Advertisement

Comparisons of Subjective and Objective Measures of Free-Living Daily Physical Activity and Sedentary Behavior in College Students

  • Ya-Wen Hsu
  • Chia-Chang Liu
  • Yen-Jung Chang
  • Yi-Ju Tsai
  • Wan-Chi Tsai
  • You FuEmail author
Original Article
  • 44 Downloads

Abstract

Purpose

To compare physical activity and sedentary behavior between four commonly used subjective and objective measures: the 7-day Physical Activity Recall (7DPAR), International Physical Activity Questionnaire (IPAQ), pedometer, and accelerometer.

Methods

A total of 130 college students completed four measures for the same 7 days. Body composition was measured using a bioelectric impedance analyzer. Wilcoxon signed rank tests and Spearman correlations were performed to compare estimates between activity measures. The Spearman correlations between different activity measures were further examined separately for the higher and the lower body fat groups.

Results

Compared with accelerometer-derived data, both the 7DPAR and the IPAQ overestimated light physical activity (P < 0.001) while underestimated sedentary behavior (P < 0.001). Across comparisons, the highest correlation was found between accelerometers and pedometers on steps/day (r = 0.72, P < 0.001). The 7DPAR and the IPAQ were correlated with each other for all physical activity variables and sedentary behavior (r = 0.37–0.45). There were low correlations (r = 0.20–0.47) between the 7DPAR, the IPAQ, and accelerometers in sedentary behavior, light physical activity, and vigorous physical activity. Higher correlations between different activity modalities were observed among individuals with lower body fat (r = 0.41–0.80) than among those with higher body fat (r = 0.31–0.65).

Conclusions

The 7DPAR and the IPAQ yielded comparable estimation of moderate physical activity relative to accelerometers. There were significant differences in sedentary time across activity measures. Body compositions should be considered when comparing the estimates of activity levels between subjective and objective instruments.

Keywords

Accelerometer Pedometer Self-report Body fat College student 

Introduction

Physical activity (PA) is defined as “any bodily movement produced by skeletal muscles that requires energy expenditure” [9], and activities such as “sitting, lying, sleeping, and watching television are collectively considered as sedentary behaviors” [33]. The lack of PA and excessive sedentary behavior are two major risk factors for chronic diseases, such as obesity, cardiovascular disease, diabetes, hypertension, and mental disorder [8, 24, 40]. Consequently, accurate measurement of PA and sedentary behavior will likely result in significant public health implications and benefits.

Researchers have been exploring optimal approaches to accurately assess PA and sedentary behaviors. Subjective method is predominantly relied on self-reported instruments. Questionnaire are commonly employed to assess PA [29] and sedentary behaviors [7, 20], given their low cost and their ability to categorize behavior by frequency, intensity, time, and type [36]. The International Physical Activity Questionnaire (IPAQ) and the Seven-Day Physical Activity Recall (7DPAR) are two of the most commonly used self-report measures of PA [30]. Despite their wide use, most self-report instruments have limitations. In addition, self-reporters are compelled to restrain their responses into the limited answer choices, and may further alter their responses for social desirability and tendency [16]. Finally, the self-report results mainly rely on personal recall of relevant instances from memory and judgement, which may be unreliable and occasionally lead to biases [37]. These inherent disadvantages may cause inaccurate data, therefore the use of subjective measures has been a concern in research.

Advancement in technology has enhanced the applicability of objective activity measures [11, 14, 26]. A pedometer is an affordable objective instrument that enables us to capture low to moderate intensity PA, such as walking and other lifestyle activities, which are difficult to accurately quantify via self-report questionnaires [11, 26]. An accelerometer is able to differentiate activity intensities and has been considered a current standard for assessing PA in various settings. Compared to literature regarding PA, limited research has employed accelerometers to assess sedentary behaviors [17].

Despite a large number of studies comparing subjective and objective activity measurements, the majority of the existing research has been conducted among Western population [10, 13, 14, 18, 27, 30, 41], with very few based on Asian populations [44]. More studies are needed to investigate the relationship between self-report and objective assessment in different populations considering that measurement properties of the questionnaire may differ by demographic characteristics, such as age, culture, and region [42]. In addition, prior studies are often limited to compare a particular questionnaire to one objective measure. Comparing multiple measures simultaneously provides a unique insight into how self-reported PA and objective estimates differ and whether or not the magnitude of difference varies across intensity levels. Another gap is that few studies have explored whether the comparisons of activity measures may differ by body compositions, a potential confounding factor. Thus, the primary purpose of this study was to compare the four widely used subjective and objective PA measures (IPAQ, 7DPAR, accelerometers, and pedometers) in assessing college students’ daily PA and sedentary behavior in Taiwan. A secondary purpose was to investigate how these measures may differ in estimating PA and sedentary behavior among individuals with different body compositions.

