initiation of a vigorous-intensity walking training program in females results in large individual variability in changes in daily steps outside of the training sessions, as well as changes in VO_2max. However, the change in VO_2max was not correlated with the change in daily steps, suggesting that change in NEPA is not an important determinant of the response variability to a vigorous-intensity exercise training in females ages 18–45 yr.