Methods

Participants

The participants of this study were a convenient sample of 135 college students recruited from a University in Taiwan. Flyers were posted on the message boards in the university and interested students could sign up to be contacted. At the screening visit, participants completed body composition measures and demographics surveys at the health promotion center on campus. Participants were excluded from the study if they (1) were previously diagnosed with any major illnesses that could affect physical abilities; or (2) took any medications that may influence body composition. Institutional Review Board approval was obtained from the Kaohsiung Medical University Chung-Ho Memorial Hospital. Consent forms were obtained from all participants.

Procedures

Eligible participants were instructed to wear the two objective measures (accelerometer and pedometer) and to complete the two subjective measures (IPAQ and 7DPAR) for the same 7-day period. During the 7 days of measurement, participants were asked to maintain their current levels of PA. Each day, they wore an accelerometer and a pedometer simultaneously during waking hours, with the exception of bathing or during aquatic activities. In addition, self-reported daily activity levels were measured using the IPAQ and the 7DPAR for the same 7 days. The research staff contacted the participants via phone calls and text messages to ensure that the accelerometers and pedometers were properly worn and that the IPAQ and 7DPAR were completed. Participants were also instructed to keep a log diary the time they wore the activity devices during the 7-day period.

Measurements

Objective Measurements

Accelerometers. ActiGraph GT3X accelerometers (ActiGraph, Pensacola, FL, USA) have proven validity and reliability in assessing PA and sedentary behavior among adults [1, 22]. Each participant wore an accelerometer on the hip of the dominant leg. Accelerometers were recorded in 60-s epochs and data were processed using ActiLife® software version 6.7.2. The nonwear periods were defined as 60 min of consecutive zero counts and were cross-validated by the self-report logs from participants. A valid day of wear was defined as having at least 10 h of wear time. Participants with 7 valid days of data were included in the analyses. Outcome variables were mean minutes per day in light PA (LPA; < 3 METs), moderate PA (MPA; 3–6 METs), vigorous PA (VPA; ≥ 6 METs), and sedentary behavior (< 100 counts/min) [4, 39]. Mean minutes per day were calculated by summing each minute spent daily and averaging across all valid days of wear. Mean steps per day were calculated by summing the daily step counts accumulated and averaging across all valid days of wear.

Pedometers. Participants wore Omron pedometers with dual piezoelectric sensors (Model HJ-720ITC, Omron Healthcare, Kyoko, Japan) for 7 days on the same side of hip that they wore the accelerometers. Data were downloaded using Omron™ software version 1.3 for total steps per day. A valid day includes at least 500 steps and 10 h of nonzero step counts. Individuals with 7 valid days of data were included in the analyses. Mean steps per day were calculated by summing the daily total step counts and averaging across all valid days of wear.

Body Composition. Weight and height were measured three times to the nearest 0.1 kg and 0.1 cm, respectively, using a beam medical scale and wall-mounted stadiometer. BMI was calculated as kg/m2. Percent body fat was measured using a bioelectric impedance analyzer (X-SCAN Plus-II). The measurement was performed in a standing position with participant barefoot and wearing light clothing.

Subjective Measurements

International Physical Activity Questionnaire (IPAQ). The modified Taiwan version of short-form IPAQ [25] was used to assess daily activity levels. Participants reported how much time in minutes per day they spent engaged in PA at each intensity: VPA (activities that make you breathe much harder than normal), MPA (activities that make you breathe somewhat harder than normal), LPA (activities that make you breathe normally), and sedentary behavior (time spent sitting at a desk, visiting friends, reading or sitting or lying down to watch television.). Mean minutes per day in VPA, MPA, LPA, and sedentary behavior were obtained by averaging across the total estimates of each intensity level from 7 days. The original IPAQ does not include the assessment of LPA; instead, there was an item asked time spent walking. Considering that walking and LPA are different constructs and should not be treated as the same [35], we modified the question for participants to report time spent in LPA so that there will be comparable estimates across activity measures included in the current study.

7-Day Physical Activity Recall (7DPAR). Participants recorded their activities using the 7DPAR for the same 7-day period as they completed three other activity measurements [34]. A modified 7DPAR was developed by adding some physical activities specific to Taiwanese culture and lifestyles (e.g., karaoke, Qigong). From a list of 86 activities provided, participants identified the main activities to describe their activity in every half-hour block for 24 h and rated the intensity level (light, moderate, vigorous) for each activity. Activity types were converted into half-hour blocks of either LPA (< 3 METs), MPA (3–6 METs), or VPA (≥ 6 METs) using a combination of the intensity ratings provided by the participants and the compendium of physical activities [2]. Half-hour intervals spent watching television/movies, playing video games, using computers/surfing the internet, sitting in class, and talking on the telephone were coded separately as sedentary behavior. Total time spent in LPA, MPA, VPA, and sedentary behavior was obtained by summing over the half-hour intervals for 1 day. Mean minutes per day was obtained by averaging total minutes across 7 days.

Statistical Analysis

The distribution of the data was checked using the Shapiro–Wilk test for normality (≤ 0.05). As the PA and sedentary behavior data were non-normally distributed, non-parametric statistical tests were used. Wilcoxon matched-pairs signed rank tests were performed to compare (1) total time in LPA, MPA, VPA, and sedentary behavior between the IPAQ, 7DPAR, and accelerometers and (2) steps/day between accelerometers and pedometers. Cohen’s d effect size was calculated to assess the magnitude of the difference between activity measures [12]. The Spearman correlation coefficient was computed to determine the relationships between time spent in LPA, MPA, VPA, and sedentary behavior across the IPAQ, 7DPAR, and accelerometers, as well as the steps/day between accelerometers and pedometers. Stratifying by gender, the same sets of correlations were further analyzed using partial Spearman correlation coefficients, adjusting for age and percent body fat. In addition, the Spearman correlations between different activity measures for PA variables and sedentary behavior were examined separately for the higher and the lower body fat groups. In order to have similar sample size for proper between-group comparisons, participants were further dichotomized into higher and lower body fat groups at the gender-specific median value for percent body fat (male: 18.3%; female: 23.5%). All analyses were conducted using SPSS version 18.0 (IBM Corp., Armonk, NY, USA), with P ≤ 0.05 considered as statistically significant.

Results

Among the 135 students who completed the screening visit, 5 were excluded because of missing data. The participants’ mean BMI was 21.9 kg/m2 with 22.3% (n = 29) classified as overweight (≥ 24 kg/m2) and 77.7% (n = 101) as non-overweight (< 24 kg/m2) [19]. The mean percent body fat was 18.8% among males and 24.2% among females. Demographic information is shown in Table 1.
Table 1

Demographic characteristics

 

Mean (± SD)/n (%)

Gender

 Female

101 (78%)

 Male

29 (22%)

Age (years)

21.1(± 1.2)

Percent body fat (%)

 

 Female

24.2(± 5.3)

 Male

18.8(± 7.8)

Percent body fat group

 

 Higher

65 (49.6%)

 Lower

66 (50.4%)

BMI (kg/m2)

21.9 (± 3.9)

Participants were further dichotomized into high and low body fat group at the gender-specific median value for percent body fat (male: 18.3%; female: 23.5%)

Mean Differences Between Measures

The comparisons of subjective and objective measures are presented in Table 2. A significant difference in mean daily steps was found between the pedometer and accelerometer (mean difference = 474 steps/day, P < 0.001). Between the two self-report questionnaires, the mean mins/day in LPA was higher based on the 7DPAR (P < 0.001) while the mean mins/day in VPA and sedentary behavior were higher based on the IPAQ (P < 0.001). Using accelerometers as criterion measure, the 7DPAR yielded overestimations on LPA (P < 0.001), but underestimations on VPA (P < 0.001) and sedentary behavior (P < 0.001). On the other hand, the IPAQ overestimated mean mins/day in LPA (P < 0.001) while underestimated time spent in sedentary behavior than the accelerometers (P < 0.001).
Table 2

Comparisons of activity levels by measurement

Accelerometers vs. pedometers

Mean (± SD)

Mean (± SD)

 

Effect Size (Cohen’s d)

Steps (steps/day)

5708.8 ± 1915.7

6182.6 ± 2351.5

***

0.22

7DPAR vs. IPAQ

 LPA (mins/day)

596.4 ± 419.6

302.0 ± 197.4

***

0.90

 MPA(mins/day)

36.9 ± 40.1

44.1 ± 54.0

 

0.15

 VPA (mins/day)

1.9 ± 5.8

8.6 ± 19.2

***

0.47

 Sedentary Behavior(mins/day)

311.1 ± 133.2

570.0 ± 190.2

***

1.58

7DPAR vs. accelerometers

 LPA (mins/day)

596.4 ± 419.6

168.2 ± 54.2

***

1.43

 MPA(mins/day)

36.9 ± 40.1

38.0 ± 22.7

 

0.03

 VPA (mins/day)

1.9 ± 5.8

3.0 ± 3.5

***

0.23

 Sedentary Behavior(mins/day)

311.1 ± 133.2

718.1 ± 108.4

***

3.35

IPAQ vs. accelerometers

 LPA (mins/day)

302.0 ± 197.4

168.2 ± 54.2

***

0.92

 MPA(mins/day)

44.1 ± 54.0

38.0 ± 22.7

 

0.15

 VPA (mins/day)

8.6 ± 19.2

3.0 ± 3.5

 

0.41

 Sedentary Behavior(mins/day)

570.0 ± 190.2

718.1 ± 108.4

***

0.96

LPA light physical activity, MPA moderate physical activity, VPA vigorous physical activity, 7DPAR 7-day physical activity recall, IPAQ International Physical Activity Questionnaire

*P < 0.05, **P < 0.01, ***P < 0.001

Table 3 demonstrates the correlation coefficients between activity measurements. There was a strong positive correlation of mean steps/day between accelerometers and pedometers (r = 0.72, P < 0.001). Between the 7DPAR and IPAQ, small to medium positive correlations were reported on time in LPA (r = 0.37, P < 0.001), MPA (r = 0.39, P < 0.001), VPA (r = 0.44, P < 0.001) and sedentary behavior (r = 0.45, P < 0.001). When comparing estimates between accelerometers and the 7DPAR, positive correlations were observed on LPA (r = 0.45, P < 0.001) and sedentary behavior (r = 0.20, P = 0.02). Estimates derived from IPAQ and accelerometers were positively correlated with each other for LPA (r = 0.24, P = 0.007) and VPA (r = 0.34, P < 0.001). The partial correlations, when stratified by gender, all of the aforementioned significant correlations remained significant after adjusting for age and percent body fat among females. There were fewer significant adjusted correlations among males.
Table 3

Spearman correlation coefficients and partial correlation coefficients between activity measurements

 

Accelerometers vs. pedometers

7DPAR vs. IPAQ

7DPAR vs. accelerometers

IPAQ vs. accelerometers

r

r

r

r

Unadjusted

 Steps (steps/day)

0.72***

 LPA (mins/day)

0.37***

0.45***

0.24**

 MPA(mins/day)

0.39***

0.04

0.05

 VPA (mins/day)

0.44***

0.12

0.34***

 Sedentary behavior(mins/day)

0.45***

0.20*

0.04

Adjusted for age and percent body fat

 Female

  Steps (steps/day)

0.70***

  LPA (mins/day)

0.45***

0.44***

0.29**

  MPA(mins/day)

0.35***

0.12

0.03

  VPA (mins/day)

0.41***

0.06

0.31**

  Sedentary behavior(mins/day)

0.48***

0.24*

0.04

 Male

  Steps (steps/day)

0.74***

  LPA (mins/day)

0.18

0.43 *

0.28

  MPA(mins/day)

0.58**

0.38

0.02

  VPA (mins/day)

0.57**

0.11

0.45*

  Sedentary behavior(mins/day)

0.39

0.14

0.09

LPA light physical activity, MPA moderate physical activity, VPA vigorous physical activity, 7DPAR 7-day physical activity recall, IPAQ international physical activity questionnaire

*P < 0.05, **P < 0.01, ***P < 0.001

When participants were dichotomized by higher or lower body fat group, stronger correlations between different types of activity modalities were observed among individuals with lower body fat than with higher body fat (Table 4). For the mean steps/day, the Spearman correlation coefficient between accelerometers and pedometers was 0.80 (P < 0.001) among the lower body fat group and 0.65 (P < 0.001) among the higher body fat group. Similarly, those with lower body fat (r = 0.42–0.50) had higher correlations between the 7DPAR and IPAQ for all levels of PA and sedentary behavior than those with higher body fat (r = 0.31–0.40). In addition, a stronger correlation between the 7DPAR and the accelerometer was also observed in LPA for the lower body fat group than the higher body fact group (lower body fat: r = 0.41, P < 0.001; higher body fat: r = 0.32, P < 0.001).
Table 4

Spearman correlation coefficients between activity measurements by low and high body fat

 

Accelerometers vs. pedometers

7DPAR vs. IPAQ

7DPAR vs. accelerometers

IPAQ vs. accelerometers

 

r

r

r

r

Low body fat group

 Steps (steps/day)

0.80***

 LPA (mins/day)

0.48***

0.41***

0.45***

 MPA(mins/day)

0.46***

0.03

–0.06

 VPA (mins/day)

0.42***

0.12

0.24

 Sedentary behavior(mins/day)

0.50***

0.21

0.06

High body fat group

 Steps (steps/day)

0.65***

 LPA (mins/day)

0.32**

0.32***

0.03

 MPA(mins/day)

0.31*

–0.07

0.11

 VPA (mins/day)

0.40***

0.10

0.11

 Sedentary behavior(mins/day)

0.34**

0.18

0.03

LPA light physical activity, MPA moderate physical activity, VPA vigorous physical activity, 7DPAR 7-day physical activity recall, IPAQ international physical activity questionnaire

*P < 0.05, **P < 0.01, ***P < 0.001

Discussion

The primary purpose of the study was to compare the four subjective and objective measures of PA and sedentary behavior in college students. A strong correlation of steps/day was reported between the two objective measures (pedometer and accelerometer). There were low-to-moderate positive correlations between the two self-report questionnaires (7DPAR and IPAQ) for LPA, MPA, VPA, and sedentary behavior [32] As compared to IPAQ, estimates in MPA and VPA based on 7DPAR were closer to those derived from accelerometers (criterion measure). For all comparisons, the correlations between all types of activity measures were higher among individuals with lower body fat than with higher body fat.

Comparisons Between Subjective Measures

Regarding the estimates from the two self-report questionnaires, the 7DPAR and IPAQ were comparable for MPA. The LPA was substantially higher based on the 7DPAR, while time in VPA and sedentary behavior were higher based on the IPAQ. These discrepancies in psychometric properties may be explained by how activity levels were assessed in the questionnaires. The 7DPAR required participants to recall types and intensities of activities from a list of activities in a time-use diary format, in which memories about daily activities were enhanced by providing cues on when and how the behavior occurred [21]. For the IPAQ, single-day summary estimations of time spent in different intensity levels were collected, and it has been suggested that such format is suitable for routine and structured activities [21]. The findings in this study revealed that for college students, the daily recall and estimates of MPA were consistent across the behavior-specific questionnaire (7DPAR) and the summary format questionnaire (IPAQ). Nevertheless, wider variations in LPA and sedentary behavior implied that it is challenging to determine the duration on low intensity activities. For instance, although descriptions of similar sedentary activities were provided in both the 7DPAR and the IPAQ, the estimation of sedentary time varied significantly due to different question designs. The summary format approach (IPAQ) resulted in higher estimates of sedentary time, while the behavior-specific approach (7DPAR) led to higher estimates of LPA. Another plausible explanation might be that the participants in the current study spent the majority of their time in LPA and sedentary behavior, leaving them a broader range of activities and time that needed to be recalled. Given that LPA and sedentary behavior are often unstructured (e.g. reading, walking, surfing internet), it is somewhat difficult to differentiate and determine the exact length of these activities.

Comparisons Between Subjective and Objective Measures

For comparisons between the questionnaires and accelerometers, both the 7DPAR and the IPAQ yielded similar estimates of MPA to those of the accelerometeric data. In addition, the estimates of VPA were comparable between the IPAQ and accelerometers. Interestingly, the self-report questionnaires yielded less time in sedentary behavior than the accelerometers. It is plausible that hip-worn accelerometers overestimate sedentary time due to its limitations in detecting motion of upper limb movement. Thus, certain upper body motions would be misclassified as sedentary behavior [45]. Another possible explanation is the challenge accelerometers face in distinguishing between sitting, lying, and standing still [3, 23]. Furthermore, it remained difficult to differentiate daytime napping from sedentary behavior. Although a diary log was utilized in this study to obtain participants’ sleep schedule and accelerometer wear time, napping time may have been recognized as sedentary time by accelerometers if participants forgot to report napping time in the log.

To date, few studies have evaluated the correlations between subjective and objective measures in assessing sedentary behavior. One study reported a strong association between the PAR and accelerometers in sedentary time (ActivPAL) [28], which is inconsistent with the current study. Such incongruence may be explained by the different types of accelerometers used, and the body locations where the devices were placed. Specifically, research by Matthews et al. [28] used the ActivPAL accelerometer, which was placed on the front of the participant’s thigh. Unlike the hip-worn accelerometers used in the current study, the thigh-worn accelerometers detect movements above the thigh. Thus, they tend to recognize more episodes of sitting, lying, standing, and sit-to-stand transitions [3, 15], resulting in a higher correlation with the PAR in evaluating sedentary behavior.

Consistent with prior studies, another major finding of this study was that there were discrepancies in PA assessed by accelerometer and self-report questionnaires based on body composition. Using accelerometers as criteria, we found that self-report measures could not properly assess PA among individuals with higher body fat. This finding has been soundly echoed by previous research [38, 43]. The present study indicated that there were weaker correlations between self-reported time and accelerometer-assessed time in individuals with higher body fat compared to those with lower body fat. Perhaps, individuals with higher body fat may demonstrate lower physical fitness levels and poorer motor skills [31], so they are more likely to subjectively report certain PAs as more intense than individuals with low body fat would report. For instance, participants with higher body fat may identify running at the speed of 5 m/h as vigorous intensity, while a low-percent body fat individual may rate his/her perceived exertion as moderate or light intensity for the same activity. On the other hand, activity measured by accelerometers, is objectively device-determined regardless of body fat consideration, and is less prone to be affected by human bias. Our findings highlight that there are greater discrepancies between subjective and objective instruments in estimating PA among individuals with higher body fat. Thus, body composition should be considered when comparing activity estimates between subjective and objective measurements. Accelerometers detect “absolute” energy expenditure without being affected by body weight, while self-report data take into consideration of “relative” intensities of behavior [43]. It is important to note that although we did not compare PA measures by overweight status, the fact that we were able to observe differences in correlations between the lower and the higher body fat group suggested that future research should consider the adjustment of body compositions.

Comparisons Between Objective Measures

In this study, pedometers were used as the criterion for daily step counts because of their acceptable reliability [5]. Similar to another study that found that step counts detected by accelerometers were lower than those detected by pedometers [5], we found that accelerometers underestimated step counts by 474 steps per day on average (7.7%). It is worth mentioning that the mean step count is between 5708 to 6182 steps/day in our study. According to the steps/PA guidelines in healthy adults [6], the Taiwanese college students in this study would be considered physically inactive (5000–7499 steps/day). More studies are needed to identify the facilitators of PA and sedentary behavior for college students, as university is a critical period for individuals to form lifestyle patterns, which may have a sustaining impact on health.

Strengths and Limitations

Among the limited studies that compare concordance between activity measures, these studies only assess activity levels for a short period of time or base their assessment on proxy measures. This is the first study that compared the estimates of PA and sedentary behavior across two objective measures and two self-report instruments simultaneously over the same 7 days. This design takes into consideration the habitual variability of individual activity behavior and provides data that are representative of individual activity patterns. Another key strength is the comparisons of sedentary time between objective monitors and self-report questionnaire. This topic is less discussed in the literature [43]. Limitations of the study include the small sample size, which may limit the statistical power. Second, participants were volunteers and may not be the representative of the entire college student populations. Third, the majority of the study sample was female, and the results may differ with a more diverse sample. Fourth, the Bland–Altman analysis was not applied to compare agreements across measurements, because the activity estimates in the current study did not meet the normality assumption.

Conclusion

Although objective monitors have been more widely used, self-report instruments are still dominant in large sample studies due to their low cost and usability. Discrepancies observed between accelerometers, pedometers, the 7DPAR, and the IPAQ may be because they capture different dimensions of PA, such as type, frequency, intensity, and type. It is recommended that estimates of steps and time in activities should not be used interchangeably. This is especially important for sedentary time, in which significant variations were found. Last but not the least, body composition should be considered when comparing activity estimates across measurements.

Notes

Acknowledgements

This research was supported by the MOST 102-2410-H-041-009, Taiwan, China. We would like to acknowledge the participants and graduate students who assisted in data collection.

References

  1. 1.
    Aadland E, Ylvisåker E. Reliability of objectively measured sedentary time and physical activity in adults. PLoS ONE. 2015;10(7):e0133296.PubMedPubMedCentralCrossRefGoogle Scholar
  2. 2.
    Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, O'Brien WL, Bassett DR Jr, Schmitz KH, Emplaincourt PO, Jacobs DR Jr, Leon AS. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc. 2000;32(9):S498–504.PubMedCrossRefPubMedCentralGoogle Scholar
  3. 3.
    Aminian S, Hinckson EA. Examining the validity of the ActivPAL monitor in measuring posture and ambulatory movement in children. Int J Behav Nutr Phys Act. 2012;9:119.PubMedPubMedCentralCrossRefGoogle Scholar
  4. 4.
    Arem H, Moore SC, Patel A, Hartge P, Berrington de Gonzalez A, Visvanathan K, Campbell PT, Freedman M, Weiderpass E, Adami HO, Linet MS, Lee IM, Matthews CE. Leisure time physical activity and mortality: a detailed pooled analysis of the dose-response relationship. JAMA. Intern Med. 2015;175(6):959–67.Google Scholar
  5. 5.
    Barriera TV, Tudor-Locke C, Champagne CM, Broyles ST, Johnson WD, Katzmarzyk PT. Comparison of GT3X accelerometer and YAMAX pedometer steps/day in a free-living sample of overweight and obese adults. J Sport Health Sci. 2013;10(2):263–70.Google Scholar
  6. 6.
    Bassett DR, Toth LP, LaMunion SR, Crouter SE. Step counting: a review of measurement considerations and health-related applications. Sports Med. 2017;47(7):1303–15.PubMedCrossRefPubMedCentralGoogle Scholar
  7. 7.
    Brown WJ, Miller YD, Miller R. Sitting time and work patterns as indicators of overweight and obesity in Australian adults. Int J Obes Relat Metab Disord. 2003;27(11):1340–6.PubMedCrossRefPubMedCentralGoogle Scholar
  8. 8.
    Carlson SA, Fulton JE, Pratt M, Yang Z, Adams EK. Inadequate physical activity and health care expenditures in the United States. Prog Cardiovasc Dis. 2015;57(4):315–23.PubMedCrossRefPubMedCentralGoogle Scholar
  9. 9.
    Caspersen CJ, Powell KE, Christenson GE. Physical activity, exercise, and physical fitness: definitions and distinctions for health-related research. Public Health Rep. 1985;100(2):126–31.PubMedPubMedCentralGoogle Scholar
  10. 10.
    Cleland CL, Hunter RF, Kee F, Cupples ME, Sallis JF, Tully MA. Validity of the global physical activity questionnaire (GPAQ) in assessing levels and change in moderate-vigorous physical activity and sedentary behaviour. BMC Public Health. 2014;14:1255.PubMedPubMedCentralCrossRefGoogle Scholar
  11. 11.
    Clemes SA, Biddle SJ. The use of pedometers for monitoring physical activity in children and adolescents: measurement considerations. J Phys Act Health. 2013;10(2):249–62.PubMedCrossRefPubMedCentralGoogle Scholar
  12. 12.
    Cohen J. Statistical power analysis for the behavioral sciences. Hillsdale: Lawrence Erlbaum; 1988.Google Scholar
  13. 13.
    Dinger MK, Behrens TK, Han JL. Validity and reliability of the international physical activity questionnaire in college students. Am J Health Educ. 2006;37:337–43.CrossRefGoogle Scholar
  14. 14.
    Dyrstad SM, Hansen BH, Holme IM, Anderssen SA. Comparison of self-reported versus accelerometer-measured physical activity. Med Sci Sports Exerc. 2014;46(1):99–106.PubMedCrossRefPubMedCentralGoogle Scholar
  15. 15.
    Edwardson CL, Winkler EAH, Bodicoat DH, Yates T, Davies MJ, Dunstan DW, Healy GN. Considerations when using the ActivPAL monitor in field-based research with adult populations. J Sport Health Sci. 2017;6(2):162–78.PubMedCrossRefPubMedCentralGoogle Scholar
  16. 16.
    Finger JD, Gisle L, Mimilidis H, Santos-Hoevener C, Kruusmaa EK, Matsi A, Oja L, Balarajan M, Gray M, Kratz AL, Lange C. How well do physical activity questions perform? A European cognitive testing study. Arch of Public Health. 2015;73:57.CrossRefGoogle Scholar
  17. 17.
    Hart TL, Ainsworth BE, Tudor-Locke C. Objective and subjective measures of sedentary behavior and physical activity. Med Sci Sports Exerc. 2011;43(3):449–56.PubMedCrossRefGoogle Scholar
  18. 18.
    Hayden-Wade HA, Coleman KJ, Sallis JF, Armstrong C. Validation of the telephone and in-person interview versions of the 7-day PAR. Med Sci Sports Exerc. 2003;35:801–9.PubMedCrossRefGoogle Scholar
  19. 19.
    Health Promotion Administration, Ministry of Health and Welfare. Assessing your weight. https://www.hpa.gov.tw/Pages/Detail.aspx?nodeid=542&pid=705 (2016).
  20. 20.
    Hu FB, Li TY, Colditz GA, Willett WC, Manson JE. Television watching and other sedentary behaviors in relation to risk of obesity and type 2 diabetes mellitus in women. JAMA. 2003;289(14):1785–91.PubMedCrossRefPubMedCentralGoogle Scholar
  21. 21.
    Johnson-Kozlow M, Sallis JF, Gilpin EA, Rock CL, Pierce JP. Comparative validation of the IPAQ and the 7-day PAR among women diagnosed with breast cancer. Int J Behav Nutr Phys Act. 2006;3:7.PubMedPubMedCentralCrossRefGoogle Scholar
  22. 22.
    Kelly LA, McMillan DE, Anderson A, Fippinger M, Fillerup G, Rider J. Validity of Actigraphs uniaxial and triaxial accelerometers for assessment of physical activity in adults in laboratory conditions. BMC Med Phys. 2013;13(1):5.PubMedPubMedCentralCrossRefGoogle Scholar
  23. 23.
    Lee IM, Shiroma EJ. Using accelerometers to measure physical activity in large-scale epidemiological studies: issues and challenges. Br J Sports Med. 2014;48(3):197–201.PubMedCrossRefPubMedCentralGoogle Scholar
  24. 24.
    Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT, Lancet Physical Activity Series Working Group. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet. 2012;380(9838):219–29.PubMedPubMedCentralCrossRefGoogle Scholar
  25. 25.
    Liou YM, Jwo CJ, Yao KG, Chiang LC, Huang LH. Selection of appropriate Chinese terms to represent intensity and types of physical activity terms for use in the Taiwan version of IPAQ. J Nurs Res. 2008;16(4):252–63.PubMedCrossRefPubMedCentralGoogle Scholar
  26. 26.
    Lubans DR, Plotnikoff RC, Miller A, Scott JJ, Thompson D, Tudor-Locke C. Using pedometers for measuring and increasing physical activity in children and adolescents: the next step. Am J Lifestyle Med. 2014;9(6):418–27.CrossRefGoogle Scholar
  27. 27.
    Magistro D, Brustio PR, Ivaldi M, Esliger DW, Zecca M, Rainoldi A, Boccia G. Validation of the ADAMO care watch for step counting in older adults. PLoS ONE. 2018;13(2):e0190753.PubMedPubMedCentralCrossRefGoogle Scholar
  28. 28.
    Matthews CE, Keadle SK, Sampson J, Lyden K, Bowles HR, Moore SC, Libertine A, Freedson PS, Fowke JH. Validation of a previous-day recall measure of active and sedentary behaviors. Med Sci Sports Exerc. 2013;45(8):1629–38.PubMedPubMedCentralCrossRefGoogle Scholar
  29. 29.
    Matthews CE, Welk GJ. Use of self-report instruments to assess physical activity. In: Welk GJ, editor. Physical activity assessments for health-related research. Champaign: Human Kinetics; 2002.Google Scholar
  30. 30.
    Metcalf KM, Baquero BI, Coronado Garcia ML, Francis SL, Janz KF, Laroche HH, Sewell DK. Calibration of the global physical activity questionnaire to accelerometry measured physical activity and sedentary behavior. BMC Public Health. 2018;18:412.PubMedPubMedCentralCrossRefGoogle Scholar
  31. 31.
    Morano M, Colella D, Capranica L. Body image, perceived and actual physical abilities in normal-weight and overweight boys involved in individual and team sports. J Sports Sci. 2011;29(4):355–62.PubMedCrossRefPubMedCentralGoogle Scholar
  32. 32.
    Mukaka MM. Statistics corner: a guide to appropriate use of correlation coefficient in medical research. Malawi Med J. 2012;24(3):69–71.PubMedPubMedCentralGoogle Scholar
  33. 33.
    Pate RR, O’Neill JR, Lobelo F. The evolving definition of “sedentary”. Exerc Sport Sci Rev. 2008;36(4):173–8.PubMedCrossRefPubMedCentralGoogle Scholar
  34. 34.
    Pate RR, Ross R, Dowda M, Trost SG, Sirard JR. Validation of a 3-day physical activity recall instrument in female youth. Pediatr Exerc Sci. 2003;15(3):257–65.CrossRefGoogle Scholar
  35. 35.
    Rääsk T, Mäestu J, Lätt E, Jürimäe J, Jürimäe T, Vainik U, Konstabel K. Comparison of IPAQ-SF and two other physical activity questionnaires with accelerometer in adolescent boys. PLoS ONE. 2017;12(1):e0169527.PubMedPubMedCentralCrossRefGoogle Scholar
  36. 36.
    Sallis JF, Saelens BE. Assessment of physical activity by self-report: status, limitations, and future directions. Res Q Exerc Sport. 2000;71(Suppl 2):1–14.PubMedCrossRefPubMedCentralGoogle Scholar
  37. 37.
    Schmidt MD, Cleland VJ, Thomson RJ, Dwyer T, Venn AJ. A comparison of subjective and objective measures of physical activity and fitness in identifying associations with cardiometabolic risk factors. Ann Epidemiol. 2008;18(5):378–86.PubMedCrossRefPubMedCentralGoogle Scholar
  38. 38.
    Slootmaker SM, Schuit AJ, Chinapaw MJ, Seidell JC, van Mechelen W. Disagreement in physical activity assessed by accelerometer and self-report in subgroups of age, gender, education and weight status. Int J Behav Nutr Phys Act. 2009;6:17.PubMedPubMedCentralCrossRefGoogle Scholar
  39. 39.
    Troiano RP, Berrigan D, Dodd KW, Mâsse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. 2008;40(1):181–8.PubMedCrossRefPubMedCentralGoogle Scholar
  40. 40.
    U.S. Department of Health & Human Services. 2008 Physical activity guidelines advisory committee report. (2008).Google Scholar
  41. 41.
    van der Ploeg HP, Tudor-Locke C, Marshall AL, Craig C, Hagströmer M, Sjöström M, Bauman A. Reliability and validity of the international physical activity questionnaire for assessing walking. Res Q Exerc Sport. 2010;81(1):97–101.PubMedCrossRefPubMedCentralGoogle Scholar
  42. 42.
    van Poppel MN, Chinapaw MJ, Mokkink LB, van Mechelen W, Terwee CB. Physical activity questionnaires for adults: a systematic review of measurement properties. Sports Med. 2010;40(7):565–600.PubMedCrossRefPubMedCentralGoogle Scholar
  43. 43.
    Welk GJ, Kim Y, Stanfill B, Osthus DA, Calabro MA, Nusser SM, Carriquiry A. Validity of 24-h physical activity recall: physical activity measurement survey. Med Sci Sports Exerc. 2014;46(10):2014–24.PubMedPubMedCentralCrossRefGoogle Scholar
  44. 44.
    Wang C, Chen P, Zhuang J. Validity and reliability of international physical activity questionnaire-short form in Chinese youth. Res Q Exerc Sport. 2013;84(Suppl 2):S80–6.PubMedCrossRefPubMedCentralGoogle Scholar
  45. 45.
    Yang CC, Hsu YL. A review of accelerometry-based wearable motion detectors for physical activity monitoring. Sensors (Basel). 2010;10(8):7772–88.CrossRefGoogle Scholar

Copyright information

© Beijing Sport University 2020

Authors and Affiliations

  1. 1.School of Community Health SciencesUniversity of Nevada, RenoRenoUSA
  2. 2.Department of Hospital and Health Care AdministrationChia Nan University of Pharmacy and ScienceTainan, TaiwanChina
  3. 3.Department of Health Promotion and Health EducationNational Taiwan Normal UniversityTaipei, TaiwanChina
  4. 4.Department of Physical Therapy, College of MedicineNational Cheng Kung UniversityTainan, TaiwanChina
  5. 5.Department of Medical Laboratory Science and BiotechnologyKaohsiung Medical UniversityKaohsiung, TaiwanChina

Personalised recommendations